From bwalton at opencsw.org Tue Jun 1 02:43:51 2010 From: bwalton at opencsw.org (Ben Walton) Date: Mon, 31 May 2010 20:43:51 -0400 Subject: [csw-maintainers] old gcc4* packages in catalog Message-ID: <1275352845-sup-6422@pinkfloyd.chass.utoronto.ca> Hi Phil, I just noticed that we seem to have overlapping gcc4 packages in the catalog. This is likely due to a name change that slipped through unnoticed, but maybe there is another reason. Either way, two packages are shipping overlapping files: http://www.opencsw.org/packages/gcc4g95 http://www.opencsw.org/packages/gcc4gfortran Given that this has been in the wild for some time now, it may be best to release a dummy gcc4g95 package with a dependency on gcc4gfortran, but correct this if you feel otherwise. Thanks -Ben -- Ben Walton Systems Programmer - CHASS University of Toronto C:416.407.5610 | W:416.978.4302 From phil at bolthole.com Tue Jun 1 03:51:27 2010 From: phil at bolthole.com (Philip Brown) Date: Mon, 31 May 2010 18:51:27 -0700 Subject: [csw-maintainers] gtk - dump 64bit for now? Message-ID: <20100601015127.GA76098@bolthole.com> So.... I'm looking into a "quickie" redo of libgtk , to get us into the 21st century. I see the older package has 64bit libs. I also see that unlike many other 64bit capable things, it's going to be a pain for me to do a nice small clean implementation of that. BUT... in checking out our installed /opt/csw/lib/sparcv9 on our build machines... it looks like there's no other libs that actually USE the 64bit libgtk. So I'm thinking of dropping it to keep life simple. In the future (possibly in the near future) it is entirely possible for someone to take over libgtk and provide 64bit libs again if desired. (heck, i would greatly prefer that :) But for the near term, I think it's more important for us to get a modern libgtk, QUICKLY, rather than wait another few weeks for someone else. Opinions? Can I get a yea or a nay from anyone else out there on this? From bwalton at opencsw.org Tue Jun 1 04:10:11 2010 From: bwalton at opencsw.org (Ben Walton) Date: Mon, 31 May 2010 22:10:11 -0400 Subject: [csw-maintainers] packages with conflicting files Message-ID: <1275358157-sup-7239@pinkfloyd.chass.utoronto.ca> Hi All, I've just discovered a file conflict between CSWlibev and CSWlibevent-devel. Thoughts on resolving this? Thanks -Ben -- Ben Walton Systems Programmer - CHASS University of Toronto C:416.407.5610 | W:416.978.4302 From phil at bolthole.com Tue Jun 1 04:28:52 2010 From: phil at bolthole.com (Philip Brown) Date: Mon, 31 May 2010 19:28:52 -0700 Subject: [csw-maintainers] packages with conflicting files In-Reply-To: <1275358157-sup-7239@pinkfloyd.chass.utoronto.ca> References: <1275358157-sup-7239@pinkfloyd.chass.utoronto.ca> Message-ID: they are technically different btw. different implementation of same API if recall. not okay for them to overlap though On Monday, May 31, 2010, Ben Walton wrote: > > Hi All, > > I've just discovered a file conflict between CSWlibev and > CSWlibevent-devel. ?Thoughts on resolving this? > > Thanks > -Ben > -- > Ben Walton > Systems Programmer - CHASS > University of Toronto > C:416.407.5610 | W:416.978.4302 > > _______________________________________________ > maintainers mailing list > maintainers at lists.opencsw.org > https://lists.opencsw.org/mailman/listinfo/maintainers > .:: This mailing list's archive is public. ::. > From phil at bolthole.com Tue Jun 1 07:22:37 2010 From: phil at bolthole.com (Philip Brown) Date: Mon, 31 May 2010 22:22:37 -0700 Subject: [csw-maintainers] old gcc4* packages in catalog In-Reply-To: <1275352845-sup-6422@pinkfloyd.chass.utoronto.ca> References: <1275352845-sup-6422@pinkfloyd.chass.utoronto.ca> Message-ID: sounds good to me but check with current maintainer first. there might have been a reason for both On Monday, May 31, 2010, Ben Walton wrote: > > Hi Phil, > > I just noticed that we seem to have overlapping gcc4 packages in the > catalog. ?This is likely due to a name change that slipped through > unnoticed, but maybe there is another reason. ?Either way, two > packages are shipping overlapping files: > > http://www.opencsw.org/packages/gcc4g95 > http://www.opencsw.org/packages/gcc4gfortran > > Given that this has been in the wild for some time now, it may be best > to release a dummy gcc4g95 package with a dependency on gcc4gfortran, > but correct this if you feel otherwise. > > Thanks > -Ben > > -- > Ben Walton > Systems Programmer - CHASS > University of Toronto > C:416.407.5610 | W:416.978.4302 > > _______________________________________________ > maintainers mailing list > maintainers at lists.opencsw.org > https://lists.opencsw.org/mailman/listinfo/maintainers > .:: This mailing list's archive is public. ::. > From maciej at opencsw.org Tue Jun 1 13:51:03 2010 From: maciej at opencsw.org (Maciej (Matchek) Blizinski) Date: Tue, 1 Jun 2010 12:51:03 +0100 Subject: [csw-maintainers] cswclassutils: location of template init scripts In-Reply-To: References: Message-ID: No dia 1 de Fevereiro de 2010 21:31, Philip Brown escreveu: > I think it would be better for our users, if we standardize on a > location under /opt/csw > For example, /opt/csw/etc/init.d Let's consider the following scenario: There's a global zone, and a non-global, sparse zone which shares /opt. During package installation, the global zone is processed first. The class action script copies the init file to /opt/csw/etc/init.d/foo, generates the SMF manifest, imports it and starts the service. Then, pkgadd proceeds to the non-global zone. The /opt filesystem is shared, so the /opt/csw/etc/init.d/foo is already there. The rest of this scenario is left as an exercise for the reader. From phil at bolthole.com Tue Jun 1 15:04:47 2010 From: phil at bolthole.com (Philip Brown) Date: Tue, 1 Jun 2010 06:04:47 -0700 Subject: [csw-maintainers] cswclassutils: location of template init scripts In-Reply-To: References: Message-ID: On Tuesday, June 1, 2010, Maciej (Matchek) Blizinski wrote: > No dia 1 de Fevereiro de 2010 21:31, Philip Brown escreveu: >> I think it would be better for our users, if we standardize on a >> location under /opt/csw >> For example, /opt/csw/etc/init.d > > Let's consider the following scenario: > > There's a global zone, and a non-global, sparse zone which shares > /opt. ?During package installation, the global zone is processed > first. ?The class action script copies the init file to > /opt/csw/etc/init.d/foo,... err... that's not what it's supposed to do. the script is supposed to copy it FROM there to whereever it needs From maciej at opencsw.org Tue Jun 1 17:06:40 2010 From: maciej at opencsw.org (Maciej (Matchek) Blizinski) Date: Tue, 1 Jun 2010 16:06:40 +0100 Subject: [csw-maintainers] cswclassutils: location of template init scripts In-Reply-To: References: Message-ID: No dia 1 de Junho de 2010 14:04, Philip Brown escreveu: > On Tuesday, June 1, 2010, Maciej (Matchek) Blizinski wrote: >> No dia 1 de Fevereiro de 2010 21:31, Philip Brown escreveu: >>> I think it would be better for our users, if we standardize on a >>> location under /opt/csw >>> For example, /opt/csw/etc/init.d >> >> Let's consider the following scenario: >> >> There's a global zone, and a non-global, sparse zone which shares >> /opt. ?During package installation, the global zone is processed >> first. ?The class action script copies the init file to >> /opt/csw/etc/init.d/foo,... > > err... that's not what it's supposed to do. the script is supposed to > copy it FROM there to whereever it needs Not sure what you mean. Right now, in the code[1] there is: while read src dest do (...) /usr/bin/cp $src $dest || exit 2 (...) done Did you mean the following? while read src dest do (...) /usr/bin/cp $dest $wherever_it_needs (...) done I think it would result in "cp: cannot access /opt/csw/etc/init.d/foo". In any case, "/usr/bin/cp $src $dest" is what it is doing right now, and I've presented an actual scenario. Any thoughts on it? [1] http://sourceforge.net/apps/trac/gar/browser/csw/mgar/pkg/cswclassutils/trunk/files/CSWcswclassutils.i.cswinitsmf From phil at bolthole.com Tue Jun 1 17:25:04 2010 From: phil at bolthole.com (Philip Brown) Date: Tue, 1 Jun 2010 08:25:04 -0700 Subject: [csw-maintainers] cswclassutils: location of template init scripts In-Reply-To: References: Message-ID: On Tue, Jun 1, 2010 at 8:06 AM, Maciej (Matchek) Blizinski wrote: >> err... that's not what it's supposed to do. the script is supposed to >> copy it FROM there to whereever it needs > > Not sure what you mean. ?Right now, in the code[1] there is: > > while read src dest > do > ?(...) > ?/usr/bin/cp $src $dest || exit 2 > ?(...) > done Ah. forgot about that. There are two aspects to class action scripts, speaking generally. 1. optional "special processing" done on the file. 2. mandatory "hey, install this" stuff, that every class action script has to do. I was thinking of the #1 part. Whereas you were rightly pointing out a bug in #2. What we SHOULD be doing, is something like. if test -f $dest ; then if cmp $src $dest ; then echo $dest already exists - continuing else echo ERROR: $dest exists and is not correct file exit 1 #? fi else cp $src $dest fi # and then later, do the "copy FROM $dest, to appropriate place(s)" as we presumably # do already. From maciej at opencsw.org Tue Jun 1 17:40:37 2010 From: maciej at opencsw.org (Maciej (Matchek) Blizinski) Date: Tue, 1 Jun 2010 16:40:37 +0100 Subject: [csw-maintainers] cswclassutils: location of template init scripts In-Reply-To: References: Message-ID: No dia 1 de Junho de 2010 16:25, Philip Brown escreveu: > On Tue, Jun 1, 2010 at 8:06 AM, Maciej (Matchek) Blizinski > wrote: > >>> err... that's not what it's supposed to do. the script is supposed to >>> copy it FROM there to whereever it needs >> >> Not sure what you mean. ?Right now, in the code[1] there is: >> >> while read src dest >> do >> ?(...) >> ?/usr/bin/cp $src $dest || exit 2 >> ?(...) >> done > > > Ah. forgot about that. > > There are two aspects to class action scripts, speaking generally. > > 1. optional "special processing" done on the file. > > 2. mandatory "hey, install this" stuff, that every class action script > has to do. > > I was thinking of the #1 part. Whereas you were rightly pointing out a > bug in #2. > > What we SHOULD be doing, is something like. > > if test -f $dest ; then > ? if cmp $src $dest ; then > ? ? ?echo $dest already exists - continuing > ? else > ? ? ?echo ERROR: $dest exists and is not correct file > ? ? ?exit 1 #? > ? fi > else > ?cp $src $dest > fi > > # and then later, do the "copy FROM $dest, to appropriate place(s)" as > we presumably > # do already. I think I know what you think it is, and it's not that. ;-) You need to think about what happens in the case of the "none" class, and then extrapolate it to our CAS. From phil at bolthole.com Tue Jun 1 17:50:03 2010 From: phil at bolthole.com (Philip Brown) Date: Tue, 1 Jun 2010 08:50:03 -0700 Subject: [csw-maintainers] cswclassutils: location of template init scripts In-Reply-To: References: Message-ID: On Tue, Jun 1, 2010 at 8:40 AM, Maciej (Matchek) Blizinski wrote: > No dia 1 de Junho de 2010 16:25, Philip Brown escreveu: >> What we SHOULD be doing, is something like. >> >> if test -f $dest ; then >> ? if cmp $src $dest ; then >> ? ? ?echo $dest already exists - continuing >> ? else >> ? ? ?echo ERROR: $dest exists and is not correct file >> ? ? ?exit 1 #? >> ? fi >> else >> ?cp $src $dest >> fi >> >> # and then later, do the "copy FROM $dest, to appropriate place(s)" as >> we presumably >> # do already. > > I think I know what you think it is, and it's not that. ;-) > > You need to think about what happens in the case of the "none" class, > and then extrapolate it to our CAS. > I think what I just wrote, should work just fine. Unless the class action script does not get called in a subzone, when the file already exists. If you are concerned for other reasons, please give specifics as to why, rather than suggesting "extrapolation". PS: the remove part of cswinitsmf would have to be adjusted also, to not do "rm", when not needed. From maciej at opencsw.org Tue Jun 1 18:05:52 2010 From: maciej at opencsw.org (Maciej (Matchek) Blizinski) Date: Tue, 1 Jun 2010 17:05:52 +0100 Subject: [csw-maintainers] cswclassutils: location of template init scripts In-Reply-To: References: Message-ID: No dia 1 de Junho de 2010 16:50, Philip Brown escreveu: > I think what I just wrote, should work just fine. Unless the class > action script does not get called in a subzone, when the file already > exists. Bingo. The class action script doesn't get called in a non-global zone, it doesn't generate the manifest, nor it imports it. As the result, the service only runs in the global zone. So our current policy (init scripts below /opt) together with our CAS result in broken packages. Suggested course of action? From maciej at opencsw.org Tue Jun 1 18:17:31 2010 From: maciej at opencsw.org (Maciej (Matchek) Blizinski) Date: Tue, 1 Jun 2010 17:17:31 +0100 Subject: [csw-maintainers] cswclassutils: location of template init scripts In-Reply-To: References: Message-ID: No dia 1 de Junho de 2010 17:05, Maciej (Matchek) Blizinski escreveu: > No dia 1 de Junho de 2010 16:50, Philip Brown escreveu: >> I think what I just wrote, should work just fine. Unless the class >> action script does not get called in a subzone, when the file already >> exists. > > Bingo. ?The class action script doesn't get called in a non-global > zone, it doesn't generate the manifest, nor it imports it. ?As the > result, the service only runs in the global zone. To be more specific, looks like the class action scripts don't get called for files that are located on the inherited (read-only) file systems. (The file existence per se doesn't matter.) From phil at bolthole.com Tue Jun 1 18:22:49 2010 From: phil at bolthole.com (Philip Brown) Date: Tue, 1 Jun 2010 09:22:49 -0700 Subject: [csw-maintainers] cswclassutils: location of template init scripts In-Reply-To: References: Message-ID: On Tue, Jun 1, 2010 at 9:05 AM, Maciej (Matchek) Blizinski wrote: > No dia 1 de Junho de 2010 16:50, Philip Brown escreveu: >> I think what I just wrote, should work just fine. Unless the class >> action script does not get called in a subzone, when the file already >> exists. > > Bingo. ?The class action script doesn't get called in a non-global > zone, it doesn't generate the manifest, nor it imports it. ?As the > result, the service only runs in the global zone. > > So our current policy (init scripts below /opt) together with our CAS > result in broken packages. ?Suggested course of action? > Gaaaah. ugly.not sure. So I'll state some goals, and ask for suggestions. Optimal Requirements: 1. CAS script must run in every zone 2. An init file must be left in"global" area under /opt/csw somewhere. Thoughts: CAS script will only run, if (class tagged) file does not exist in the zone. This previously "worked", because by default, files under /etc are counted as local-to-zone, so even though /opt/csw gets mouted shared-read-only, /etc/opt/csw does not. /opt/csw/etc/init.d namespace is now somewhat "polluted". We might attempt to clean it up.. or to somewhat ignore it from now on. OOO, I Have An Idea! We dont actually HAVE to do the "cp". in which case, CAS will always get triggered! Details: /opt/csw/etc/virtual (or /opt/csw/etc/cas, if preferred) f cswinitsmf /opt/csw/etc/virtual/cswinitscripthere blahblah Then.. CAS does the following: 1. a copy to /opt/csw/etc/init.d/cswinitscript, IF not already present. 2. either symlinks from /etc/rc to #1, or use SMF, as appropriate 3. calls "installf", to "register" /opt/csw/etc/virtual/cswinitscript.. but does NOT create the file in the actual filesystem Thus, the CAS should be triggered for each zone alsol Muahahahah..... If we go for the "ignore, and choose new clean one", perhaps something like the following: Perhaps we need a new global-space area to handle this. From phil at bolthole.com Tue Jun 1 18:24:45 2010 From: phil at bolthole.com (Philip Brown) Date: Tue, 1 Jun 2010 09:24:45 -0700 Subject: [csw-maintainers] cswclassutils: location of template init scripts In-Reply-To: References: Message-ID: ERm... the prior email had editing trash at the end of it. Ignore everything after the "Muahahaha" ;-) [ obviously, I need some evil henchmen to clean up after me :-D ] From phil at bolthole.com Tue Jun 1 18:40:49 2010 From: phil at bolthole.com (Philip Brown) Date: Tue, 1 Jun 2010 09:40:49 -0700 Subject: [csw-maintainers] testing9s.bo.opencsw.org Message-ID: FYI: I'm going to be wreaking havok on testing9s. In the unlikelyi event you were planning to use it today... I'd suggest making other plans :) From maciej at opencsw.org Tue Jun 1 18:41:07 2010 From: maciej at opencsw.org (Maciej (Matchek) Blizinski) Date: Tue, 1 Jun 2010 17:41:07 +0100 Subject: [csw-maintainers] cswclassutils: location of template init scripts In-Reply-To: References: Message-ID: No dia 1 de Junho de 2010 17:22, Philip Brown escreveu: > Gaaaah. ugly.not sure. So I'll state some goals, and ask for suggestions. > > Optimal Requirements: > ?1. CAS script must run in every zone > ?2. An init file must be left in"global" area under /opt/csw somewhere. > > Thoughts: > ?CAS script will only run, if (class tagged) file does not exist in the zone. I think it depends on the path of the $dest file, not the existence of the $dest file. Which means... > We dont actually HAVE to do the "cp". in which case, CAS will always > get triggered! ...that the above won't work. To be sure, we could just test it, by writing a class that doesn't copy the file, but leaves a log of activity. I'd bet my money on CAS not called for inherited filesystems. From phil at bolthole.com Tue Jun 1 19:43:30 2010 From: phil at bolthole.com (Philip Brown) Date: Tue, 1 Jun 2010 10:43:30 -0700 Subject: [csw-maintainers] cswclassutils: location of template init scripts In-Reply-To: References: Message-ID: On Tue, Jun 1, 2010 at 9:41 AM, Maciej (Matchek) Blizinski wrote: > No dia 1 de Junho de 2010 17:22, Philip Brown escreveu: >> >> We dont actually HAVE to do the "cp". in which case, CAS will always >> get triggered! > > ...that the above won't work. ?To be sure, we could just test it, by > writing a class that doesn't copy the file, but leaves a log of > activity. ?I'd bet my money on CAS not called for inherited > filesystems. > well... its even worse than that. pkgadd insists that the file get created. Installing class ... calling installf to file /usr/lib/fakefilehere [ verifying class ] ERROR: attribute verification of failed pathname does not exist back to the drawing board... From phil at bolthole.com Tue Jun 1 19:49:44 2010 From: phil at bolthole.com (Philip Brown) Date: Tue, 1 Jun 2010 10:49:44 -0700 Subject: [csw-maintainers] cswclassutils: location of template init scripts In-Reply-To: References: Message-ID: On Tue, Jun 1, 2010 at 9:41 AM, Maciej (Matchek) Blizinski wrote: > > ...that the above won't work. ?To be sure, we could just test it, by > writing a class that doesn't copy the file, but leaves a log of > activity. ?I'd bet my money on CAS not called for inherited > filesystems. > interestingly, though... the CAS **does** get called. Just, with a null list of files :-/ ## Installing package in zone .. [blah blah] 1 package pathname is already properly installed. Installing testpkg as ## Installing part 1 of 1. i.testclass: Installing class ... [ verifying class ] So I guess you still lose money. Half of it? :-D From maciej at opencsw.org Tue Jun 1 20:16:51 2010 From: maciej at opencsw.org (Maciej (Matchek) Blizinski) Date: Tue, 1 Jun 2010 19:16:51 +0100 Subject: [csw-maintainers] cswclassutils: location of template init scripts In-Reply-To: References: Message-ID: No dia 1 de Junho de 2010 18:49, Philip Brown escreveu: > interestingly, though... the CAS **does** get called. > Just, with a null list of files :-/ > > ## Installing package in zone > ?.. [blah blah] > > ? 1 package pathname is already properly installed. > > Installing testpkg as > > ## Installing part 1 of 1. > i.testclass: Installing class ... > [ verifying class ] > > > So I guess you still lose money. > Half of it? :-D I say I win! :-) It depends on how you interpret what I said. The exact phrasing was "CAS not called for inherited filesystems" which would be a shorthand for "CAS not processing files with paths on the inherited filesystems". I'd lose if I said "CAS not called at all when there are no files outside inherited filesystems". But you see that there are on-screen messages saying that it's processing the cswinitsmf class. In any case, I've submitted the cups package with the init script in /etc/opt/csw/init.d, until we figure how do to it better. I've removed the initfile-wrong-location check from checkpkg. I'm thinking of putting in a check that throws a message for cswinitsmf class files under /opt - we now know that it guarantees a broken package. What do you think? From rupert at opencsw.org Tue Jun 1 21:04:37 2010 From: rupert at opencsw.org (rupert THURNER) Date: Tue, 1 Jun 2010 21:04:37 +0200 Subject: [csw-maintainers] old gcc4* packages in catalog In-Reply-To: <1275352845-sup-6422@pinkfloyd.chass.utoronto.ca> References: <1275352845-sup-6422@pinkfloyd.chass.utoronto.ca> Message-ID: On Tue, Jun 1, 2010 at 02:43, Ben Walton wrote: > > Hi Phil, > > I just noticed that we seem to have overlapping gcc4 packages in the > catalog. ?This is likely due to a name change that slipped through > unnoticed, but maybe there is another reason. ?Either way, two > packages are shipping overlapping files: > > http://www.opencsw.org/packages/gcc4g95 > http://www.opencsw.org/packages/gcc4gfortran > > Given that this has been in the wild for some time now, it may be best > to release a dummy gcc4g95 package with a dependency on gcc4gfortran, > but correct this if you feel otherwise. > could this be the reason that cmake-2.8.1 has a failing fortran test ? rupert. From bwalton at opencsw.org Tue Jun 1 21:15:40 2010 From: bwalton at opencsw.org (Ben Walton) Date: Tue, 01 Jun 2010 15:15:40 -0400 Subject: [csw-maintainers] old gcc4* packages in catalog In-Reply-To: References: <1275352845-sup-6422@pinkfloyd.chass.utoronto.ca> Message-ID: <1275419668-sup-2085@pinkfloyd.chass.utoronto.ca> Excerpts from rupert THURNER's message of Tue Jun 01 15:04:37 -0400 2010: > could this be the reason that cmake-2.8.1 has a failing fortran test > ? Possibly. It would require forcing installation of one package when the other was already there though as there is some file conflict. Thanks -Ben -- Ben Walton Systems Programmer - CHASS University of Toronto C:416.407.5610 | W:416.978.4302 From phil at bolthole.com Tue Jun 1 21:14:17 2010 From: phil at bolthole.com (Philip Brown) Date: Tue, 1 Jun 2010 12:14:17 -0700 Subject: [csw-maintainers] cswclassutils: location of template init scripts In-Reply-To: References: Message-ID: On Tue, Jun 1, 2010 at 11:16 AM, Maciej (Matchek) Blizinski wrote: > > In any case, I've submitted the cups package with the init script in > /etc/opt/csw/init.d, until we figure how do to it better. > > I've removed the initfile-wrong-location check from checkpkg. I'm > thinking of putting in a check that throws a message for cswinitsmf > class files under /opt - we now know that it guarantees a broken > package. What do you think? I think we should use "cups needs an update" as motivation for "lets figure out how to do it best,then fix cups to use it". My thoughts on potential fixes so far are two fold. either: a) We partially revert to prior behaviour: we say that init scripts should be declared in the template as in /etc/opt/init.d/foo... but that the class action script will and must copy it to /opt/csw/init.d/foo if not already there. or b) we figure out a way to have the class action script always call itself "appropriately" somehow. Which we COULD technically do. If we define the new actions as: "init scripts always go in /opt/csw/etc/init.d/$PKGINST". Then the class action script, when called, copies files given on stdin in normal behaviour... (and ignores, if destination target already there) but then after that, ignores the list, and does its usual processin on "all files in /opt/csw/etc/init.d/$PKGINST". This second approach would have one very definite advantage, in that we would then no longer have to write a separate script for the case of, "Hey, I just copied over (or NFS mounted, or whatever) /opt/csw to a bare machine... how do I get any init scripts taken care of?)" You just run the CAS with PKGINST set, and it Does The Right Thing. and/or it would be trivial to provide a utility wrapper consisting of #!/bin/sh cd /opt/csw/etc/init.d for p in CSW* ; do PKGINST=$p /usr/sadm/install/scripts/i.cswinitsmf ; done The drawback, would be in converting existing packages that use the old style, to the new style. Would be nice to have some sort of middle ground somehow... From phil at bolthole.com Tue Jun 1 21:23:18 2010 From: phil at bolthole.com (Philip Brown) Date: Tue, 1 Jun 2010 12:23:18 -0700 Subject: [csw-maintainers] wierd problem with unresolved symbols, gtk, firefox, dynamic loader Message-ID: So.... I have a new libgtk, that seems to be relatively good. except that its having mysterous unresolved symbol problems with firefox. ld.so.1: firefox-bin: fatal: relocation error: file /opt/csw/lib/sparcv8/libgtk-x11-2.0.so.0: symbol gdk_font_equal: referenced symbol not found ld.so.1: firefox-bin: fatal: relocation error: file /opt/csw/mozilla/firefox/lib/libxul.so: symbol gtk_micro_version: referenced symbol not found I use the older version of libgtk, it works fine. I swap out nothing but CSWgtk to new lib, and it blows up with many errors like the above. yet here's the thing: THE SYMBOLS ARE IN THE GTK/GDK libraries! its just that for some reason, (probably to do with the firefox stuff being defined with LAZYLOAD flags), libgtk is not being pulled in for libxul, and libgdk is not being pulled in for libgtk. Yet the wierd thing is... when I use LD_DEBUG=files... both libgdk and libgtk, have ALREADY been loaded into memory!! ?? !! Can anyone shed some light on this mess? for the record: $ ldd /opt/csw/mozilla/firefox/lib/libxul.so|grep libg libgobject-2.0.so.0 => /opt/csw/lib/sparcv8/libgobject-2.0.so.0 libglib-2.0.so.0 => /opt/csw/lib/sparcv8/libglib-2.0.so.0 libgtk-x11-2.0.so.0 => /opt/csw/lib/sparcv8/libgtk-x11-2.0.so.0 libgdk-x11-2.0.so.0 => /opt/csw/lib/sparcv8/libgdk-x11-2.0.so.0 libgdk_pixbuf-2.0.so.0 => /opt/csw/lib/sparcv8/libgdk_pixbuf-2.0.so.0 .... so its not like the linker doesnt know how to find libs, for that particular library. ?? From maciej at opencsw.org Wed Jun 2 11:16:13 2010 From: maciej at opencsw.org (Maciej (Matchek) Blizinski) Date: Wed, 2 Jun 2010 10:16:13 +0100 Subject: [csw-maintainers] cswclassutils: location of template init scripts In-Reply-To: References: Message-ID: No dia 1 de Junho de 2010 20:14, Philip Brown escreveu: > On Tue, Jun 1, 2010 at 11:16 AM, Maciej (Matchek) Blizinski > wrote: >> >> In any case, I've submitted the cups package with the init script in >> /etc/opt/csw/init.d, until we figure how do to it better. >> >> I've removed the initfile-wrong-location check from checkpkg. ?I'm >> thinking of putting in a check that throws a message for cswinitsmf >> class files under /opt - we now know that it guarantees a broken >> package. ?What do you think? > > > I think we should use "cups needs an update" as motivation for "lets > figure out how to do it best,then fix cups to use it". OK, in this case I'll let you figure that out. I'll update the packages once it's implemented and documented. From phil at bolthole.com Wed Jun 2 15:25:44 2010 From: phil at bolthole.com (Philip Brown) Date: Wed, 2 Jun 2010 06:25:44 -0700 Subject: [csw-maintainers] testing9s problems Message-ID: after spending a bunch of time syncing csw packages, I discover that the OS is out of sync :( so please don't modify or use testing9s until further notice From phil at bolthole.com Wed Jun 2 18:19:08 2010 From: phil at bolthole.com (Philip Brown) Date: Wed, 2 Jun 2010 09:19:08 -0700 Subject: [csw-maintainers] cswclassutils: location of template init scripts In-Reply-To: References: Message-ID: I have an idea for both "making it work right", and also preserving backward compatibility. How about this for an idea: cswinitsmf will be modified, so that it creates a new file somewhere under /opt/csw, to keep track of file(s) registered for a package. /opt/csw/etc/cswinitsmf/CSWxxx (a directory It will also copy or symlink relevant files into there. For global zone, it will otherwise act as normal. Zone behaviour If the init files were in /etc/opt/csw, then copy into above mentioned directory, and proceed as normal. If the init files were in /opt/csw/etc/init.d, then it will get called, but without a normal list of files on stdin. We can then detect "hey I was called with no files",and instead have it go look in /opt/csw/etc/cswinitsmf/$PKGINST, and process THOSE files, as its file list for the zone. Whaddyathink? From phil at bolthole.com Wed Jun 2 18:22:25 2010 From: phil at bolthole.com (Philip Brown) Date: Wed, 2 Jun 2010 09:22:25 -0700 Subject: [csw-maintainers] cswclassutils: location of template init scripts In-Reply-To: References: Message-ID: Urk. posted too quickly. here's rev 2, slightly cleaned up. cswinitsmf will be modified, so that it keeps track of file(s) registered for a package, under /opt/csw, in /opt/csw/etc/cswinitsmf/CSWxxx (a directory) It will copy or symlink relevant files into there. For global zone, it will otherwise act as normal. Zone behaviour If the init files were in /etc/opt/csw, then copy into above mentioned directory(if writable. or otherwise, verify that they are already there), and proceed as normal. If the init files were in /opt/csw/etc/init.d, then it will get called, but without a normal list of files on stdin. We can then detect "hey I was called with no files",and instead have it go look in /opt/csw/etc/cswinitsmf/$PKGINST, and process those files, as its file list for the zone. From phil at bolthole.com Thu Jun 3 23:58:09 2010 From: phil at bolthole.com (Philip Brown) Date: Thu, 3 Jun 2010 14:58:09 -0700 Subject: [csw-maintainers] banner news, for gtk, x11 Message-ID: Good news, folks: We have a proven working new gtk+firefox chain! sitting in experimental/x11-reloaded The bad news is: I took some nasty hacks to get to this point, (along with a lot of help from other folks such as Dago, Maciej, and Ben).. and we still arent up to firefox 3.6.. just our same old firefox 3.0 :( but this is still really important news... because it shows that we can proceed with our "reloaded" plan, and move forward with an updated GNOME/gtk chain, based on SUN X11 libs. Which in theory means, we could have a new GNOME environment after this, too! Anyways, I'm off to a meeting now :) From jeff at cjsa.com Fri Jun 4 05:12:14 2010 From: jeff at cjsa.com (Jeffery Small) Date: Fri, 4 Jun 2010 03:12:14 GMT Subject: [csw-maintainers] banner news, for gtk, x11 References: Message-ID: Philip Brown writes: >Which in theory means, we could have a new GNOME environment after this, too! Could you work on updating gimp before tackling the much more involved Gnome project? Regards, -- Jeff C. Jeffery Small CJSA LLC 206-232-3338 jeff at cjsa.com 7000 E Mercer Way, Mercer Island, WA 98040 From phil at bolthole.com Fri Jun 4 05:51:30 2010 From: phil at bolthole.com (Philip Brown) Date: Thu, 3 Jun 2010 20:51:30 -0700 Subject: [csw-maintainers] banner news, for gtk, x11 In-Reply-To: References: Message-ID: On Thursday, June 3, 2010, Jeffery Small wrote: > Philip Brown writes: > >>Which in theory means, we could have a new GNOME environment after this, too! > > Could you work on updating gimp before tackling the much more involved Gnome > project? > gimp is just one package... you could always tackle it yourself ;-) From pfelecan at opencsw.org Sat Jun 5 14:28:29 2010 From: pfelecan at opencsw.org (Peter FELECAN) Date: Sat, 05 Jun 2010 14:28:29 +0200 Subject: [csw-maintainers] old gcc4* packages in catalog In-Reply-To: <1275352845-sup-6422@pinkfloyd.chass.utoronto.ca> (Ben Walton's message of "Mon, 31 May 2010 20:43:51 -0400") References: <1275352845-sup-6422@pinkfloyd.chass.utoronto.ca> Message-ID: Ben Walton writes: > Hi Phil, > > I just noticed that we seem to have overlapping gcc4 packages in the > catalog. This is likely due to a name change that slipped through > unnoticed, but maybe there is another reason. Either way, two > packages are shipping overlapping files: > > http://www.opencsw.org/packages/gcc4g95 > http://www.opencsw.org/packages/gcc4gfortran gcc4g95 is for the first version of the 4th branch for which I was the maintainer until 4.0.2. Afterward, Mike took up the maintenance of this branch and provided the gcc4gfortran which is the newer FORTRAN front end supported by then GCC Suite. -- Peter From pfelecan at opencsw.org Sat Jun 5 14:30:33 2010 From: pfelecan at opencsw.org (Peter FELECAN) Date: Sat, 05 Jun 2010 14:30:33 +0200 Subject: [csw-maintainers] gtk - dump 64bit for now? In-Reply-To: <20100601015127.GA76098@bolthole.com> (Philip Brown's message of "Mon, 31 May 2010 18:51:27 -0700") References: <20100601015127.GA76098@bolthole.com> Message-ID: Philip Brown writes: > So.... I'm looking into a "quickie" redo of libgtk , to get us into the > 21st century. > I see the older package has 64bit libs. > > I also see that unlike many other 64bit capable things, it's going to be a > pain for me to do a nice small clean implementation of that. > > BUT... in checking out our installed /opt/csw/lib/sparcv9 on our build > machines... it looks like there's no other libs that actually USE the 64bit > libgtk. > So I'm thinking of dropping it to keep life simple. > > In the future (possibly in the near future) it is entirely possible for > someone to take over libgtk and provide 64bit libs again if desired. > (heck, i would greatly prefer that :) > But for the near term, I think it's more important for us to get a modern > libgtk, QUICKLY, rather than wait another few weeks for someone else. > > Opinions? Can I get a yea or a nay from anyone else out there on this? As nobody answered and today I feel generous, here is a "yea", as a reward for stepping in the new century! -- Peter From pfelecan at opencsw.org Sat Jun 5 14:33:59 2010 From: pfelecan at opencsw.org (Peter FELECAN) Date: Sat, 05 Jun 2010 14:33:59 +0200 Subject: [csw-maintainers] banner news, for gtk, x11 In-Reply-To: (Philip Brown's message of "Thu, 3 Jun 2010 14:58:09 -0700") References: Message-ID: Philip Brown writes: > Good news, folks: We have a proven working new gtk+firefox chain! > > sitting in experimental/x11-reloaded > > The bad news is: I took some nasty hacks to get to this point, (along > with a lot of help from other folks such as Dago, Maciej, and Ben).. > and we still arent up to firefox 3.6.. just our same old firefox 3.0 > :( > but this is still really important news... because it shows that we > can proceed with our "reloaded" plan, and move forward with an updated > GNOME/gtk chain, based on SUN X11 libs. > > Which in theory means, we could have a new GNOME environment after this, too! > > > Anyways, I'm off to a meeting now :) After your meeting, can you start hacking KDE? It's asking for entry in the new century also. -- Peter From phil at bolthole.com Sun Jun 6 00:52:54 2010 From: phil at bolthole.com (Philip Brown) Date: Sat, 5 Jun 2010 15:52:54 -0700 Subject: [csw-maintainers] banner news, for gtk, x11 In-Reply-To: References: Message-ID: > After your meeting, can you start hacking KDE? It's asking for entry in > the new century also. > -- very true. but sadly I'm going to burn up most if not all of my csw free time getting gtk out the door. it looks as though I'll have to do 64bit builds after all. sigh. oh well. on the brighter side, you may end up being interested in the simplified make system I'm working on. it's single-makefile based. the more interesting one is in svn-- pkg/TEMPLATES/createpkg/Makefile.lib (which is not a makefile "library". but rather "a makefile for building libraries) not finished yet, but you should be able to tell already whether you like the style or not From phil at bolthole.com Sun Jun 6 19:30:31 2010 From: phil at bolthole.com (Philip Brown) Date: Sun, 6 Jun 2010 10:30:31 -0700 Subject: [csw-maintainers] old gcc4* packages in catalog In-Reply-To: References: <1275352845-sup-6422@pinkfloyd.chass.utoronto.ca> Message-ID: > > gcc4g95 is for the first version of the 4th branch for which I was the > maintainer until 4.0.2. Afterward, Mike took up the maintenance of this branch and > provided the gcc4gfortran which is the newer FORTRAN front end supported > by then GCC Suite. > -- > Peter thanks . so we just need a definitie statement, by those who acually know and care abot fortran, wheher there s any value i n the older branch or whether we can officially replace it with the newer? From pfelecan at opencsw.org Sun Jun 6 20:01:22 2010 From: pfelecan at opencsw.org (Peter FELECAN) Date: Sun, 06 Jun 2010 20:01:22 +0200 Subject: [csw-maintainers] old gcc4* packages in catalog In-Reply-To: (Philip Brown's message of "Sun, 6 Jun 2010 10:30:31 -0700") References: <1275352845-sup-6422@pinkfloyd.chass.utoronto.ca> Message-ID: Philip Brown writes: >> >> gcc4g95 is for the first version of the 4th branch for which I was the >> maintainer until 4.0.2. Afterward, Mike took up the maintenance of this branch and >> provided the gcc4gfortran which is the newer FORTRAN front end supported >> by then GCC Suite. >> -- >> Peter > > thanks . so we just need a definitie statement, by those who acually > know and care abot fortran, wheher there s any value i n the older > branch or whether we can officially replace it with the newer? I'm not a FORTRAN expert, the last time that I wrote code in that language was in the very beginning of the 9th decade of the 20th century when the FORTRAN 77 was the standard. However, I'm for keeping only the newer version of the corresponding front-end, i.e., gcc4gfortran int the 4th branch; for those needing an older version they can use the gcc3g77 or even gcc2g77. -- Peter From phil at bolthole.com Sun Jun 6 22:48:47 2010 From: phil at bolthole.com (Philip Brown) Date: Sun, 6 Jun 2010 13:48:47 -0700 Subject: [csw-maintainers] old gcc4* packages in catalog In-Reply-To: References: <1275352845-sup-6422@pinkfloyd.chass.utoronto.ca> Message-ID: okay. well someone please submit an empty packed of the new one that depends on the old one On Sunday, June 6, 2010, Peter FELECAN wrote: > Philip Brown writes: > >>> >>> gcc4g95 is for the first version of the 4th ?branch for which I was the >>> maintainer until 4.0.2. Afterward, Mike took up the maintenance of this branch and >>> provided the gcc4gfortran which is the newer FORTRAN front end supported >>> by then GCC Suite. >>> -- >>> Peter >> >> thanks . so we just need a definitie statement, by those who acually >> know and care abot fortran, wheher there s any value i n the older >> branch or whether we can officially replace it with the newer? > > I'm not a FORTRAN expert, the last time that I wrote code in that > language was in the very beginning of the 9th decade of the 20th century > when the FORTRAN 77 was the standard. However, I'm for keeping only the > newer version of the corresponding front-end, i.e., gcc4gfortran int the > 4th branch; for those needing an older version they can use the gcc3g77 > or even gcc2g77. > > -- > Peter > _______________________________________________ > maintainers mailing list > maintainers at lists.opencsw.org > https://lists.opencsw.org/mailman/listinfo/maintainers > .:: This mailing list's archive is public. ::. > From jeff at cjsa.com Mon Jun 7 03:40:03 2010 From: jeff at cjsa.com (Jeffery Small) Date: Mon, 7 Jun 2010 01:40:03 GMT Subject: [csw-maintainers] Question regarding di(1) utility Message-ID: I have been using an older self-compiled version of the di utility, but just downloaded the current 4.24,REV=2010.06.02 version. I am finding one odd behavior with the new version that i wanted to fly past others using this tool. The manual states: "Unless the -a flag is specified, the following mounted filesystems will not normally be displayed: filesystems with total blocks <= 0; filesystems marked by the operating system as "ignore"; automounted filesystems that are duplicates of other normally mounted filesystems; loopback filesystems that are part of a zone (Solaris)." I'm on a SPARC Solaris 10 system with no zones other than the default one. When I use my self-compiled version 3.10 utility I observe the above behavior. However, when using the new csw di, I am seeing some loopback mounts listed. For example, with the older di: % di -H -ss -f "S Mbv 1 i P" Filesystem Mount Size Avail %used Inodes %used /dev/md/dsk/d0 / 59.4G 14.3G 75% 7501312 6% /dev/md/dsk/d3 /var 3.9G 3.7G 5% 500864 1% /dev/md/dsk/d5 /v 67.3G 26.9G 59% 8497216 0% /dev/md/dsk/d7 /x 67.3G 15.9G 75% 8497216 0% swap /etc/svc/volatile 5.3G 5.3G 0% 582790 0% swap /tmp 5.7G 5.3G 6% 582790 0% swap /var/run 5.3G 5.3G 0% 582790 0% But with the CSW di, in addition to the above I also see displayed: % /opt/csw/bin/di -H -ss -f "S Mbv 1 i P" /platform/sun4u-us3/lib/libc_psr/libc_psr_hwcap1.so.1 /platform/sun4u-us3/lib/libc_psr.so.1 59.4G 14.3G 75% 7501312 6% /platform/sun4u-us3/lib/sparcv9/libc_psr/libc_psr_hwcap1.so.1 /platform/sun4u-us3/lib/sparcv9/libc_psr.so.1 59.4G 14.3G 75% 7501312 6% /sadm /var/sadm 59.4G 14.3G 75% 7501312 6% /v/JS/images /u/jeff/lib/images 67.3G 26.9G 59% 8497216 0% /x/JS/music /u/jeff/lib/music 67.3G 15.9G 75% 8497216 0% The first two additional entries are standard Solaris loopbacks, and the last three are personal loopback mounts for relocated data hierarchies. These loopback mounts are seen with the older di if the -a flag is specified. Can people check this behavior on other systems please. Based upon the manual, I believe that the current operation is in error. Regards, -- Jeff C. Jeffery Small CJSA LLC 206-232-3338 jeff at cjsa.com 7000 E Mercer Way, Mercer Island, WA 98040 From phil at bolthole.com Mon Jun 7 04:11:09 2010 From: phil at bolthole.com (Philip Brown) Date: Sun, 6 Jun 2010 19:11:09 -0700 Subject: [csw-maintainers] Question regarding di(1) utility In-Reply-To: References: Message-ID: I would guess it's because yours was compiled on sol10 and ours was sol9 From dam at opencsw.org Mon Jun 7 10:26:26 2010 From: dam at opencsw.org (Dagobert Michelsen) Date: Mon, 7 Jun 2010 10:26:26 +0200 Subject: [csw-maintainers] Question regarding di(1) utility In-Reply-To: References: Message-ID: <1FF7E50C-2049-4377-8D87-3F1F3B5E4684@opencsw.org> Hi Jeff, I am forwarding this to the author, Brad Lanam. I am pretty sure he can answer that in a jiffy :-) Brad: Please reply-all as you can't write to maintainers at . Best regards -- Dago > Am 07.06.2010 um 04:11 schrieb Philip Brown: >> Am 07.06.2010 um 03:40 schrieb Jeffery Small: >> I have been using an older self-compiled version of the di utility, >> but >> just downloaded the current 4.24,REV=2010.06.02 version. I am >> finding one >> odd behavior with the new version that i wanted to fly past others >> using >> this tool. The manual states: >> >> "Unless the -a flag is specified, the following mounted >> filesystems >> will not normally be displayed: filesystems with total blocks <= >> 0; >> filesystems marked by the operating system as "ignore"; >> automounted >> filesystems that are duplicates of other normally mounted >> filesystems; >> loopback filesystems that are part of a zone (Solaris)." >> >> I'm on a SPARC Solaris 10 system with no zones other than the default >> one. When I use my self-compiled version 3.10 utility I observe >> the above >> behavior. However, when using the new csw di, I am seeing some >> loopback >> mounts listed. For example, with the older di: >> >> % di -H -ss -f "S Mbv 1 i P" >> Filesystem Mount Size Avail %used >> Inodes %used >> /dev/md/dsk/d0 / 59.4G 14.3G 75% >> 7501312 6% >> /dev/md/dsk/d3 /var 3.9G 3.7G 5% >> 500864 1% >> /dev/md/dsk/d5 /v 67.3G 26.9G 59% >> 8497216 0% >> /dev/md/dsk/d7 /x 67.3G 15.9G 75% >> 8497216 0% >> swap /etc/svc/volatile 5.3G 5.3G 0% >> 582790 0% >> swap /tmp 5.7G 5.3G 6% >> 582790 0% >> swap /var/run 5.3G 5.3G 0% >> 582790 0% >> >> But with the CSW di, in addition to the above I also see displayed: >> >> % /opt/csw/bin/di -H -ss -f "S Mbv 1 i P" >> /platform/sun4u-us3/lib/libc_psr/libc_psr_hwcap1.so.1 >> /platform/sun4u-us3/lib/libc_psr.so.1 >> 59.4G 14.3G 75% 7501312 6% >> /platform/sun4u-us3/lib/sparcv9/libc_psr/libc_psr_hwcap1.so.1 >> /platform/sun4u-us3/lib/sparcv9/libc_psr.so.1 >> 59.4G 14.3G 75% 7501312 6% >> /sadm >> /var/sadm >> 59.4G 14.3G 75% 7501312 6% >> /v/JS/images >> /u/jeff/lib/images >> 67.3G 26.9G 59% 8497216 0% >> /x/JS/music >> /u/jeff/lib/music >> 67.3G 15.9G 75% 8497216 0% >> >> The first two additional entries are standard Solaris loopbacks, >> and the >> last three are personal loopback mounts for relocated data >> hierarchies. >> These loopback mounts are seen with the older di if the -a flag is >> specified. >> >> Can people check this behavior on other systems please. Based upon >> the >> manual, I believe that the current operation is in error. > > I would guess it's because yours was compiled on sol10 and ours was > sol9 From bonivart at opencsw.org Mon Jun 7 10:33:24 2010 From: bonivart at opencsw.org (Peter Bonivart) Date: Mon, 7 Jun 2010 10:33:24 +0200 Subject: [csw-maintainers] Why are packages named in a "creative", non-consistent way? Message-ID: I'm talking about the sparconly tag on these two recently updated packages: libxft2:2.1.14,REV=2010.06.03.sparconly - A client-side font API for X applications xpm:3.5.8,REV=2010.05.30_sparconly - X11 pixmap library libxft2 used to be named like this: libxft2-2.1.14,REV=2010.05.30-SunOS5.9-sparc-CSW.pkg.gz libxft2-2.1.14,REV=2010.05.30-SunOS5.9-i386-CSW.pkg.gz Why not just update the sparc package with a current date and not mess with the rest? Does this have to do with the "release process" in any way? Why is one separated with a dot and the other with an underscore? If this is needed at all why not put the tag with the software version, e.g. 2.1.14sparconly,REV=2010.06.03? -- /peter From dam at opencsw.org Mon Jun 7 11:50:35 2010 From: dam at opencsw.org (Dagobert Michelsen) Date: Mon, 7 Jun 2010 11:50:35 +0200 Subject: [csw-maintainers] banner news, for gtk, x11 In-Reply-To: References: Message-ID: <999BE378-4B7C-49A3-99CE-9E7DF1D4BFD8@opencsw.org> Hi Phil, Am 06.06.2010 um 00:52 schrieb Philip Brown: >> After your meeting, can you start hacking KDE? It's asking for >> entry in >> the new century also. >> -- > > very true. but sadly I'm going to burn up most if not all of my csw > free time getting gtk out the door. it looks as though I'll have to do > 64bit builds after all. sigh. Probably not. Now as the compile order is understood I would rather prefer doing it the standard way and updating the GAR recipes. Best regards -- Dago From phil at bolthole.com Mon Jun 7 15:08:34 2010 From: phil at bolthole.com (Philip Brown) Date: Mon, 7 Jun 2010 06:08:34 -0700 Subject: [csw-maintainers] Why are packages named in a "creative", non-consistent way? In-Reply-To: References: Message-ID: On Monday, June 7, 2010, Peter Bonivart wrote: > I'm talking about the sparconly tag on these... that's something we've actually done for a very long time, when only one arch of a package got an update for purely packaging reasons. as far as . vs _ in that trailing position ... didn't think anyone cared much From phil at bolthole.com Mon Jun 7 15:12:42 2010 From: phil at bolthole.com (Philip Brown) Date: Mon, 7 Jun 2010 06:12:42 -0700 Subject: [csw-maintainers] banner news, for gtk, x11 In-Reply-To: <999BE378-4B7C-49A3-99CE-9E7DF1D4BFD8@opencsw.org> References: <999BE378-4B7C-49A3-99CE-9E7DF1D4BFD8@opencsw.org> Message-ID: On Monday, June 7, 2010, Dagobert Michelsen wrote: > Hi Phil, > > Am 06.06.2010 um 00:52 schrieb Philip Brown: > > After your meeting, can you start hacking KDE? It's asking for entry in > the new century also. > -- > > > very true. but sadly I'm going to burn up most if not all of my csw > free time getting gtk out the door. it looks as though I'll have to do > 64bit builds after all. sigh. > > > Probably not. Now as the compile order is understood I would rather > prefer doing it the standard way and updating the GAR recipes. > well wonderfull ! are you (or someone else) going to reroll them? today or tomorrow? as I've already said, I'd prefer someone else do them. From dam at opencsw.org Mon Jun 7 15:26:47 2010 From: dam at opencsw.org (Dagobert Michelsen) Date: Mon, 7 Jun 2010 15:26:47 +0200 Subject: [csw-maintainers] banner news, for gtk, x11 In-Reply-To: References: <999BE378-4B7C-49A3-99CE-9E7DF1D4BFD8@opencsw.org> Message-ID: Hi Phil, Am 07.06.2010 um 15:12 schrieb Philip Brown: > On Monday, June 7, 2010, Dagobert Michelsen wrote: >> Am 06.06.2010 um 00:52 schrieb Philip Brown: >> >> After your meeting, can you start hacking KDE? It's asking for >> entry in >> the new century also. >> -- >> >> >> very true. but sadly I'm going to burn up most if not all of my csw >> free time getting gtk out the door. it looks as though I'll have to >> do >> 64bit builds after all. sigh. >> >> >> Probably not. Now as the compile order is understood I would rather >> prefer doing it the standard way and updating the GAR recipes. >> > > > well wonderfull ! are you (or someone else) going to reroll them? > today or tomorrow? Probably. What package do you want to have rerolled? My time is limited, but I could trigger a compile once in a while. Best regards -- Dago From bonivart at opencsw.org Mon Jun 7 15:50:26 2010 From: bonivart at opencsw.org (Peter Bonivart) Date: Mon, 7 Jun 2010 15:50:26 +0200 Subject: [csw-maintainers] Why are packages named in a "creative", non-consistent way? In-Reply-To: References: Message-ID: On Mon, Jun 7, 2010 at 3:08 PM, Philip Brown wrote: > On Monday, June 7, 2010, Peter Bonivart wrote: >> I'm talking about the sparconly tag on these... > > > that's something we've actually done for a very long time, when only > one arch of a package got an update for purely packaging reasons. > > as far as . vs _ in that trailing position ? ... didn't think anyone cared much Neither of those arguments are very good. -- /peter From phil at bolthole.com Mon Jun 7 15:51:30 2010 From: phil at bolthole.com (Philip Brown) Date: Mon, 7 Jun 2010 06:51:30 -0700 Subject: [csw-maintainers] banner news, for gtk, x11 In-Reply-To: References: <999BE378-4B7C-49A3-99CE-9E7DF1D4BFD8@opencsw.org> Message-ID: libcairo would be nice for a start. and then followed by pango if you have time. as I mentioned we need full 64bit From phil at bolthole.com Mon Jun 7 17:58:15 2010 From: phil at bolthole.com (Philip Brown) Date: Mon, 7 Jun 2010 08:58:15 -0700 Subject: [csw-maintainers] banner news, for gtk, x11 In-Reply-To: References: <999BE378-4B7C-49A3-99CE-9E7DF1D4BFD8@opencsw.org> Message-ID: oh I think libxft may also need redo, so that libcairo amd64 gets it On Monday, June 7, 2010, Philip Brown wrote: > libcairo would be nice for a start. and then followed by pango if you > have time. as I mentioned we need full 64bit > From phil at bolthole.com Mon Jun 7 18:29:44 2010 From: phil at bolthole.com (Philip Brown) Date: Mon, 7 Jun 2010 09:29:44 -0700 Subject: [csw-maintainers] banner news, for gtk, x11 In-Reply-To: References: <999BE378-4B7C-49A3-99CE-9E7DF1D4BFD8@opencsw.org> Message-ID: On Mon, Jun 7, 2010 at 8:58 AM, Philip Brown wrote: > oh I think libxft may also need redo, so that libcairo amd64 gets it > oops.. I see you just redid libcairo for x11-reloaded friday, which was really nice... and it appears that I misspoke: libcairo does not use libxft2, but libpango does. So if you could redo libxft2 for full 64bit, and then hit Pango, that would be really helpful :-) From pfelecan at opencsw.org Mon Jun 7 19:25:36 2010 From: pfelecan at opencsw.org (Peter FELECAN) Date: Mon, 07 Jun 2010 19:25:36 +0200 Subject: [csw-maintainers] old gcc4* packages in catalog In-Reply-To: (Philip Brown's message of "Sun, 6 Jun 2010 13:48:47 -0700") References: <1275352845-sup-6422@pinkfloyd.chass.utoronto.ca> Message-ID: Philip Brown writes: > On Sunday, June 6, 2010, Peter FELECAN wrote: >> Philip Brown writes: >> >>>> >>>> gcc4g95 is for the first version of the 4th ?branch for which I was the >>>> maintainer until 4.0.2. Afterward, Mike took up the maintenance of this branch and >>>> provided the gcc4gfortran which is the newer FORTRAN front end supported >>>> by then GCC Suite. >>>> -- >>>> Peter >>> >>> thanks . so we just need a definitie statement, by those who acually >>> know and care abot fortran, wheher there s any value i n the older >>> branch or whether we can officially replace it with the newer? >> >> I'm not a FORTRAN expert, the last time that I wrote code in that >> language was in the very beginning of the 9th decade of the 20th century >> when the FORTRAN 77 was the standard. However, I'm for keeping only the >> newer version of the corresponding front-end, i.e., gcc4gfortran int the >> 4th branch; for those needing an older version they can use the gcc3g77 >> or even gcc2g77. >> > okay. well someone please submit an empty packed of the new one that > depends on the old one In case that you agreed with my proposal, it's the other way: the old one is empty and depends on the new one. What you say is to keep gcc4g95 (the old one) but we should keep gcc4gfortran (the new one) -- Peter From phil at bolthole.com Mon Jun 7 19:53:24 2010 From: phil at bolthole.com (Philip Brown) Date: Mon, 7 Jun 2010 10:53:24 -0700 Subject: [csw-maintainers] old gcc4* packages in catalog In-Reply-To: References: <1275352845-sup-6422@pinkfloyd.chass.utoronto.ca> Message-ID: On Mon, Jun 7, 2010 at 10:25 AM, Peter FELECAN wrote: > Philip Brown writes: > >> okay. well someone please submit an empty packed of the new one that >> depends on the old one > oops. I mistyped. (durn iphone! yeah, that's it...) I meant submit old, empty package that depends on new one. make gcc4g95 pull in gcc4gfortran. From phil at opencsw.org Mon Jun 7 20:03:52 2010 From: phil at opencsw.org (Philip Brown) Date: Mon, 7 Jun 2010 11:03:52 -0700 Subject: [csw-maintainers] problems with upstream notification Message-ID: So... i got this email message, and there are two problems with it. 1. Subject line claims "Glib" but it is not. The filename would appear to be associated with pm_gtk2. 2. it should not be telling me about it. Theres a side issue, though, that I dont even see any USTREAM_WATCH, or whatever, magic in cpan/Gtk2/trunk/Makefile, so it seems even more confused than the above two items indicate. I'd offer more info, but that's about as far as I understand this stuff. What's going on....? ---------- Forwarded message ---------- From: Upstream Package Watch Date: Sun, May 30, 2010 at 5:55 PM Subject: [svn] Glib upstream update notification To: phil at opencsw.org Hello dear Glib maintainer, The upstream notification job has detected the availability of new files for Glib. The following upstream file(s): Glib-1.223.tar.gz is/are available at the following url(s): http://search.cpan.org/CPAN/authors/id/T/TS/TSCH/ ftp://ftp.nrc.ca/pub/CPAN/authors/id/T/TS/TSCH/ ftp://ftp.nas.nasa.gov/pub/perl/CPAN/authors/id/T/TS/TSCH/ http://mirrors.ibiblio.org/pub/mirrors/CPAN/authors/id/T/TS/TSCH/ ftp://cpan.pair.com/pub/CPAN/authors/id/T/TS/TSCH/ http://mirrors.kernel.org/cpan/authors/id/T/TS/TSCH/ Please consider updating your package. Questions, problem report or help should be sent to mailto:maintainers at lists.opencsw.org -- Kindest regards upstream notification job From pfelecan at opencsw.org Mon Jun 7 20:52:43 2010 From: pfelecan at opencsw.org (Peter FELECAN) Date: Mon, 07 Jun 2010 20:52:43 +0200 Subject: [csw-maintainers] old gcc4* packages in catalog In-Reply-To: (Philip Brown's message of "Mon, 7 Jun 2010 10:53:24 -0700") References: <1275352845-sup-6422@pinkfloyd.chass.utoronto.ca> Message-ID: Philip Brown writes: > On Mon, Jun 7, 2010 at 10:25 AM, Peter FELECAN wrote: >> Philip Brown writes: >> >>> okay. well someone please submit an empty packed of the new one that >>> depends on the old one >> > oops. I mistyped. (durn iphone! yeah, that's it...) Are you following the WWDC announcement? > I meant submit old, empty package that depends on new one. > make gcc4g95 pull in gcc4gfortran. Yup. -- Peter From bwalton at opencsw.org Mon Jun 7 20:58:56 2010 From: bwalton at opencsw.org (Ben Walton) Date: Mon, 07 Jun 2010 14:58:56 -0400 Subject: [csw-maintainers] old gcc4* packages in catalog In-Reply-To: References: <1275352845-sup-6422@pinkfloyd.chass.utoronto.ca> Message-ID: <1275937103-sup-8772@pinkfloyd.chass.utoronto.ca> Excerpts from Peter FELECAN's message of Mon Jun 07 14:52:43 -0400 2010: > Are you following the WWDC announcement? With all those extra pixels, there'll be no more excuses for mis-reading things! :) -Ben -- Ben Walton Systems Programmer - CHASS University of Toronto C:416.407.5610 | W:416.978.4302 From william at wbonnet.net Tue Jun 8 00:31:07 2010 From: william at wbonnet.net (William Bonnet) Date: Tue, 08 Jun 2010 00:31:07 +0200 Subject: [csw-maintainers] problems with upstream notification In-Reply-To: References: Message-ID: <4C0D732B.6010401@wbonnet.net> Hi Phil > I'd offer more info, but that's about as far as I understand this stuff. > > What's going on....? > It's still the same problem, both glib 1 and 2 have the same GARNAME... :( Thus it is difficult with the current version of upstream watch to make distinction between both version. cheers W. /me popping late emails... > > ---------- Forwarded message ---------- > From: Upstream Package Watch > Date: Sun, May 30, 2010 at 5:55 PM > Subject: [svn] Glib upstream update notification > To: phil at opencsw.org > > > > Hello dear Glib maintainer, > > The upstream notification job has detected the availability of new > files for Glib. > > The following upstream file(s): > Glib-1.223.tar.gz > > is/are available at the following url(s): > http://search.cpan.org/CPAN/authors/id/T/TS/TSCH/ > ftp://ftp.nrc.ca/pub/CPAN/authors/id/T/TS/TSCH/ > ftp://ftp.nas.nasa.gov/pub/perl/CPAN/authors/id/T/TS/TSCH/ > http://mirrors.ibiblio.org/pub/mirrors/CPAN/authors/id/T/TS/TSCH/ > ftp://cpan.pair.com/pub/CPAN/authors/id/T/TS/TSCH/ > http://mirrors.kernel.org/cpan/authors/id/T/TS/TSCH/ > > Please consider updating your package. > > Questions, problem report or help should be sent to > mailto:maintainers at lists.opencsw.org > > -- > Kindest regards > upstream notification job > _______________________________________________ > maintainers mailing list > maintainers at lists.opencsw.org > https://lists.opencsw.org/mailman/listinfo/maintainers > .:: This mailing list's archive is public. ::. > -- William http://www.wbonnet.net http://www.sunwizard.net Le site fran?ais des amateurs de stations Unix http://www.opencsw.org Community SoftWare for Solaris http://www.guses.org French speaking Solaris User Group From phil at bolthole.com Tue Jun 8 00:41:38 2010 From: phil at bolthole.com (Philip Brown) Date: Mon, 7 Jun 2010 15:41:38 -0700 Subject: [csw-maintainers] problems with upstream notification In-Reply-To: <4C0D732B.6010401@wbonnet.net> References: <4C0D732B.6010401@wbonnet.net> Message-ID: On Mon, Jun 7, 2010 at 3:31 PM, William Bonnet wrote: > Hi Phil > >> I'd offer more info, but that's about as far as I understand this stuff. >> >> What's going on....? >> > > It's still the same problem, both glib 1 and 2 have the same GARNAME... :( But there's more than that... since neither "glib1" nor "glib 2", reference the file "Glib-1.223.tar.gz: > Thus it is difficult with the current version of upstream watch to make > distinction between both version. Could you put in a little more debug please, so that it explicitly mentions which file in gar it is parsing, to generate its output? From william at wbonnet.net Tue Jun 8 00:58:35 2010 From: william at wbonnet.net (William Bonnet) Date: Tue, 08 Jun 2010 00:58:35 +0200 Subject: [csw-maintainers] problems with upstream notification In-Reply-To: References: <4C0D732B.6010401@wbonnet.net> Message-ID: <4C0D799B.3010403@wbonnet.net> Hi > But there's more than that... since neither "glib1" nor "glib 2", > reference the file > "Glib-1.223.tar.gz: > > So there are three packages with the same garname :( > Could you put in a little more debug please, so that it explicitly > mentions which file in gar it is parsing, to generate its output? > sure cheers W. -- William http://www.wbonnet.net http://www.sunwizard.net Le site fran?ais des amateurs de stations Unix http://www.opencsw.org Community SoftWare for Solaris http://www.guses.org French speaking Solaris User Group From phil at bolthole.com Tue Jun 8 03:56:36 2010 From: phil at bolthole.com (Philip Brown) Date: Mon, 7 Jun 2010 18:56:36 -0700 Subject: [csw-maintainers] banner news, for gtk, x11 In-Reply-To: References: <999BE378-4B7C-49A3-99CE-9E7DF1D4BFD8@opencsw.org> Message-ID: On Mon, Jun 7, 2010 at 9:29 AM, Philip Brown wrote: > ..... > and it appears that I misspoke: libcairo does not use libxft2, but > libpango does. > > So if you could redo libxft2 for full 64bit, and then hit Pango, that > would be really helpful :-) Guess you were busy today. So I updated libxft2 myself. Since the prior, almost identical version was already public (it doesnt depend on cswx11), it will be hitting public mirrors in an hour or so. If you have the time, it would be nice if you could rebuild pango in the next 16 hours or so. Otherwise, I guess that will be my project for tomorrow. From dam at opencsw.org Tue Jun 8 09:35:25 2010 From: dam at opencsw.org (Dagobert Michelsen) Date: Tue, 8 Jun 2010 09:35:25 +0200 Subject: [csw-maintainers] problems with upstream notification In-Reply-To: <4C0D732B.6010401@wbonnet.net> References: <4C0D732B.6010401@wbonnet.net> Message-ID: <1E142D88-1A48-4BE1-A132-1AEB477E9C0C@opencsw.org> Hi William, Am 08.06.2010 um 00:31 schrieb William Bonnet: >> I'd offer more info, but that's about as far as I understand this >> stuff. >> >> What's going on....? >> > > It's still the same problem, both glib 1 and 2 have the same > GARNAME... :( > > Thus it is difficult with the current version of upstream watch to > make distinction between both version. BTW, why are you using the GARNAME and not CATALOGNAME of the first produced package? Best regards -- Dago From dam at opencsw.org Tue Jun 8 09:40:04 2010 From: dam at opencsw.org (Dagobert Michelsen) Date: Tue, 8 Jun 2010 09:40:04 +0200 Subject: [csw-maintainers] banner news, for gtk, x11 In-Reply-To: References: <999BE378-4B7C-49A3-99CE-9E7DF1D4BFD8@opencsw.org> Message-ID: <7F47F2E5-95EA-4284-AF93-1C2E12A9FE18@opencsw.org> Hi Phil, Am 08.06.2010 um 03:56 schrieb Philip Brown: > On Mon, Jun 7, 2010 at 9:29 AM, Philip Brown > wrote: >> ..... >> and it appears that I misspoke: libcairo does not use libxft2, but >> libpango does. >> >> So if you could redo libxft2 for full 64bit, and then hit Pango, that >> would be really helpful :-) > > Guess you were busy today. So I updated libxft2 myself. > Since the prior, almost identical version was already public (it > doesnt depend on cswx11), it will be hitting public mirrors in an hour > or so. > > If you have the time, it would be nice if you could rebuild pango in > the next 16 hours or so. Otherwise, I guess that will be my project > for tomorrow. If it is an easy build I can kick it on, but I won't have time to dig :-P Best regards -- DAgo From dam at opencsw.org Tue Jun 8 09:56:20 2010 From: dam at opencsw.org (Dagobert Michelsen) Date: Tue, 8 Jun 2010 09:56:20 +0200 Subject: [csw-maintainers] banner news, for gtk, x11 In-Reply-To: References: <999BE378-4B7C-49A3-99CE-9E7DF1D4BFD8@opencsw.org> Message-ID: <55B7D3D0-6128-4AA9-99CF-BC1B36EFC765@opencsw.org> Hi Phil, Am 07.06.2010 um 17:58 schrieb Philip Brown: > oh I think libxft may also need redo, so that libcairo amd64 gets it > > On Monday, June 7, 2010, Philip Brown wrote: >> libcairo would be nice for a start. and then followed by pango if you >> have time. as I mentioned we need full 64bit pango has compilation errors: "/opt/csw/include/glib-2.0/glib/gthread.h", line 348: Error: The size in an array declaration cannot be negative. "hb-font.cc", line 179: Warning (Anachronism): Assigning _hb_blob_t*(*) (unsigned,void*) to extern "C" _hb_blob_t*(*)(unsigned,void*). 1 Error(s) and 1 Warning(s) detected. Sorry, I can't dig right now... Best regards -- Dago From maciej at opencsw.org Tue Jun 8 12:46:04 2010 From: maciej at opencsw.org (Maciej (Matchek) Blizinski) Date: Tue, 8 Jun 2010 11:46:04 +0100 Subject: [csw-maintainers] [csw-pkgsubmissions] newpkgs buildbot, colormake, coreutils, cswut(...) In-Reply-To: References: <201006030143.o531hgwc009167@login.bo.opencsw.org> Message-ID: No dia 4 de Junho de 2010 01:05, Philip Brown escreveu: > Hurray for new collision detection... I didn't know it was already working! Is there any available code so that people can read it and helpfully point out any flaws? From phil at bolthole.com Tue Jun 8 17:57:00 2010 From: phil at bolthole.com (Philip Brown) Date: Tue, 8 Jun 2010 08:57:00 -0700 Subject: [csw-maintainers] banner news, for gtk, x11 In-Reply-To: <55B7D3D0-6128-4AA9-99CF-BC1B36EFC765@opencsw.org> References: <999BE378-4B7C-49A3-99CE-9E7DF1D4BFD8@opencsw.org> <55B7D3D0-6128-4AA9-99CF-BC1B36EFC765@opencsw.org> Message-ID: Okay. thanks for trying, and the quick status. I'll get it done today On Tue, Jun 8, 2010 at 12:56 AM, Dagobert Michelsen wrote: > Hi > pango has compilation errors: > > "/opt/csw/include/glib-2.0/glib/gthread.h", line 348: Error: The size in an > array declaration cannot be negative. > "hb-font.cc", line 179: Warning (Anachronism): Assigning > _hb_blob_t*(*)(unsigned,void*) to extern "C" _hb_blob_t*(*)(unsigned,void*). > 1 Error(s) and 1 Warning(s) detected. > > Sorry, I can't dig right now... > From bonivart at opencsw.org Tue Jun 8 18:22:17 2010 From: bonivart at opencsw.org (Peter Bonivart) Date: Tue, 8 Jun 2010 18:22:17 +0200 Subject: [csw-maintainers] [csw-pkgrequests] package request In-Reply-To: References: <20100607150628.A3F39C0B@mail.opencsw.org> <2C8C35B1-200C-4484-9183-68E5AA33E8BA@opencsw.org> Message-ID: On Tue, Jun 8, 2010 at 11:00 AM, Peter Bonivart wrote: > On Tue, Jun 8, 2010 at 8:59 AM, Dagobert Michelsen wrote: >> Interesting. Want to start a project and set up a webpage? Sounds like a fun >> thing to do whenever there are a couple of spare minutes. [Posting to maintainers list to get more help] I have set up a project page on the wiki: http://wiki.opencsw.org/project-koha Here's four modules I have built today: http://mirror.opencsw.org/experimental.html#koha Please visit the project page and help out if you have some spare cycles. Perl modules are usually real easy builds. -- /peter From phil at bolthole.com Tue Jun 8 18:53:57 2010 From: phil at bolthole.com (Philip Brown) Date: Tue, 8 Jun 2010 09:53:57 -0700 Subject: [csw-maintainers] package collision detection Message-ID: On Tue, Jun 8, 2010 at 3:46 AM, Maciej (Matchek) Blizinski wrote: > No dia 4 de Junho de 2010 01:05, Philip Brown escreveu: >> Hurray for new collision detection... > > I didn't know it was already working! Is there any available code so > that people can read it and helpfully point out any flaws? > there are flaws aplenty, unfortunately :-) At the moment, its functional by virtue of a side table. The table is.. incompletely populated. The reason being, it doesnt allow collisions... which means, it does not allow all packages to be registered, because there are collisions!! But it at least works, for packages that have successfully been entered into there. How things work now is; At package registration/"batch" time, the contents of the package get inserted into the table. if there are collisions, then it fails insert, and I throw the package back to you :) If you wished, I'd be happy to give you a mysqldump of the table or something. From phil at bolthole.com Tue Jun 8 20:48:27 2010 From: phil at bolthole.com (Philip Brown) Date: Tue, 8 Jun 2010 11:48:27 -0700 Subject: [csw-maintainers] package collision detection In-Reply-To: References: Message-ID: On Tue, Jun 8, 2010 at 9:53 AM, Philip Brown wrote: > On Tue, Jun 8, 2010 at 3:46 AM, Maciej (Matchek) Blizinski > wrote: >> No dia 4 de Junho de 2010 01:05, Philip Brown escreveu: >>> Hurray for new collision detection... >> >> I didn't know it was already working! Is there any available code so >> that people can read it and helpfully point out any flaws? >> ps: This runs over the pkgmap file. nawk ' $2 ~ /^[bceflpsvx]$/ { if(($2=="s")||($2=="l")){ $4=substr($4,1,index($4,"=")-1); } if(index($4,"/")==1) { longname=$4; } else{ longname="'$basedir'/"$4; } printf("insert into xxx values ('\'$pkgname\'', '\''%s'\'');\n", longname); }' From william at wbonnet.net Tue Jun 8 22:24:56 2010 From: william at wbonnet.net (William Bonnet) Date: Tue, 08 Jun 2010 22:24:56 +0200 Subject: [csw-maintainers] Next summer camp Message-ID: <4C0EA718.4060509@wbonnet.net> Hi all During the last Winter camp we talked about the location of the next Summer camp, and Paris was the consensus. It is now time to start to topics, organization and date of the summer camp. This camp will take place in the city of Puteaux, aka La D?fense, one of the business place of Paris. Puteaux is located west of Paris. About 2O to 25 minutes from the center of Paris (metro line 1). And about 1 hour from airports or train stations. The meeting will be held in the facilities of "Groupe ON-X". A local consulting company. We will have at our disposal a meeting room, internet acces, a place to eat almost without leaving the meeting room :) and several nearby hotel and restaurant. I'll do local organization, and provide you needed information. Once the date chose, i'll try to find nearby hotel and organize prebooking. I have created a wiki page : http://wiki.opencsw.org/summercamp-2010 Many things are still To Be Defined (TBD). Discussions are now open :) One of the first thing to decide is the date. Please use the following poll : http://www.doodle.com/vsaehg5rxyrcgxcq If have set more date than my initial choice. If needed i'll try to change my availability in july. cheers W. -- William http://www.wbonnet.net http://www.sunwizard.net Le site fran?ais des amateurs de stations Unix http://www.opencsw.org Community SoftWare for Solaris http://www.guses.org French speaking Solaris User Group From skayser at opencsw.org Tue Jun 8 22:55:54 2010 From: skayser at opencsw.org (Sebastian Kayser) Date: Tue, 08 Jun 2010 22:55:54 +0200 Subject: [csw-maintainers] Next summer camp In-Reply-To: <4C0EA718.4060509@wbonnet.net> References: <4C0EA718.4060509@wbonnet.net> Message-ID: <4C0EAE5A.9080005@opencsw.org> William Bonnet wrote on 08.06.2010 22:24: > During the last Winter camp we talked about the location of the next > Summer camp, and Paris was the consensus. It is now time to start to > topics, organization and date of the summer camp. > [...] > > One of the first thing to decide is the date. Please use the following poll : http://www.doodle.com/vsaehg5rxyrcgxcq Thanks William for taking care of the camp! I have just added the few spare weekends of mine, the rest is already booked - mostly with marriage parties ... or as a friend put it "watch out, this is spreading!" ;). Can we aim for a full Sunday this time and departure on Monday? Sebastian From william at wbonnet.net Tue Jun 8 23:09:29 2010 From: william at wbonnet.net (William Bonnet) Date: Tue, 08 Jun 2010 23:09:29 +0200 Subject: [csw-maintainers] Next summer camp In-Reply-To: <4C0EAE5A.9080005@opencsw.org> References: <4C0EA718.4060509@wbonnet.net> <4C0EAE5A.9080005@opencsw.org> Message-ID: <4C0EB189.3090201@wbonnet.net> Hi > Can we aim for a full Sunday this time and departure on Monday? > The meeting room will be available on Sunday, all day long (until late...). It is up to the participant. But i fully agree and support this proposal, it would be nice to have two full days of meetings. cheers W. -- William http://www.wbonnet.net http://www.sunwizard.net Le site fran?ais des amateurs de stations Unix http://www.opencsw.org Community SoftWare for Solaris http://www.guses.org French speaking Solaris User Group From bonivart at opencsw.org Tue Jun 8 23:44:32 2010 From: bonivart at opencsw.org (Peter Bonivart) Date: Tue, 8 Jun 2010 23:44:32 +0200 Subject: [csw-maintainers] Help with bug report? (perl issue) Message-ID: Hi, I got a bug report today: http://www.opencsw.org/mantis/view.php?id=4446 It seemed like a simple one with mismatch between old/new Perl vs. modules not compiled with it. Seen a few of those when we went from 5.8.8 to 5.10.1. However, the info he provides doesn't confirm that suspicion. Looks like he has a well updated system. Anyone know what to ask for so we can get to the bottom of this? -- /peter From maciej at opencsw.org Wed Jun 9 13:44:48 2010 From: maciej at opencsw.org (Maciej (Matchek) Blizinski) Date: Wed, 9 Jun 2010 12:44:48 +0100 Subject: [csw-maintainers] cswclassutils: location of template init scripts In-Reply-To: References: Message-ID: No dia 2 de Junho de 2010 17:22, Philip Brown escreveu: > Urk. posted too quickly. here's rev 2, slightly cleaned up. > > cswinitsmf will be modified, so that it keeps track of file(s) > registered for a package, under /opt/csw, > in > > /opt/csw/etc/cswinitsmf/CSWxxx (a directory) > > It will copy or symlink relevant files into there. > For global zone, it will otherwise act as normal. > > Zone behaviour > If the init files were in /etc/opt/csw, then copy into above mentioned > directory(if writable. or otherwise, verify that they are already > there), and proceed as normal. > > If the init files were in /opt/csw/etc/init.d, then it will get > called, but without a normal list of files on stdin. > We can then detect "hey I was called with no files",and instead have > it go look in > /opt/csw/etc/cswinitsmf/$PKGINST, and process those files, as its file > list for the zone. After reading this carefully, on one hand I understand the idea behind handing the special case with no files supplied; but I realized that I don't know enough about the intended way for the NFS sharing to work. I've read our user guide page[1], and the wiki page[2]. IIUC, the NFS sharing is done in such a way that there is a master machine onto which CSW packages are installed and which serves the /opt/csw directory in a read-only fashion. Other machines mount this directory via NFS and don't install the CSW packages at all. It's enough to add /opt/csw/bin to $PATH in order to access the utilities. If you tried to install a package to a machine with read-only mounted /opt/csw, I'm guessing that pkgadd would fail (is that true?), because writing to the file wouldn't be possible. If you don't run pkgadd on client machines, how are daemons started? The "empty input cswinitsmf" idea above wouldn't solve that. Sorry about backtracking, but I have to understand the background before I can make any useful comments on a particular idea. [1] http://www.opencsw.org/userguide/sharingcsw [2] http://wiki.opencsw.org/shared-opt-csw-setup From phil at bolthole.com Wed Jun 9 20:16:26 2010 From: phil at bolthole.com (Philip Brown) Date: Wed, 9 Jun 2010 11:16:26 -0700 Subject: [csw-maintainers] cswclassutils: location of template init scripts In-Reply-To: References: Message-ID: thank you for taking the time to understand the issue fully. more details below On Wednesday, June 9, 2010, Maciej (Matchek) Blizinski wrote: > > > .....?IIUC, the NFS > sharing is done in such a way that there is a master machine onto > which CSW packages are installed and which serves the /opt/csw > directory in a read-only fashion. ?Other machines mount this directory > via NFS and don't install the CSW packages at all. ?It's enough to add > /opt/csw/bin to $PATH in order to access the utilities. > > If you tried to install a package to a machine with read-only mounted > /opt/csw, I'm guessing that pkgadd would fail (is that true?), because > writing to the file wouldn't be possible. ?If you don't run pkgadd on > client machines, how are daemons started? ?The "empty input > cswinitsmf" idea above wouldn't solve that. you would never "install a package" on one of these nfs client machines. the empty input case is to solve two use cases. 1. when it gets automatically called in a zone, from a global pkgadd, when the maintainer specified a file in the class under /opt/csw 2. to allow the script to be called directly, by the admins of those nfs client machines, to "do the right thing" for demon setup on the clients. From maciej at opencsw.org Wed Jun 9 23:03:21 2010 From: maciej at opencsw.org (Maciej (Matchek) Blizinski) Date: Wed, 9 Jun 2010 22:03:21 +0100 Subject: [csw-maintainers] cswclassutils: location of template init scripts In-Reply-To: References: Message-ID: No dia 9 de Junho de 2010 19:16, Philip Brown escreveu: > you would never "install a package" on one of these nfs client > machines. the empty input case is to solve two use cases. > > 1. when it gets automatically called in a zone, from a global pkgadd, > when the maintainer specified a file in the class under /opt/csw > > 2. to allow the script to be called directly, by the admins of those > nfs client machines, to "do the right thing" for demon setup on the > clients. I see; the init file under /opt/csw is there for the NFS client machines. There's also the argument that you don't need to vary the init script between the zones, but it's not a strong one. The absence of the init file from /opt is a show stopper for an NFS setup; in the sparse zones scenario it doesn't pose any practical problem at all. I think that handling SMF via straightforward CAS is a simple an elegant solution, and it works well with the current file layout. The main point of the elegance being the 1:1 correspondence between the init files and CAS executions. Losing it would be a significant cost from the SMF handling point of view, if you think about debugging situations for instance. IIUC, the main problem with the NFS setup is the absence of the init script from /opt. If so, why not include one more copy of the init file under /opt/csw/etc/cswinitsmf/$PKGINST so that it can be used by admins of an NFS client system? I know that the counterargument is the duplication of data; but it's not much of a practical problem. If you worry about users not knowing which init file is used, you can add a README file in the /opt/csw/etc/cswinitsmf directory, saying that these are only for the shared NFS purposes and the main location is /etc/opt/csw/init.d. In this way, NFS admins would have a script to run, and all the packages using init scripts could just carry on with their lives. Maciej From phil at bolthole.com Wed Jun 9 23:56:48 2010 From: phil at bolthole.com (Philip Brown) Date: Wed, 9 Jun 2010 14:56:48 -0700 Subject: [csw-maintainers] cswclassutils: location of template init scripts In-Reply-To: References: Message-ID: On Wed, Jun 9, 2010 at 2:03 PM, Maciej (Matchek) Blizinski wrote: > > I see; the init file under /opt/csw is there for the NFS client machines. > > There's also the argument that you don't need to vary the init script > between the zones, but it's not a strong one. and more importantly: if the init script needs per-machine modifications... the maintainer should be allowing it by way of config files or smf variables, etc. The init script itself should never have to be edited by a sysadmin. > ?The > main point of the elegance being the 1:1 correspondence between the > init files and CAS executions. ?Losing it would be a significant cost > from the SMF handling point of view, if you think about debugging > situations for instance. yup. > > IIUC, the main problem with the NFS setup is the absence of the init > script from /opt. ?If so, why not include one more copy of the init > file under /opt/csw/etc/cswinitsmf/$PKGINST so that it can be used by > admins of an NFS client system? > We seem to be mostly agreeing in principle, and just discussing implementations. I am actually not sure if you are agreeing with my prior suggestion of "have the CAS duplicate it into /opt/csw/etc/cswinitsmf", or whether you are proposing an actual duplicate ship in the package itself. Please clarify. There is also potentialy a third option, which I have not tested yet (and need to run out the door soon :) In theory, if we had no prior installs/use of /etc/opt/csw/init.d, the straightforward thing to do might be "always put one file, in /opt/csw/etc/cswinitsmf" That gives the cleanliness of single-file use, and with a well written CAS, will work for everyone. So it seems the main question is how to handle our "legacy" /etc/opt/csw/init.d packages, without having to force rebuild of quite a few. I had an "interesting" thought: svr4 hackery ;-) Perhaps we can have the CAS script, dynamically RELOCATE, rather than copy, init scripts from /etc/opt/csw/init.d, to /opt/csw/etc/[...] By judicious use of removef, rm, and installf, I think this may be possible. Then, packages deployed after the new CSW installation, woudl have a uniform install location, even if the package itself was pre-standard. From phil at bolthole.com Thu Jun 10 07:11:20 2010 From: phil at bolthole.com (Philip Brown) Date: Wed, 9 Jun 2010 22:11:20 -0700 Subject: [csw-maintainers] new gtk2: in experimental, and on testing machines Message-ID: <20100610051120.GA67871@bolthole.com> I have just created new gtk packages for both x86 and sparc.. and with 64bit libs. Maintainers are hereby invited; nay, entreated ... to give them a spin. on testing9s, testing9x. and/or via your own personal hosts, and the package chain at (pkgutil -t, pkg-get -U -s ) http://mirror.opencsw.org/opencsw/experimental/x11-reloaded From dam at opencsw.org Thu Jun 10 15:35:27 2010 From: dam at opencsw.org (Dagobert Michelsen) Date: Thu, 10 Jun 2010 15:35:27 +0200 Subject: [csw-maintainers] Question regarding di(1) utility In-Reply-To: References: <1FF7E50C-2049-4377-8D87-3F1F3B5E4684@opencsw.org> Message-ID: <385B6035-E302-4987-8389-13FC9FFACEFA@opencsw.org> Hi Brad, Am 07.06.2010 um 15:47 schrieb Brad Lanam: > On Mon, Jun 7, 2010 at 01:26, Dagobert Michelsen > wrote: >>> Am 07.06.2010 um 04:11 schrieb Philip Brown: >>>> Am 07.06.2010 um 03:40 schrieb Jeffery Small: >>>> behavior. However, when using the new csw di, I am seeing some >>>> loopback >>>> mounts listed. For example, with the older di: > > Apparently I was writing code from memory and not double checking. > > Line 434 of di.c: #if ! SunOS > Should be: #if ! sun /* will do for now */ > > Looks like I introduced that bug in di-4.20. > Unfortunately, there's no easy fix other than recompiling. Recompiling *is* an easy fix ;-) An updated package is available from Please let me know if it works as expected so I can release it to current/. BTW, checking for 'sun' doesn't help, these are defined by Sun Studio: (from cc(1)): > These predefinitions are valid in all modes: > __BUILTIN_VA_ARG_INCR > __SUNPRO_C=0x590 > __SVR4(SPARC) > __SunOS(Solaris) > __amd64(x86 with-m64) > __gnu__linux(linux) > __i386(x86) > __linux(linux) > __linux__(linux) > __sparc(SPARC) > __sparcv9(with-m64) > __sun(Solaris) > __unix > __`uname -s`_`uname -r`(Solaris) > __x86_64(x86) > linux(x86,linux) I now check for __SunOS: > From 08eae7402b5bbc3dd9e3c3d1227f22c140014c00 Mon Sep 17 00:00:00 2001 > From: Dagobert Michelsen > Date: Thu, 10 Jun 2010 15:23:45 +0200 > Subject: [PATCH] Fix Solaris detection > > --- > di.c | 2 +- > 1 files changed, 1 insertions(+), 1 deletions(-) > > diff --git a/di.c b/di.c > index d5adef2..32c3502 100644 > --- a/di.c > +++ b/di.c > @@ -431,7 +431,7 @@ main (argc, argv) > /* change default display format here */ > diopts->dispBlockSize = DI_VAL_1024 * DI_VAL_1024; > diopts->flags = 0; > -#if ! SunOS /* Solaris loopback devices should be excluded */ > +#if ! __SunOS /* Solaris loopback devices should be excluded */ > diopts->flags |= DI_F_INCLUDE_LOOPBACK; > #endif > strcpy (diopts->sortType, "m"); /* default - sort by mount > point */ > -- > 1.7.1 > Best regards -- Dago From phil at bolthole.com Thu Jun 10 18:11:14 2010 From: phil at bolthole.com (Philip Brown) Date: Thu, 10 Jun 2010 09:11:14 -0700 Subject: [csw-maintainers] Question regarding di(1) utility In-Reply-To: <385B6035-E302-4987-8389-13FC9FFACEFA@opencsw.org> References: <1FF7E50C-2049-4377-8D87-3F1F3B5E4684@opencsw.org> <385B6035-E302-4987-8389-13FC9FFACEFA@opencsw.org> Message-ID: On Thu, Jun 10, 2010 at 6:35 AM, Dagobert Michelsen wrote: > > BTW, checking for 'sun' doesn't help, these are defined by Sun Studio: > (from cc(1)): > Some corrections: >> ? ? ? ? ?These predefinitions are valid in all modes: >... >> ? ? ? ? ? ? ? ? ? __SVR4(SPARC) __SVR4 is defined on both sparc and x86, with sun studio . both 11 and 12. as is __sun The nice thing is that gcc ALSO defines these. along with the more balanced __sun__. but that's not important :) Thats why I have documented on http://www.bolthole.com/solaris, that the best most portable way to auto-detect solaris, is #if defined (__SVR4) && defined (__sun) (using either one of them by itself, may technically trigger the clause on additional machine types, such as SVR4==HPUX(?), and __sun on ye olde sunos 1.x From jeff at cjsa.com Fri Jun 11 01:47:39 2010 From: jeff at cjsa.com (Jeffery Small) Date: Thu, 10 Jun 2010 23:47:39 GMT Subject: [csw-maintainers] Question regarding di(1) utility References: <1FF7E50C-2049-4377-8D87-3F1F3B5E4684@opencsw.org> <385B6035-E302-4987-8389-13FC9FFACEFA@opencsw.org> Message-ID: Dagobert Michelsen writes: >Recompiling *is* an easy fix ;-) An updated package is available from > >Please let me know if it works as expected so I can release it to >current/. > [...] I just installed rev. 4.24,REV=2010.06.02. Unfortunately, I'm still seeing the same results on my SPARC system. ------------------------------------------------------------------------- 23-> /opt/csw/bin/di Filesystem Mount /dev/md/dsk/d0 / swap /etc/svc/volatile /platform/sun4u-us /platform/sun4u-us3/lib/libc_psr.so.1 /platform/sun4u-us /platform/sun4u-us3/lib/sparcv9/libc_psr.so.1 swap /tmp /v/JS/images /u/jeff/lib/images /x/JS/music /u/jeff/lib/music /dev/md/dsk/d5 /v /dev/md/dsk/d3 /var swap /var/run /sadm /var/sadm /dev/md/dsk/d7 /x ------------------------------------------------------------------------- I truncated the above output lines for readability. It is still showing the loopback mounts. ;-( Regards, -- Jeff C. Jeffery Small CJSA LLC 206-232-3338 jeff at cjsa.com 7000 E Mercer Way, Mercer Island, WA 98040 From dam at opencsw.org Fri Jun 11 10:53:50 2010 From: dam at opencsw.org (Dagobert Michelsen) Date: Fri, 11 Jun 2010 10:53:50 +0200 Subject: [csw-maintainers] Coreutils and CSWgfile Message-ID: Hi, it looks like the new CSWcoreutils is marked compatible with the old CSWgfile. Unfortunately, there is now a new CSWgfile which is AFAIK no longer incompatible as it contains other things: > A local copy of CSWcoreutils-8.4,REV=2010.05.20 exists and is of > matching size. > 3 incompatible packages to remove. Do you want to continue? [Y,n] > => Removing incompatible package CSWgfile > > The following package is currently installed: > CSWgfile fileutils - GNU file utilities > (sparc) 4.1,REV=2003.01.23 > Ben, Maciej, what do you think? Best regards -- Dago From bonivart at opencsw.org Fri Jun 11 11:00:52 2010 From: bonivart at opencsw.org (Peter Bonivart) Date: Fri, 11 Jun 2010 11:00:52 +0200 Subject: [csw-maintainers] Coreutils and CSWgfile In-Reply-To: References: Message-ID: On Fri, Jun 11, 2010 at 10:53 AM, Dagobert Michelsen wrote: > Hi, > > it looks like the new CSWcoreutils is marked compatible with the old > CSWgfile. > Unfortunately, there is now a new CSWgfile which is AFAIK no longer > incompatible > as it contains other things: > >> A local copy of CSWcoreutils-8.4,REV=2010.05.20 exists and is of matching >> size. >> 3 incompatible packages to remove. Do you want to continue? [Y,n] >> => Removing incompatible package CSWgfile >> >> The following package is currently installed: >> ? CSWgfile ? ? ? ?fileutils - GNU file utilities >> ? ? ? ? ? ? ? ? ? (sparc) 4.1,REV=2003.01.23 On a related note I just had to uninstall CSWfile on testing9s and testing9x to get CSWlibmagic on there. CSWlibmagic collided with files from CSWfile. CSWlibmagic should have CSWfile listed as an I-dep. -- /peter From maciej at opencsw.org Fri Jun 11 11:02:08 2010 From: maciej at opencsw.org (Maciej (Matchek) Blizinski) Date: Fri, 11 Jun 2010 10:02:08 +0100 Subject: [csw-maintainers] Coreutils and CSWgfile In-Reply-To: References: Message-ID: No dia 11 de Junho de 2010 09:53, Dagobert Michelsen escreveu: > Hi, > > it looks like the new CSWcoreutils is marked compatible with the old > CSWgfile. > Unfortunately, there is now a new CSWgfile which is AFAIK no longer > incompatible > as it contains other things: > >> A local copy of CSWcoreutils-8.4,REV=2010.05.20 exists and is of matching >> size. >> 3 incompatible packages to remove. Do you want to continue? [Y,n] >> => Removing incompatible package CSWgfile >> >> The following package is currently installed: >> ? CSWgfile ? ? ? ?fileutils - GNU file utilities >> ? ? ? ? ? ? ? ? ? (sparc) 4.1,REV=2003.01.23 >> > > > Ben, Maciej, what do you think? I think it's best to remove the incompatibility flag. I've submitted a bug already: http://www.opencsw.org/bugtrack/view.php?id=4448 From maciej at opencsw.org Fri Jun 11 11:08:08 2010 From: maciej at opencsw.org (Maciej (Matchek) Blizinski) Date: Fri, 11 Jun 2010 10:08:08 +0100 Subject: [csw-maintainers] Coreutils and CSWgfile In-Reply-To: References: Message-ID: No dia 11 de Junho de 2010 10:00, Peter Bonivart escreveu: > On a related note I just had to uninstall CSWfile on testing9s and > testing9x to get CSWlibmagic on there. CSWlibmagic collided with files > from CSWfile. CSWlibmagic should have CSWfile listed as an I-dep. CSWfile is no more, there's only CSWgfile now. Yes, it had conflicting files at some point. It's best to remove all its instances. From bonivart at opencsw.org Fri Jun 11 11:10:22 2010 From: bonivart at opencsw.org (Peter Bonivart) Date: Fri, 11 Jun 2010 11:10:22 +0200 Subject: [csw-maintainers] Coreutils and CSWgfile In-Reply-To: References: Message-ID: On Fri, Jun 11, 2010 at 11:08 AM, Maciej (Matchek) Blizinski wrote: > No dia 11 de Junho de 2010 10:00, Peter Bonivart > escreveu: >> On a related note I just had to uninstall CSWfile on testing9s and >> testing9x to get CSWlibmagic on there. CSWlibmagic collided with files >> from CSWfile. CSWlibmagic should have CSWfile listed as an I-dep. > > CSWfile is no more, there's only CSWgfile now. ?Yes, it had > conflicting files at some point. ?It's best to remove all its > instances. You mean it's out of the catalog? It can still be on customer machines though (just like it was on our testing servers). I think an I-dep would help. -- /peter From maciej at opencsw.org Fri Jun 11 11:36:50 2010 From: maciej at opencsw.org (Maciej (Matchek) Blizinski) Date: Fri, 11 Jun 2010 10:36:50 +0100 Subject: [csw-maintainers] Coreutils and CSWgfile In-Reply-To: References: Message-ID: No dia 11 de Junho de 2010 10:10, Peter Bonivart escreveu: > You mean it's out of the catalog? No, it's not in the catalog, it never was. > It can still be on customer machines > though (just like it was on our testing servers). I think an I-dep > would help. There's no worry of CSWfile on consumer machines. From bwalton at opencsw.org Fri Jun 11 13:56:19 2010 From: bwalton at opencsw.org (Ben Walton) Date: Fri, 11 Jun 2010 07:56:19 -0400 Subject: [csw-maintainers] Coreutils and CSWgfile In-Reply-To: References: Message-ID: <1276257283-sup-2586@pinkfloyd.chass.utoronto.ca> Excerpts from Maciej (Matchek) Blizinski's message of Fri Jun 11 05:02:08 -0400 2010: > I think it's best to remove the incompatibility flag. I've > submitted a bug already: > http://www.opencsw.org/bugtrack/view.php?id=4448 The change is in flight right now. I should have updated packages tonight. This will cause problems for sites that have (old) CSWgfile installed and then install CSWcoreutils without doing a full update... Thanks -Ben -- Ben Walton Systems Programmer - CHASS University of Toronto C:416.407.5610 | W:416.978.4302 From phil at bolthole.com Fri Jun 11 15:05:57 2010 From: phil at bolthole.com (Philip Brown) Date: Fri, 11 Jun 2010 06:05:57 -0700 Subject: [csw-maintainers] Coreutils and CSWgfile In-Reply-To: References: Message-ID: > On a related note I just had to uninstall CSWfile on testing9s and > testing9x to get CSWlibmagic on there. CSWlibmagic collided with files > from CSWfile. CSWlibmagic should have CSWfile listed as an I-dep. > > hmmm. that's odd. if there are file conflicts with CSWfile(aka fileutils) I would expect it to also conflict with coreutils. no? From phil at bolthole.com Fri Jun 11 15:11:47 2010 From: phil at bolthole.com (Philip Brown) Date: Fri, 11 Jun 2010 06:11:47 -0700 Subject: [csw-maintainers] Coreutils and CSWgfile In-Reply-To: <1276257283-sup-2586@pinkfloyd.chass.utoronto.ca> References: <1276257283-sup-2586@pinkfloyd.chass.utoronto.ca> Message-ID: On Friday, June 11, 2010, Ben Walton wrote: > Excerpts from Maciej (Matchek) Blizinski's message of Fri Jun 11 05:02:08 -0400 2010: > >> I think it's best to remove the incompatibility flag. ?I've >> submitted a bug already: >> http://www.opencsw.org/bugtrack/view.php?id=4448 > > The change is in flight right now. ?I should have updated packages > tonight. > > This will cause problems for sites that have (old) CSWgfile installed > and then install CSWcoreutils without doing a full update... please remember that CSWfile and CSWgfile are/were different things From bwalton at opencsw.org Fri Jun 11 15:26:23 2010 From: bwalton at opencsw.org (Ben Walton) Date: Fri, 11 Jun 2010 09:26:23 -0400 Subject: [csw-maintainers] Coreutils and CSWgfile In-Reply-To: References: <1276257283-sup-2586@pinkfloyd.chass.utoronto.ca> Message-ID: <1276262748-sup-3873@pinkfloyd.chass.utoronto.ca> Excerpts from Philip Brown's message of Fri Jun 11 09:11:47 -0400 2010: > please remember that CSWfile and CSWgfile are/were different things Please remember that this part of the thread is dealing with the I-dep currently in CSWcoreutils against CSWgfile. :) Thanks -Ben -- Ben Walton Systems Programmer - CHASS University of Toronto C:416.407.5610 | W:416.978.4302 From maciej at opencsw.org Fri Jun 11 15:38:32 2010 From: maciej at opencsw.org (Maciej (Matchek) Blizinski) Date: Fri, 11 Jun 2010 14:38:32 +0100 Subject: [csw-maintainers] cswclassutils: location of template init scripts In-Reply-To: References: Message-ID: No dia 9 de Junho de 2010 22:56, Philip Brown escreveu: > On Wed, Jun 9, 2010 at 2:03 PM, Maciej (Matchek) Blizinski > wrote: >> >> I see; the init file under /opt/csw is there for the NFS client machines. >> >> There's also the argument that you don't need to vary the init script >> between the zones, but it's not a strong one. > > and more importantly: if the init script needs per-machine > modifications... the maintainer should be allowing it by way of config > files or smf variables, etc. The init script itself should never have > to be edited by a sysadmin. Right, this means that it can be installed into /opt, but it doesn't mean that it has to. It's sometimes convenient to edit the init file for debugging purposes. In debugging context, private init file is better than a global one (you don't influence other zones). >> ?The >> main point of the elegance being the 1:1 correspondence between the >> init files and CAS executions. ?Losing it would be a significant cost >> from the SMF handling point of view, if you think about debugging >> situations for instance. > > yup. > > >> >> IIUC, the main problem with the NFS setup is the absence of the init >> script from /opt. ?If so, why not include one more copy of the init >> file under /opt/csw/etc/cswinitsmf/$PKGINST so that it can be used by >> admins of an NFS client system? >> > > We seem to be mostly agreeing in principle, and just discussing implementations. > I am actually not sure if you are agreeing with my prior suggestion of > "have the CAS duplicate it into /opt/csw/etc/cswinitsmf", or whether > you are proposing an actual duplicate ship in the package itself. > Please clarify. For the simplest solution, I suggest shipping duplicates, just because it's easy to implement. It has some drawbacks, but now show stoppers. Things underneath /etc are mostly about triggering script execution, and giving them data (init script paths). The convenience lies in pkgadd understanding the concept of zones and abstracting that away. > > > There is also potentialy a third option, which I have not tested yet > (and need to run out the door soon :) > > In theory, if we had no prior installs/use of /etc/opt/csw/init.d, the > straightforward thing to do might be "always put one file, in > /opt/csw/etc/cswinitsmf" > That gives the cleanliness of single-file use, and with a well written > CAS, will work for everyone. > > So it seems the main question is how to handle our "legacy" > /etc/opt/csw/init.d packages, without having to force rebuild of quite > a few. Adding a duplicate in /opt/csw/etc/cswinitsmf would avoid those problems. > I had an "interesting" thought: svr4 hackery ;-) > > Perhaps we can have the CAS script, dynamically RELOCATE, rather than > copy, init scripts from /etc/opt/csw/init.d, to /opt/csw/etc/[...] > > By judicious use of ?removef, rm, and installf, I think this may be possible. > > Then, packages deployed after the new CSW installation, woudl have a > uniform install location, even if the package itself was pre-standard. Can you tell more about this idea? What kind of code would handle that? (A shell script? Where would it be?) My last thought is that if we want there to be a single location of the init scripts, perhaps we shouldn't use CAS, but just postinstall scripts. For convenience, we could use some sort of automation around pulling SMF handling function from a common source and pasting them into the individual package's postinstall. In any case, I'm weary of throwing away or significantly modifying what we have right now, just because of the NFS case. I'd rather add whatever NFS setup needs, without disturbing the existing solution. Maciej From phil at bolthole.com Fri Jun 11 16:15:20 2010 From: phil at bolthole.com (Philip Brown) Date: Fri, 11 Jun 2010 07:15:20 -0700 Subject: [csw-maintainers] Coreutils and CSWgfile In-Reply-To: <1276262748-sup-3873@pinkfloyd.chass.utoronto.ca> References: <1276257283-sup-2586@pinkfloyd.chass.utoronto.ca> <1276262748-sup-3873@pinkfloyd.chass.utoronto.ca> Message-ID: I'm rather confused why it had a conflict declared since we never had gfile before On Friday, June 11, 2010, Ben Walton wrote: > Excerpts from Philip Brown's message of Fri Jun 11 09:11:47 -0400 2010: > >> please remember that CSWfile and CSWgfile are/were different things > > Please remember that this part of the thread is dealing with the I-dep > currently in CSWcoreutils against CSWgfile. :) > > Thanks > -Ben > -- > Ben Walton > Systems Programmer - CHASS > University of Toronto > C:416.407.5610 | W:416.978.4302 > > _______________________________________________ > maintainers mailing list > maintainers at lists.opencsw.org > https://lists.opencsw.org/mailman/listinfo/maintainers > .:: This mailing list's archive is public. ::. > From bwalton at opencsw.org Fri Jun 11 16:25:47 2010 From: bwalton at opencsw.org (Ben Walton) Date: Fri, 11 Jun 2010 10:25:47 -0400 Subject: [csw-maintainers] Coreutils and CSWgfile In-Reply-To: References: <1276257283-sup-2586@pinkfloyd.chass.utoronto.ca> <1276262748-sup-3873@pinkfloyd.chass.utoronto.ca> Message-ID: <1276266246-sup-5438@pinkfloyd.chass.utoronto.ca> Excerpts from Philip Brown's message of Fri Jun 11 10:15:20 -0400 2010: > I'm rather confused why it had a conflict declared since we never > had gfile before I think this highlights my point about re-using this package name rather nicely! We _did_ have CSWgfile before. It was the gnu fileutils package. We now have a brand new (hot on the heels of the old) CSWgfile providing the well known `file` utility. The I-dep in CSWcoreutils is to prevent it from being installed along side the deprecated CSWgfile (fileutils). -Ben -- Ben Walton Systems Programmer - CHASS University of Toronto C:416.407.5610 | W:416.978.4302 From phil at bolthole.com Fri Jun 11 17:50:32 2010 From: phil at bolthole.com (Philip Brown) Date: Fri, 11 Jun 2010 08:50:32 -0700 Subject: [csw-maintainers] Coreutils and CSWgfile In-Reply-To: <1276266246-sup-5438@pinkfloyd.chass.utoronto.ca> References: <1276257283-sup-2586@pinkfloyd.chass.utoronto.ca> <1276262748-sup-3873@pinkfloyd.chass.utoronto.ca> <1276266246-sup-5438@pinkfloyd.chass.utoronto.ca> Message-ID: On Fri, Jun 11, 2010 at 7:25 AM, Ben Walton wrote: > Excerpts from Philip Brown's message of Fri Jun 11 10:15:20 -0400 2010: > >> I'm rather confused why it had a conflict declared since we never >> had gfile before > > I think this highlights my point about re-using this package name > rather nicely! ?We _did_ have CSWgfile before. ?It was the gnu > fileutils package. Nuts. I thought that was "CSWfile", not "CSWgfile", otherwise, I would not have accepted in "CSWgfile" as the pkg name for the new "file" package! What is the "best" way to fix this?!?!?? GAaaaahhhhhh..... *brain explodes*. From phil at bolthole.com Fri Jun 11 18:22:15 2010 From: phil at bolthole.com (Philip Brown) Date: Fri, 11 Jun 2010 09:22:15 -0700 Subject: [csw-maintainers] cswclassutils: location of template init scripts In-Reply-To: References: Message-ID: On Fri, Jun 11, 2010 at 6:38 AM, Maciej (Matchek) Blizinski wrote: > >> So it seems the main question is how to handle our "legacy" >> /etc/opt/csw/init.d packages, without having to force rebuild of quite >> a few. > > Adding a duplicate in /opt/csw/etc/cswinitsmf would avoid those problems. yup. i just think we would be better off automating the duplication, via CAS >> I had an "interesting" thought: svr4 hackery ;-) >> >> Perhaps we can have the CAS script, dynamically RELOCATE, rather than >> copy, init scripts from /etc/opt/csw/init.d, to /opt/csw/etc/[...] >> >> By judicious use of ?removef, rm, and installf, I think this may be possible. >> >> Then, packages deployed after the new CSW installation, woudl have a >> uniform install location, even if the package itself was pre-standard. > > Can you tell more about this idea? What kind of code would handle > that? ?(A shell script? Where would it be?) The CAS. pseudo-code: [start of read filename from stdin loop] case $filename in /etc/opt/csw/init.d/*) NEWfilename=/opt/csw/init.d/cswinitsmf/$PKGINST/`basename $filename` cp $filename $NEWfilename installf $NEWfilename removef $filename # not sure this will be allowed though ;-) filename=$NEWfilename ;; esac ## now proceed to do "normal" init processing on $filename, either way. ## Of course, the above would have to be tweaked for chroot, blah blah blah. ## If the removef is disallowed, above.. .then we simply "ignore" files under /etc/opt/csw ## when it comes to remove time. we presume that they get handled by their automatically ## duplicated copy in /opt/csw/init.d ? IF /opt/csw/init.d/cswinitsmf exists, that is. ## That handles the case of legacy installs done before newer version of cswinitsmf installed. > In any case, I'm weary of throwing away or significantly modifying > what we have right now, just because of the NFS case. ?I'd rather add > whatever NFS setup needs, without disturbing the existing solution. > I believe the above proposal meets that criteria. The only "significant modification" is a one-time alteration of the cswinitsmf CAS package. From maciej at opencsw.org Fri Jun 11 18:49:53 2010 From: maciej at opencsw.org (Maciej (Matchek) Blizinski) Date: Fri, 11 Jun 2010 17:49:53 +0100 Subject: [csw-maintainers] Coreutils and CSWgfile In-Reply-To: References: <1276257283-sup-2586@pinkfloyd.chass.utoronto.ca> <1276262748-sup-3873@pinkfloyd.chass.utoronto.ca> <1276266246-sup-5438@pinkfloyd.chass.utoronto.ca> Message-ID: No dia 11 de Junho de 2010 16:50, Philip Brown escreveu: > On Fri, Jun 11, 2010 at 7:25 AM, Ben Walton wrote: >> Excerpts from Philip Brown's message of Fri Jun 11 10:15:20 -0400 2010: >> >>> I'm rather confused why it had a conflict declared since we never >>> had gfile before >> >> I think this highlights my point about re-using this package name >> rather nicely! ?We _did_ have CSWgfile before. ?It was the gnu >> fileutils package. > > Nuts. I thought that was "CSWfile", not "CSWgfile", otherwise, I would > not have accepted in > "CSWgfile" as the pkg name for the new "file" package! I'd like to ask some more questions. IIUC, it can't happen automatically, the administrator has to know about the upgrade and handle certain steps in some way outside the standard tools. The update to the current CSWcoreutils needs to go as follows: 1. If they are installed, remove CSWshutils, CSWgfile and CSWtextutils 2. Update the local catalog cache 3. Install CSWcoreutils or do an automated upgrade, during which CSWcoreutils will be pulled as a dependency to some other package. In what way having new CSWgfile breaks this? Maciej From bwalton at opencsw.org Fri Jun 11 18:55:47 2010 From: bwalton at opencsw.org (Ben Walton) Date: Fri, 11 Jun 2010 12:55:47 -0400 Subject: [csw-maintainers] Coreutils and CSWgfile In-Reply-To: References: <1276257283-sup-2586@pinkfloyd.chass.utoronto.ca> <1276262748-sup-3873@pinkfloyd.chass.utoronto.ca> <1276266246-sup-5438@pinkfloyd.chass.utoronto.ca> Message-ID: <1276275211-sup-4805@pinkfloyd.chass.utoronto.ca> Excerpts from Maciej (Matchek) Blizinski's message of Fri Jun 11 12:49:53 -0400 2010: > 1. If they are installed, remove CSWshutils, CSWgfile and CSWtextutils > 2. Update the local catalog cache > 3. Install CSWcoreutils or do an automated upgrade, during which > CSWcoreutils will be pulled as a dependency to some other package. > > In what way having new CSWgfile breaks this? Say they've got CSWgfile (fileutils) installed. They update their catalog and then install CSWcoreutils. This is _without_ doing a full upgrade, just an install. As I'll be removing the I between CSWgfile and CSWcoreutils, the admin could now see file conflicts between the two packages. It all comes down to _how_ the admin adds CSWcoreutils... -Ben -- Ben Walton Systems Programmer - CHASS University of Toronto C:416.407.5610 | W:416.978.4302 From maciej at opencsw.org Fri Jun 11 19:09:17 2010 From: maciej at opencsw.org (Maciej (Matchek) Blizinski) Date: Fri, 11 Jun 2010 18:09:17 +0100 Subject: [csw-maintainers] Coreutils and CSWgfile In-Reply-To: <1276275211-sup-4805@pinkfloyd.chass.utoronto.ca> References: <1276257283-sup-2586@pinkfloyd.chass.utoronto.ca> <1276262748-sup-3873@pinkfloyd.chass.utoronto.ca> <1276266246-sup-5438@pinkfloyd.chass.utoronto.ca> <1276275211-sup-4805@pinkfloyd.chass.utoronto.ca> Message-ID: No dia 11 de Junho de 2010 17:55, Ben Walton escreveu: > Excerpts from Maciej (Matchek) Blizinski's message of Fri Jun 11 12:49:53 -0400 2010: > >> 1. If they are installed, remove CSWshutils, CSWgfile and CSWtextutils >> 2. Update the local catalog cache >> 3. Install CSWcoreutils or do an automated upgrade, during which >> CSWcoreutils will be pulled as a dependency to some other package. >> >> In what way having new CSWgfile breaks this? > > Say they've got CSWgfile (fileutils) installed. ?They update their > catalog and then install CSWcoreutils. ?This is _without_ doing a full > upgrade, just an install. ?As I'll be removing the I between CSWgfile > and CSWcoreutils, the admin could now see file conflicts between the > two packages. ?It all comes down to _how_ the admin adds > CSWcoreutils... I see. It's the case in which the admin doesn't remove any packages by hand, but only installs CSWcoreutils. In this way, the old conflicting CSWgfile might still be installed. I don't think it's the right way to manage packages (mixing packages from old and new catalogs), but I agree that it might happen in practice. How about shipping a preinstall script which detects the version of CSWgfile and displays a helpful error message when CSWgfile is old? From phil at bolthole.com Fri Jun 11 19:40:39 2010 From: phil at bolthole.com (Philip Brown) Date: Fri, 11 Jun 2010 10:40:39 -0700 Subject: [csw-maintainers] Coreutils and CSWgfile In-Reply-To: References: <1276257283-sup-2586@pinkfloyd.chass.utoronto.ca> <1276262748-sup-3873@pinkfloyd.chass.utoronto.ca> <1276266246-sup-5438@pinkfloyd.chass.utoronto.ca> <1276275211-sup-4805@pinkfloyd.chass.utoronto.ca> Message-ID: the main problem here is that the old fileutils was misnamed. we try very hard not to reuse pkg names. if you can think of an appropriate new one, that may be the best solution we have sometimes used the 'n' prefix. if you chose that then it would be appropriate to rename the /opt/csw/bin binary as well, to nfile From Darin.Perusich at cognigencorp.com Fri Jun 11 19:51:14 2010 From: Darin.Perusich at cognigencorp.com (Darin Perusich) Date: Fri, 11 Jun 2010 13:51:14 -0400 Subject: [csw-maintainers] i.cswinet install script failing Message-ID: <4C127792.6000608@cognigencorp.com> When I install the Amanda package the cswinet class script doesn't complete successfully, which wasn't happening previously. I didn't see anything about this in the bug tracker so I figured I'd ping the list before submitting a bug. The output I'm getting is: Installing class ... /opt/csw/etc/pkg/CSWamanda/services [ verifying class ] Installing class ... /opt/csw/etc/pkg/CSWamanda/inetd.conf svcname: amanda; proto: udp Service 'amanda' not found in /etc/inet/services. Won't setup inetd for it. pkgadd: ERROR: class action script did not complete successfully [ verifying class ] The amanda services are defined in the services file. everest:~$ grep 1008[023] /etc/inet/services amanda 10080/tcp # CSWamanda amanda 10080/udp # CSWamanda amandaidx 10082/tcp # CSWamanda amidxtape 10083/tcp # CSWamanda -- Darin Perusich Unix Systems Administrator Cognigen Corporation 395 Youngs Rd. Williamsville, NY 14221 Phone: 716-633-3463 Email: darinper at cognigencorp.com From bwalton at opencsw.org Fri Jun 11 19:58:20 2010 From: bwalton at opencsw.org (Ben Walton) Date: Fri, 11 Jun 2010 13:58:20 -0400 Subject: [csw-maintainers] i.cswinet install script failing In-Reply-To: <4C127792.6000608@cognigencorp.com> References: <4C127792.6000608@cognigencorp.com> Message-ID: <1276278926-sup-163@pinkfloyd.chass.utoronto.ca> Excerpts from Darin Perusich's message of Fri Jun 11 13:51:14 -0400 2010: Hi Darin, > Service 'amanda' not found in /etc/inet/services. Won't setup inetd for it. > pkgadd: ERROR: class action script did not complete successfully > [ verifying class ] Which version of cswclassutils are you running? There were a few bugs fixed around this issue recently. If you're not running the current version of the package, update and let us know if the issue persists. Thanks -Ben -- Ben Walton Systems Programmer - CHASS University of Toronto C:416.407.5610 | W:416.978.4302 From Darin.Perusich at cognigencorp.com Fri Jun 11 20:01:50 2010 From: Darin.Perusich at cognigencorp.com (Darin Perusich) Date: Fri, 11 Jun 2010 14:01:50 -0400 Subject: [csw-maintainers] i.cswinet install script failing In-Reply-To: <1276278926-sup-163@pinkfloyd.chass.utoronto.ca> References: <4C127792.6000608@cognigencorp.com> <1276278926-sup-163@pinkfloyd.chass.utoronto.ca> Message-ID: <4C127A0E.4000005@cognigencorp.com> Hey Ben, On 06/11/2010 01:58 PM, Ben Walton wrote: > > Which version of cswclassutils are you running? There were a few bugs > fixed around this issue recently. If you're not running the current > version of the package, update and let us know if the issue persists. I'm running 1.37,REV=2010.05.27 which is the latest release -- Darin Perusich Unix Systems Administrator Cognigen Corporation 395 Youngs Rd. Williamsville, NY 14221 Phone: 716-633-3463 Email: darinper at cognigencorp.com From bwalton at opencsw.org Fri Jun 11 22:32:29 2010 From: bwalton at opencsw.org (Ben Walton) Date: Fri, 11 Jun 2010 16:32:29 -0400 Subject: [csw-maintainers] i.cswinet install script failing In-Reply-To: <4C127A0E.4000005@cognigencorp.com> References: <4C127792.6000608@cognigencorp.com> <1276278926-sup-163@pinkfloyd.chass.utoronto.ca> <4C127A0E.4000005@cognigencorp.com> Message-ID: <1276288316-sup-7855@pinkfloyd.chass.utoronto.ca> Excerpts from Darin Perusich's message of Fri Jun 11 14:01:50 -0400 2010: > I'm running 1.37,REV=2010.05.27 which is the latest release Doh! No easy out. I'll try and look at this tonight. Can you poke an official bug report in for it? Thanks -Ben -- Ben Walton Systems Programmer - CHASS University of Toronto C:416.407.5610 | W:416.978.4302 From bwalton at opencsw.org Sat Jun 12 02:37:41 2010 From: bwalton at opencsw.org (Ben Walton) Date: Fri, 11 Jun 2010 20:37:41 -0400 Subject: [csw-maintainers] i.cswinet install script failing In-Reply-To: <1276288316-sup-7855@pinkfloyd.chass.utoronto.ca> References: <4C127792.6000608@cognigencorp.com> <1276278926-sup-163@pinkfloyd.chass.utoronto.ca> <4C127A0E.4000005@cognigencorp.com> <1276288316-sup-7855@pinkfloyd.chass.utoronto.ca> Message-ID: <1276302909-sup-4068@pinkfloyd.chass.utoronto.ca> Excerpts from Ben Walton's message of Fri Jun 11 16:32:29 -0400 2010: Hi Darin, Found the bug. When I switched from using grep to awk to detect lines in /etc/services, the 'filter' got less specific. It was then triggering on more than one line (where applicable). This was seeing the test for the detection fail. I've re-tightened the criteria for this check and it seems to work properly now. There is an update in my catalog, so please test it. If it's good, I'll bump Peter to release a new version. http://mirror.opencsw.org/opencsw/experimental/bwalton/ Thanks. -Ben -- Ben Walton Systems Programmer - CHASS University of Toronto C:416.407.5610 | W:416.978.4302 From maciej at opencsw.org Sat Jun 12 11:15:48 2010 From: maciej at opencsw.org (Maciej (Matchek) Blizinski) Date: Sat, 12 Jun 2010 10:15:48 +0100 Subject: [csw-maintainers] Is there any interest in vixie cron? Message-ID: vixie cron was my first OpenCSW build, although I've never released it. At the time, I was interested in cron.d support. Later on I've learned that Puppet supports crontab editing on Solaris, so cron.d support was no longer necessary. It's been sitting on a shelf for a year. I've recently seen someone on IRC mentioning efforts to port vixie cron, so there might be still interest in vixie cron from other people. The main nice features of vixie cron is cron.d support (which comes from the Debian patch) and fractional notation support, e.g. */5 denotes "every five minutes". cron.d is especially nice, because it allows to drop files into directories instead of editing crontabs. This way, there's no need to use CAS to have cronjob support in packages. Are there any people here interested in testing vixie cron? Maciej From bonivart at opencsw.org Sat Jun 12 11:24:04 2010 From: bonivart at opencsw.org (Peter Bonivart) Date: Sat, 12 Jun 2010 11:24:04 +0200 Subject: [csw-maintainers] Is there any interest in vixie cron? In-Reply-To: References: Message-ID: On Sat, Jun 12, 2010 at 11:15 AM, Maciej (Matchek) Blizinski wrote: > Are there any people here interested in testing vixie cron? It's not absolutely necessary but it sure is nice. :-) Did you have it working so we can test it or is it tough to port? If it's ready to test I can help test it and I think it should be released if it works. It would be one less gripe for the Linux crowd when using Solaris. -- /peter From maciej at opencsw.org Sat Jun 12 11:41:47 2010 From: maciej at opencsw.org (Maciej (Matchek) Blizinski) Date: Sat, 12 Jun 2010 10:41:47 +0100 Subject: [csw-maintainers] Is there any interest in vixie cron? In-Reply-To: References: Message-ID: No dia 12 de Junho de 2010 10:24, Peter Bonivart escreveu: > On Sat, Jun 12, 2010 at 11:15 AM, Maciej (Matchek) Blizinski > wrote: >> Are there any people here interested in testing vixie cron? > > It's not absolutely necessary but it sure is nice. :-) > > Did you have it working so we can test it or is it tough to port? I have it working on Solaris 10, but Solaris 9 needs a bit more work. Specifically, it doesn't have O_NOFOLLOW option for the open() syscall. I've found a workaround[1] that minimizes the race condition problem, by checking the device and inode numbers. I need to rewrite the section where crontabs are opened. There are more tweaks needed, but I can probably get the first version working today. [1] https://www.securecoding.cert.org/confluence/display/seccode/POS01-C.+Check+for+the+existence+of+links+when+dealing+with+files From bll6969 at gmail.com Mon Jun 7 15:47:39 2010 From: bll6969 at gmail.com (Brad Lanam) Date: Mon, 7 Jun 2010 06:47:39 -0700 Subject: [csw-maintainers] Question regarding di(1) utility In-Reply-To: <1FF7E50C-2049-4377-8D87-3F1F3B5E4684@opencsw.org> References: <1FF7E50C-2049-4377-8D87-3F1F3B5E4684@opencsw.org> Message-ID: On Mon, Jun 7, 2010 at 01:26, Dagobert Michelsen wrote: >> Am 07.06.2010 um 04:11 schrieb Philip Brown: >>> Am 07.06.2010 um 03:40 schrieb Jeffery Small: >>> behavior. ?However, when using the new csw di, I am seeing some loopback >>> mounts listed. ?For example, with the older di: Apparently I was writing code from memory and not double checking. Line 434 of di.c: #if ! SunOS Should be: #if ! sun /* will do for now */ Looks like I introduced that bug in di-4.20. Unfortunately, there's no easy fix other than recompiling. -- Brad From phil at bolthole.com Sat Jun 12 16:36:09 2010 From: phil at bolthole.com (Philip Brown) Date: Sat, 12 Jun 2010 07:36:09 -0700 Subject: [csw-maintainers] Is there any interest in vixie cron? In-Reply-To: References: Message-ID: On Saturday, June 12, 2010, Maciej (Matchek) wrote... > This way, there's no need to use CAS to have cronjob support in > packages. > errr. I think we'd still have to support the traditional cron :-} From phil at bolthole.com Sat Jun 12 18:05:19 2010 From: phil at bolthole.com (Philip Brown) Date: Sat, 12 Jun 2010 09:05:19 -0700 Subject: [csw-maintainers] need help with C++ cmopile problem Message-ID: <20100612160518.GA63751@bolthole.com> Oddity that I have run into trying to compile a library. simplified example below. The following compiles with sun cc: typedef struct fooo{ int filler; int testing []; } foostruct; it does NOT compile, however, with sun CC. "test.c", line 4: Error: In this declaration "testing" is of an incomplete type "int[]". Any suggestions on what I can do? putting it within extern "C"{} does not appear to help. Which is sad, because that IS actually how it is used in the actual library code. From maciej at opencsw.org Sat Jun 12 22:00:32 2010 From: maciej at opencsw.org (Maciej (Matchek) Blizinski) Date: Sat, 12 Jun 2010 21:00:32 +0100 Subject: [csw-maintainers] Is there any interest in vixie cron? In-Reply-To: References: Message-ID: No dia 12 de Junho de 2010 15:36, Philip Brown escreveu: > On Saturday, June 12, 2010, Maciej (Matchek) wrote... >> This way, there's no need to use CAS to have cronjob support in >> packages. >> > > > errr. I think we'd still have to support the traditional cron :-} 'We' as in OpenCSW -- yes. However, various in-house custom installations could have own packages, and additional configuration management systems used to place files on the filesystem. These would benefit from simplified cron management. The first version is available for testing: http://mirror.opencsw.org/experimental.html#maciej Known issues: - uses /etc, not /etc/opt/csw - man pages are in wrong directories - preinstall / postremove scripts assume SMF, no Solaris 9 support Otherwise, it should already give out some vibes of working. Maciej From phil at bolthole.com Sat Jun 12 23:38:22 2010 From: phil at bolthole.com (Philip Brown) Date: Sat, 12 Jun 2010 14:38:22 -0700 Subject: [csw-maintainers] new gimp! kinda... needs testing Message-ID: <20100612213822.GA88999@bolthole.com> I havent gotten any feedback on the new gtk packages I made :( So... I've cobbled together a quickie build of gimp 2.6.8 I'm HOPING this will encourage someone to try it out :-} Please folks.. i really need someone else's feedback on the new gtk packages before I offically release them. gimp is a large multipackge build to do "properly" (and its not in svn, btw) so its a little bit of a hack job. But you should be able to get it with your choice of pkg-get -U -s http://mirror.opencsw.org/opencsw/experimental/x11-reloaded \ -u gimp_gtk or pkgutil -t http://mirror.opencsw.org/opencsw/experimental/x11-reloaded -i \ gimp_gtk From jeff at cjsa.com Sun Jun 13 06:03:54 2010 From: jeff at cjsa.com (Jeffery Small) Date: Sun, 13 Jun 2010 04:03:54 GMT Subject: [csw-maintainers] new gimp! kinda... needs testing References: <20100612213822.GA88999@bolthole.com> Message-ID: Philip Brown writes: >I havent gotten any feedback on the new gtk packages I made :( >So... I've cobbled together a quickie build of gimp 2.6.8 >I'm HOPING this will encourage someone to try it out :-} >Please folks.. i really need someone else's feedback on the new gtk >packages before I offically release them. >gimp is a large multipackge build to do "properly" (and its not in svn, >btw) so its a little bit of a hack job. But you should be able to get it >with your choice of >pkg-get -U -s http://mirror.opencsw.org/opencsw/experimental/x11-reloaded \ > -u gimp_gtk >or >pkgutil -t http://mirror.opencsw.org/opencsw/experimental/x11-reloaded -i \ > gimp_gtk Phil: gimp_gtk is a new package name, so it is not clear what is going to happen if we load this, as gimp relies upon so many other packages. Is it going to pull in the new gtk2 along with other things? Does it install parallel to the existing gimp 2.4.3 or will it uninstall that package and install this in its place? How about gimplibs, gimpprint. gimp_templates, gimp_help & gimp_extras, etc.? I would love to install and test this, but I don't have a separate test machine for this kind of thing and can't afford to mung up my office system. I suppose it's time I bite the bullet and learn how to set up zones so I can have a test environment for this kind of stuff. Is that a good use for zones, and if so, can someone recommend the best place to get started reading about them? Is that an online resource or a good book that I should purchase? BTW, if I create a new zone, can I login normally to my master zone and then from there, perform a 2nd login to the zone running in some sort of window and sharing the screen real estate, or do you need to select the target zone upon initial login? (See, like Sargent Schultz, I know nothing!) Thanks. Regards, -- Jeff C. Jeffery Small CJSA LLC 206-232-3338 jeff at cjsa.com 7000 E Mercer Way, Mercer Island, WA 98040 From phil at bolthole.com Sun Jun 13 06:31:40 2010 From: phil at bolthole.com (Philip Brown) Date: Sat, 12 Jun 2010 21:31:40 -0700 Subject: [csw-maintainers] new gimp! kinda... needs testing In-Reply-To: References: <20100612213822.GA88999@bolthole.com> Message-ID: if you have an x86 box it might be better to set up virtualbox instead. I'd imagine a graphically based prog will do better that way than a zone which would be "remote" display. ps: yea it would replace the existing old gtk with the new one from experimental, ideally. that's the point :) From james at opencsw.org Sun Jun 13 10:39:30 2010 From: james at opencsw.org (James Lee) Date: Sun, 13 Jun 2010 08:39:30 GMT Subject: [csw-maintainers] need help with C++ cmopile problem In-Reply-To: <20100612160518.GA63751@bolthole.com> References: <20100612160518.GA63751@bolthole.com> Message-ID: <20100613.8393000.3666072527@gyor.oxdrove.co.uk> On 12/06/10, 17:05:19, Philip Brown wrote regarding [csw-maintainers] need help with C++ cmopile problem: > Oddity that I have run into trying to compile a library. > simplified example below. > The following compiles with sun cc: > typedef struct fooo{ > int filler; > int testing []; > } foostruct; > it does NOT compile, however, with sun CC. > "test.c", line 4: Error: In this declaration "testing" is of an > incomplete type "int[]". > Any suggestions on what I can do? typedef struct fooo{ int filler; int *testing; } foostruct; James. From bwalton at opencsw.org Sun Jun 13 14:34:07 2010 From: bwalton at opencsw.org (Ben Walton) Date: Sun, 13 Jun 2010 08:34:07 -0400 Subject: [csw-maintainers] need help with C++ cmopile problem In-Reply-To: <20100613.8393000.3666072527@gyor.oxdrove.co.uk> References: <20100612160518.GA63751@bolthole.com> <20100613.8393000.3666072527@gyor.oxdrove.co.uk> Message-ID: <1276432023-sup-9754@pinkfloyd.chass.utoronto.ca> Excerpts from James Lee's message of Sun Jun 13 04:39:30 -0400 2010: Hi James, > typedef struct fooo{ > int filler; > int *testing; > } foostruct; This was my first thought too, but I hesitated to send it as I wasn't sure about it[1]. The C compiler would most likely turn 'int testing[]' into 'int *testing' I think since defining a struct of unknown (variable) size wouldn't be useful. But, 'int *' can be different than 'int []' as the array wouldn't require a de-reference to access to the contents. Will Phil need to look at how the structure is used to ensure it's treated as 'int *' even though originally declared as 'int []' ? Thanks -Ben [1] I don't write enough C (and no C++) code to be confident about it. -- Ben Walton Systems Programmer - CHASS University of Toronto C:416.407.5610 | W:416.978.4302 From bwalton at opencsw.org Sun Jun 13 15:56:48 2010 From: bwalton at opencsw.org (Ben Walton) Date: Sun, 13 Jun 2010 09:56:48 -0400 Subject: [csw-maintainers] need help with C++ cmopile problem In-Reply-To: <1276432023-sup-9754@pinkfloyd.chass.utoronto.ca> References: <20100612160518.GA63751@bolthole.com> <20100613.8393000.3666072527@gyor.oxdrove.co.uk> <1276432023-sup-9754@pinkfloyd.chass.utoronto.ca> Message-ID: <1276437383-sup-14@pinkfloyd.chass.utoronto.ca> Excerpts from Ben Walton's message of Sun Jun 13 08:34:07 -0400 2010: > to access to the contents. Will Phil need to look at how the > structure is used to ensure it's treated as 'int *' even though > originally declared as 'int []' ? Never mind. Answered my own question with google. Thanks -Ben -- Ben Walton Systems Programmer - CHASS University of Toronto C:416.407.5610 | W:416.978.4302 From phil at bolthole.com Sun Jun 13 16:16:49 2010 From: phil at bolthole.com (Philip Brown) Date: Sun, 13 Jun 2010 07:16:49 -0700 Subject: [csw-maintainers] need help with C++ cmopile problem In-Reply-To: <1276432023-sup-9754@pinkfloyd.chass.utoronto.ca> References: <20100612160518.GA63751@bolthole.com> <20100613.8393000.3666072527@gyor.oxdrove.co.uk> <1276432023-sup-9754@pinkfloyd.chass.utoronto.ca> Message-ID: odd thng is: -features=zla should fix it... but it doesn't From james at opencsw.org Sun Jun 13 16:38:58 2010 From: james at opencsw.org (James Lee) Date: Sun, 13 Jun 2010 14:38:58 GMT Subject: [csw-maintainers] need help with C++ cmopile problem In-Reply-To: <1276432023-sup-9754@pinkfloyd.chass.utoronto.ca> References: <20100612160518.GA63751@bolthole.com> <20100613.8393000.3666072527@gyor.oxdrove.co.uk> <1276432023-sup-9754@pinkfloyd.chass.utoronto.ca> Message-ID: <20100613.14385800.2199354511@gyor.oxdrove.co.uk> On 13/06/10, 13:34:07, Ben Walton wrote regarding Re: [csw-maintainers] need help with C++ cmopile problem: > > typedef struct fooo{ > > int filler; > > int *testing; > > } foostruct; > This was my first thought too, but I hesitated to send it as I wasn't > sure about it[1]. The C compiler would most likely turn 'int > testing[]' into 'int *testing' I think since defining a struct of > unknown (variable) size wouldn't be useful. But, 'int *' can be > different than 'int []' as the array wouldn't require a de-reference > to access to the contents. Will Phil need to look at how the > structure is used to ensure it's treated as 'int *' even though > originally declared as 'int []' ? It does handle a pointer but it difference where the pointer is stored. So yes your are right, it does depend how it's used, as written the the structure is no use and can only be used to point to another piece of memory. "foostruct foo" isn't useful; "foostruct foo *" is how to use. The way to get CC to accept the array is to allocate 1 unit of storage: typedef struct foo { int length; int array[1]; } foostruct; This will work and just wastes 4 bytes. "int array[0];" is a variation I've often seen, which is accepted by gcc but not cc. #include #include typedef struct { int length; int array[1]; } foo_t; typedef struct { int length; int *array; } bar_t; void load(int length, int* data) { for (int i = 0; i < length; i++) { data[i] = i; } } void dump(int length, int* data) { for (int i = 0; i < length; i++) { printf(" %i", data[i]); } printf("\n"); } int main(int argc, char* argv[]) { int size = 9; // variable foo_t * foo; bar_t bar; printf("%i\n", sizeof (foo_t)); foo = (foo_t *) malloc(sizeof(foo_t) + size * sizeof (int)); foo->length = size; for (int i = 0; i < size; i++) { foo->array[i] = i; } load(foo->length, foo->array); dump(foo->length, foo->array); printf("%i\n", sizeof (bar_t)); bar.length = size; #ifdef __cplusplus bar.array = new int[size]; #else bar.array = malloc(size * sizeof (int)); #endif for (int i = 0; i < size; i++) { bar.array[i] = i; } load(bar.length, bar.array); dump(bar.length, bar.array); } I expect adding a 1 unit dimension is the easiest modification to existing software but if I was writing from the start I'd use pointers. Consider: typedef struct { int length; int height; int X[]; // won't compile int Y[]; } foo_t; James. From william at wbonnet.net Sun Jun 13 17:12:24 2010 From: william at wbonnet.net (William Bonnet) Date: Sun, 13 Jun 2010 17:12:24 +0200 Subject: [csw-maintainers] New website is online Message-ID: <4C14F558.9040003@wbonnet.net> Hi all, I'm pleased to announce that the new website is online. Feedbacks, improvements and bugs report are welcome :) cheers W. -- William http://www.wbonnet.net http://www.sunwizard.net Le site fran?ais des amateurs de stations Unix http://www.opencsw.org Community SoftWare for Solaris http://www.guses.org French speaking Solaris User Group From phil at bolthole.com Sun Jun 13 21:37:06 2010 From: phil at bolthole.com (Philip Brown) Date: Sun, 13 Jun 2010 12:37:06 -0700 Subject: [csw-maintainers] need help with C++ cmopile problem In-Reply-To: <20100613.14385800.2199354511@gyor.oxdrove.co.uk> References: <20100612160518.GA63751@bolthole.com> <20100613.8393000.3666072527@gyor.oxdrove.co.uk> <1276432023-sup-9754@pinkfloyd.chass.utoronto.ca> <20100613.14385800.2199354511@gyor.oxdrove.co.uk> Message-ID: turns out, sun CC now accepts [0]. (but still not [] ) From ihsan at opencsw.org Sun Jun 13 22:36:17 2010 From: ihsan at opencsw.org (Ihsan Dogan) Date: Sun, 13 Jun 2010 22:36:17 +0200 Subject: [csw-maintainers] New website is online In-Reply-To: <4C14F558.9040003@wbonnet.net> References: <4C14F558.9040003@wbonnet.net> Message-ID: <4C154141.7070504@opencsw.org> Am 13.06.10 17:12, schrieb William Bonnet: > I'm pleased to announce that the new website is online. > > Feedbacks, improvements and bugs report are welcome :) Since today, the Mantis is now accessible through SSL. http://www.opencsw.org/mantis is forwarded to https://www.opencsw.org/mantis. Ihsan -- ihsan at dogan.ch http://blog.dogan.ch/ From a.cervellin at acm.org Sun Jun 13 22:42:00 2010 From: a.cervellin at acm.org (Alessio) Date: Sun, 13 Jun 2010 22:42:00 +0200 Subject: [csw-maintainers] New website is online In-Reply-To: <4C14F558.9040003@wbonnet.net> References: <4C14F558.9040003@wbonnet.net> Message-ID: <4C154298.1020306@acm.org> William Bonnet wrote: > Hi all, > > I'm pleased to announce that the new website is online. > > Feedbacks, improvements and bugs report are welcome :) great job william! there's just a wrong tag at the top of this page: http://www.opencsw.org/use-it/ From ihsan at opencsw.org Sun Jun 13 22:50:56 2010 From: ihsan at opencsw.org (Ihsan Dogan) Date: Sun, 13 Jun 2010 22:50:56 +0200 Subject: [csw-maintainers] New website is online In-Reply-To: <4C154298.1020306@acm.org> References: <4C14F558.9040003@wbonnet.net> <4C154298.1020306@acm.org> Message-ID: <4C1544B0.2000600@opencsw.org> Am 13.06.10 22:42, schrieb Alessio: >> Feedbacks, improvements and bugs report are welcome :) > > great job william! > there's just a wrong tag at the top of this page: > http://www.opencsw.org/use-it/ Also at: http://www.opencsw.org/about/faq/ http://www.opencsw.org/get-it/mirrors/ Ihsan -- ihsan at dogan.ch http://blog.dogan.ch/ From maciej at opencsw.org Mon Jun 14 00:10:33 2010 From: maciej at opencsw.org (Maciej (Matchek) Blizinski) Date: Sun, 13 Jun 2010 23:10:33 +0100 Subject: [csw-maintainers] New website is online In-Reply-To: <4C154141.7070504@opencsw.org> References: <4C14F558.9040003@wbonnet.net> <4C154141.7070504@opencsw.org> Message-ID: No dia 13 de Junho de 2010 21:36, Ihsan Dogan escreveu: > Am 13.06.10 17:12, schrieb William Bonnet: > >> I'm pleased to announce that the new website is online. >> >> Feedbacks, improvements and bugs report are welcome :) > > Since today, the Mantis is now accessible through SSL. > http://www.opencsw.org/mantis is forwarded to > https://www.opencsw.org/mantis. http://www.opencsw.org/bugtrack led to a page which didn't have a link to mantis. I've added a link there. From william at wbonnet.net Mon Jun 14 00:35:16 2010 From: william at wbonnet.net (William Bonnet) Date: Mon, 14 Jun 2010 00:35:16 +0200 Subject: [csw-maintainers] New website is online In-Reply-To: <4C154298.1020306@acm.org> References: <4C14F558.9040003@wbonnet.net> <4C154298.1020306@acm.org> Message-ID: <4C155D24.4030601@wbonnet.net> Hi > great job william! > there's just a wrong tag at the top of this page: > http://www.opencsw.org/use-it/ Thanks a lot it's fixed cheers W. -- William http://www.wbonnet.net http://www.sunwizard.net Le site fran?ais des amateurs de stations Unix http://www.opencsw.org Community SoftWare for Solaris http://www.guses.org French speaking Solaris User Group From william at wbonnet.net Mon Jun 14 00:36:38 2010 From: william at wbonnet.net (William Bonnet) Date: Mon, 14 Jun 2010 00:36:38 +0200 Subject: [csw-maintainers] New website is online In-Reply-To: <4C1544B0.2000600@opencsw.org> References: <4C14F558.9040003@wbonnet.net> <4C154298.1020306@acm.org> <4C1544B0.2000600@opencsw.org> Message-ID: <4C155D76.8020607@wbonnet.net> Ihsan Dogan a ?crit : > Am 13.06.10 22:42, schrieb Alessio: > > >>> Feedbacks, improvements and bugs report are welcome :) >>> >> great job william! >> there's just a wrong tag at the top of this page: >> http://www.opencsw.org/use-it/ >> > > Also at: > http://www.opencsw.org/about/faq/ > http://www.opencsw.org/get-it/mirrors/ > > > > > Ihsan > > Fixed, thanks -- William http://www.wbonnet.net http://www.sunwizard.net Le site fran?ais des amateurs de stations Unix http://www.opencsw.org Community SoftWare for Solaris http://www.guses.org French speaking Solaris User Group From bwalton at opencsw.org Mon Jun 14 00:52:43 2010 From: bwalton at opencsw.org (Ben Walton) Date: Sun, 13 Jun 2010 18:52:43 -0400 Subject: [csw-maintainers] need help with C++ cmopile problem In-Reply-To: <20100613.14385800.2199354511@gyor.oxdrove.co.uk> References: <20100612160518.GA63751@bolthole.com> <20100613.8393000.3666072527@gyor.oxdrove.co.uk> <1276432023-sup-9754@pinkfloyd.chass.utoronto.ca> <20100613.14385800.2199354511@gyor.oxdrove.co.uk> Message-ID: <1276469299-sup-9805@pinkfloyd.chass.utoronto.ca> Excerpts from James Lee's message of Sun Jun 13 10:38:58 -0400 2010: Hi James, Thanks for the excellent response and example. Thanks -Ben -- Ben Walton Systems Programmer - CHASS University of Toronto C:416.407.5610 | W:416.978.4302 From bonivart at opencsw.org Mon Jun 14 10:25:46 2010 From: bonivart at opencsw.org (Peter Bonivart) Date: Mon, 14 Jun 2010 10:25:46 +0200 Subject: [csw-maintainers] New website is online In-Reply-To: <4C14F558.9040003@wbonnet.net> References: <4C14F558.9040003@wbonnet.net> Message-ID: On Sun, Jun 13, 2010 at 5:12 PM, William Bonnet wrote: > Hi all, > > I'm pleased to announce that the new website is online. > > Feedbacks, improvements and bugs report are welcome :) Great! :-) -- /peter From dam at opencsw.org Mon Jun 14 10:49:33 2010 From: dam at opencsw.org (Dagobert Michelsen) Date: Mon, 14 Jun 2010 10:49:33 +0200 Subject: [csw-maintainers] need help with C++ cmopile problem In-Reply-To: <1276432023-sup-9754@pinkfloyd.chass.utoronto.ca> References: <20100612160518.GA63751@bolthole.com> <20100613.8393000.3666072527@gyor.oxdrove.co.uk> <1276432023-sup-9754@pinkfloyd.chass.utoronto.ca> Message-ID: Hi Y'all, Am 13.06.2010 um 14:34 schrieb Ben Walton: > Excerpts from James Lee's message of Sun Jun 13 04:39:30 -0400 2010: > > Hi James, > >> typedef struct fooo{ >> int filler; >> int *testing; >> } foostruct; > > This was my first thought too, but I hesitated to send it as I wasn't > sure about it[1]. The C compiler would most likely turn 'int > testing[]' into 'int *testing' I think since defining a struct of > unknown (variable) size wouldn't be useful. But, 'int *' can be > different than 'int []' as the array wouldn't require a de-reference > to access to the contents. Will Phil need to look at how the > structure is used to ensure it's treated as 'int *' even though > originally declared as 'int []' ? > > Thanks > -Ben > > [1] I don't write enough C (and no C++) code to be confident about it. Someone wants to add this to the porting faq? Best regards -- Dago From maciej at opencsw.org Mon Jun 14 17:16:24 2010 From: maciej at opencsw.org (Maciej (Matchek) Blizinski) Date: Mon, 14 Jun 2010 16:16:24 +0100 Subject: [csw-maintainers] Important: X11 related packages, and buildfarm In-Reply-To: <20100528132042.GA34110@bolthole.com> References: <20100528132042.GA34110@bolthole.com> Message-ID: No dia 28 de Maio de 2010 14:20, Philip Brown escreveu: > We just started rerolling things like xpm, and gtk, to use the new libs. > Unfortunately, this turned out to not be the best path, for two reasons. > One of which was that we lose some hardware acceleration, by not using Sun > X11 libs. Another was that there started to be cross-version pollution of > "libX11.so". ?Some higher level applications ended up pulling in both our > version, and the sun version. This is Not Good. I was thinking about ways of detecting this on the checkpkg stage. It can't be done by examining the output of /usr/ccs/dump; it's usually one of the dependencies that pulls in the other version of libX11. This check won't be easy to write, but I'm sure there are other checks that we can write that would detect typical problems and provide helpful messages, with links to documentation. Could you identify and describe typical common problematic scenarios that could be tested for by examining just the files in the package? (One potential example would be linking against Sun libX11 and libraries in CSWlibxext at the same time.) Maciej From phil at bolthole.com Mon Jun 14 18:42:14 2010 From: phil at bolthole.com (Philip Brown) Date: Mon, 14 Jun 2010 09:42:14 -0700 Subject: [csw-maintainers] Important: X11 related packages, and buildfarm In-Reply-To: References: <20100528132042.GA34110@bolthole.com> Message-ID: hi Maciej, a shortened answer: after I release the updated chain officially... pretty much nothing should be using /opt/csw/X11 any more. if that is found on the runpath it could be considered a bug From william at wbonnet.net Tue Jun 15 00:55:27 2010 From: william at wbonnet.net (William Bonnet) Date: Tue, 15 Jun 2010 00:55:27 +0200 Subject: [csw-maintainers] New website is online In-Reply-To: References: <4C14F558.9040003@wbonnet.net> <4C154141.7070504@opencsw.org> Message-ID: <4C16B35F.3010800@wbonnet.net> Hi all I have added two icons on the right side bar which directly links to package list and mantis Emails and rss feed are not yet working on this server cheers W. Maciej (Matchek) Blizinski a ?crit : > No dia 13 de Junho de 2010 21:36, Ihsan Dogan escreveu: > >> Am 13.06.10 17:12, schrieb William Bonnet: >> >> >>> I'm pleased to announce that the new website is online. >>> >>> Feedbacks, improvements and bugs report are welcome :) >>> >> Since today, the Mantis is now accessible through SSL. >> http://www.opencsw.org/mantis is forwarded to >> https://www.opencsw.org/mantis. >> > > http://www.opencsw.org/bugtrack led to a page which didn't have a link > to mantis. I've added a link there. > _______________________________________________ > maintainers mailing list > maintainers at lists.opencsw.org > https://lists.opencsw.org/mailman/listinfo/maintainers > .:: This mailing list's archive is public. ::. > -- William http://www.wbonnet.net http://www.sunwizard.net Le site fran?ais des amateurs de stations Unix http://www.opencsw.org Community SoftWare for Solaris http://www.guses.org French speaking Solaris User Group From bwalton at opencsw.org Tue Jun 15 02:38:49 2010 From: bwalton at opencsw.org (Ben Walton) Date: Mon, 14 Jun 2010 20:38:49 -0400 Subject: [csw-maintainers] New website is online In-Reply-To: <4C16B35F.3010800@wbonnet.net> References: <4C14F558.9040003@wbonnet.net> <4C154141.7070504@opencsw.org> <4C16B35F.3010800@wbonnet.net> Message-ID: <1276562266-sup-8646@pinkfloyd.chass.utoronto.ca> Excerpts from William Bonnet's message of Mon Jun 14 18:55:27 -0400 2010: Hi William, > I have added two icons on the right side bar which directly links to > package list and mantis The new site looks great! Thanks to you and everyone else that spent cycles on this task. Thanks -Ben -- Ben Walton Systems Programmer - CHASS University of Toronto C:416.407.5610 | W:416.978.4302 From Darin.Perusich at cognigencorp.com Tue Jun 15 15:36:44 2010 From: Darin.Perusich at cognigencorp.com (Darin Perusich) Date: Tue, 15 Jun 2010 09:36:44 -0400 Subject: [csw-maintainers] i.cswinet install script failing In-Reply-To: <1276302909-sup-4068@pinkfloyd.chass.utoronto.ca> References: <4C127792.6000608@cognigencorp.com> <1276278926-sup-163@pinkfloyd.chass.utoronto.ca> <4C127A0E.4000005@cognigencorp.com> <1276288316-sup-7855@pinkfloyd.chass.utoronto.ca> <1276302909-sup-4068@pinkfloyd.chass.utoronto.ca> Message-ID: <4C1781EC.1010603@cognigencorp.com> Hi Ben, I just installed the updated cswclassutils from your repo and get the following error. Installing class ... /opt/csw/etc/pkg/CSWamanda/inetd.conf /usr/sadm/install/scripts/i.cswinetd: test: unknown operator == pkgadd: ERROR: class action script did not complete successfully On 06/11/2010 08:37 PM, Ben Walton wrote: > Excerpts from Ben Walton's message of Fri Jun 11 16:32:29 -0400 2010: > > Hi Darin, > > Found the bug. When I switched from using grep to awk to detect lines > in /etc/services, the 'filter' got less specific. It was then > triggering on more than one line (where applicable). This was seeing > the test for the detection fail. > > I've re-tightened the criteria for this check and it seems to work > properly now. There is an update in my catalog, so please test it. > If it's good, I'll bump Peter to release a new version. > > http://mirror.opencsw.org/opencsw/experimental/bwalton/ > > Thanks. > -Ben > > -- > Ben Walton > Systems Programmer - CHASS > University of Toronto > C:416.407.5610 | W:416.978.4302 > > _______________________________________________ > maintainers mailing list > maintainers at lists.opencsw.org > https://lists.opencsw.org/mailman/listinfo/maintainers > .:: This mailing list's archive is public. ::. -- Darin Perusich Unix Systems Administrator Cognigen Corporation 395 Youngs Rd. Williamsville, NY 14221 Phone: 716-633-3463 Email: darinper at cognigencorp.com From phil at bolthole.com Tue Jun 15 15:37:58 2010 From: phil at bolthole.com (Philip Brown) Date: Tue, 15 Jun 2010 06:37:58 -0700 Subject: [csw-maintainers] website issues Message-ID: there's no "webmaster" contact in the support area, and there is a Forums subtab that goes nowhere From phil at bolthole.com Tue Jun 15 17:13:45 2010 From: phil at bolthole.com (Philip Brown) Date: Tue, 15 Jun 2010 08:13:45 -0700 Subject: [csw-maintainers] i.cswinet install script failing In-Reply-To: <4C1781EC.1010603@cognigencorp.com> References: <4C127792.6000608@cognigencorp.com> <1276278926-sup-163@pinkfloyd.chass.utoronto.ca> <4C127A0E.4000005@cognigencorp.com> <1276288316-sup-7855@pinkfloyd.chass.utoronto.ca> <1276302909-sup-4068@pinkfloyd.chass.utoronto.ca> <4C1781EC.1010603@cognigencorp.com> Message-ID: aaarg. x2 1. I'm guessing this is a bashism Get over bash, Ben :-) 2. we can't have classutils breaking like this. it's a critical, supposedly trusted package. I'm thinking we need some kind of validation test suite for it. any volunteers to write one? On Tuesday, June 15, 2010, Darin Perusich wrote: > Hi Ben, > > I just installed the updated cswclassutils from your repo and get the > following error. > > Installing class ... > /opt/csw/etc/pkg/CSWamanda/inetd.conf > /usr/sadm/install/scripts/i.cswinetd: test: unknown operator == > pkgadd: ERROR: class action script did not complete successfully > > > On 06/11/2010 08:37 PM, Ben Walton wrote: >> Excerpts from Ben Walton's message of Fri Jun 11 16:32:29 -0400 2010: >> >> Hi Darin, >> >> Found the bug. ?When I switched from using grep to awk to detect lines >> in /etc/services, the 'filter' got less specific. ?It was then >> triggering on more than one line (where applicable). ?This was seeing >> the test for the detection fail. >> >> I've re-tightened the criteria for this check and it seems to work >> properly now. ?There is an update in my catalog, so please test it. >> If it's good, I'll bump Peter to release a new version. >> >> http://mirror.opencsw.org/opencsw/experimental/bwalton/ >> >> Thanks. >> -Ben >> >> -- >> Ben Walton >> Systems Programmer - CHASS >> University of Toronto >> C:416.407.5610 | W:416.978.4302 >> >> _______________________________________________ >> maintainers mailing list >> maintainers at lists.opencsw.org >> https://lists.opencsw.org/mailman/listinfo/maintainers >> .:: This mailing list's archive is public. ::. > > -- > Darin Perusich > Unix Systems Administrator > Cognigen Corporation > 395 Youngs Rd. > Williamsville, NY 14221 > Phone: 716-633-3463 > Email: darinper at cognigencorp.com > _______________________________________________ > maintainers mailing list > maintainers at lists.opencsw.org > https://lists.opencsw.org/mailman/listinfo/maintainers > .:: This mailing list's archive is public. ::. > From maciej at opencsw.org Tue Jun 15 17:29:16 2010 From: maciej at opencsw.org (Maciej (Matchek) Blizinski) Date: Tue, 15 Jun 2010 16:29:16 +0100 Subject: [csw-maintainers] i.cswinet install script failing In-Reply-To: References: <4C127792.6000608@cognigencorp.com> <1276278926-sup-163@pinkfloyd.chass.utoronto.ca> <4C127A0E.4000005@cognigencorp.com> <1276288316-sup-7855@pinkfloyd.chass.utoronto.ca> <1276302909-sup-4068@pinkfloyd.chass.utoronto.ca> <4C1781EC.1010603@cognigencorp.com> Message-ID: No dia 15 de Junho de 2010 16:13, Philip Brown escreveu: > aaarg. x2 > > 1. I'm guessing this is a bashism > Get over bash, Ben :-) I've read the script and didn't see anything glaringly bashious. I'm curious what it is. https://gar.svn.sourceforge.net/svnroot/gar/csw/mgar/pkg/cswclassutils/trunk/files/CSWcswclassutils.i.cswinetd > 2. we can't have classutils breaking like this. it's a critical, > supposedly trusted package. > ?I'm thinking we need some kind of validation test suite for it. any > volunteers to write one? I like the idea! We have a package with shutil2, it might be helpful, as it's a shell unit testing framework. The hard part might be providing mocks of commands being called. I could look at it, but no promises, I already have enough work with checkpkg. Maciej From bonivart at opencsw.org Tue Jun 15 17:32:20 2010 From: bonivart at opencsw.org (Peter Bonivart) Date: Tue, 15 Jun 2010 17:32:20 +0200 Subject: [csw-maintainers] i.cswinet install script failing In-Reply-To: <4C1781EC.1010603@cognigencorp.com> References: <4C127792.6000608@cognigencorp.com> <1276278926-sup-163@pinkfloyd.chass.utoronto.ca> <4C127A0E.4000005@cognigencorp.com> <1276288316-sup-7855@pinkfloyd.chass.utoronto.ca> <1276302909-sup-4068@pinkfloyd.chass.utoronto.ca> <4C1781EC.1010603@cognigencorp.com> Message-ID: On Tue, Jun 15, 2010 at 3:36 PM, Darin Perusich wrote: > Hi Ben, > > I just installed the updated cswclassutils from your repo and get the > following error. > > Installing class ... > /opt/csw/etc/pkg/CSWamanda/inetd.conf > /usr/sadm/install/scripts/i.cswinetd: test: unknown operator == > pkgadd: ERROR: class action script did not complete successfully This sounds like an old bug. If I look at Ben's experimental folder the version there is 1.35, we have now released 1.37. What version do you have installed? I can produce another test package to see if we can sort this out. -- /peter From maciej at opencsw.org Tue Jun 15 19:06:33 2010 From: maciej at opencsw.org (Maciej (Matchek) Blizinski) Date: Tue, 15 Jun 2010 18:06:33 +0100 Subject: [csw-maintainers] website issues In-Reply-To: References: Message-ID: 3. No links to the search subpage. From bwalton at opencsw.org Tue Jun 15 20:08:56 2010 From: bwalton at opencsw.org (Ben Walton) Date: Tue, 15 Jun 2010 14:08:56 -0400 Subject: [csw-maintainers] i.cswinet install script failing In-Reply-To: <4C1781EC.1010603@cognigencorp.com> References: <4C127792.6000608@cognigencorp.com> <1276278926-sup-163@pinkfloyd.chass.utoronto.ca> <4C127A0E.4000005@cognigencorp.com> <1276288316-sup-7855@pinkfloyd.chass.utoronto.ca> <1276302909-sup-4068@pinkfloyd.chass.utoronto.ca> <4C1781EC.1010603@cognigencorp.com> Message-ID: <1276625196-sup-5001@pinkfloyd.chass.utoronto.ca> Excerpts from Darin Perusich's message of Tue Jun 15 09:36:44 -0400 2010: > Installing class ... > /opt/csw/etc/pkg/CSWamanda/inetd.conf > /usr/sadm/install/scripts/i.cswinetd: test: unknown operator == > pkgadd: ERROR: class action script did not complete successfully I did slip this in at one point[1], but that was before the release in current. Looking at the current files that went into the package I put in experiemental, I see: --snip-- bwalton @ current9x : ~/packages/cswclassutils/trunk/files $ grep '==' * CSWcswclassutils.README.CSW:================ CSWcswclassutils.i.cswetcservices: exists=`awk "\\\$1 == \"$svcname\" && \\\$2 == \"$port_proto\" { print \"found\" }" $svcfile` CSWcswclassutils.i.cswinetd: exists=`awk "\\\$1 == \"$svcname\" && \\\$2 ~ /$proto/ {print \"found\"; exit }" /etc/services` CSWcswclassutils.i.cswinitsmf: set -- `/usr/bin/awk " \\\$1 == \"$1\" { print \\\$2 } " "$SMF_STATE_FILE"` CSWcswclassutils.i.cswusergroup: awk 'BEGIN { FS=":"; OFS=":" } $1 == "'$user'" { $2 = "NP" } { print }' /etc/shadow > /etc/shadow.$PKGINST --snip-- That looks good to me...pkgparam CSWcswclassutils ? Thanks -Ben [1] It was released accidentally in this case. -- Ben Walton Systems Programmer - CHASS University of Toronto C:416.407.5610 | W:416.978.4302 From bwalton at opencsw.org Tue Jun 15 20:09:58 2010 From: bwalton at opencsw.org (Ben Walton) Date: Tue, 15 Jun 2010 14:09:58 -0400 Subject: [csw-maintainers] i.cswinet install script failing In-Reply-To: References: <4C127792.6000608@cognigencorp.com> <1276278926-sup-163@pinkfloyd.chass.utoronto.ca> <4C127A0E.4000005@cognigencorp.com> <1276288316-sup-7855@pinkfloyd.chass.utoronto.ca> <1276302909-sup-4068@pinkfloyd.chass.utoronto.ca> <4C1781EC.1010603@cognigencorp.com> Message-ID: <1276625340-sup-8373@pinkfloyd.chass.utoronto.ca> Excerpts from Philip Brown's message of Tue Jun 15 11:13:45 -0400 2010: > 1. I'm guessing this is a bashism > Get over bash, Ben :-) Actually, my == habit is from C, not from bash. Get over yourself Phil! :) Thanks -Ben -- Ben Walton Systems Programmer - CHASS University of Toronto C:416.407.5610 | W:416.978.4302 From bwalton at opencsw.org Tue Jun 15 20:13:05 2010 From: bwalton at opencsw.org (Ben Walton) Date: Tue, 15 Jun 2010 14:13:05 -0400 Subject: [csw-maintainers] i.cswinet install script failing In-Reply-To: References: <4C127792.6000608@cognigencorp.com> <1276278926-sup-163@pinkfloyd.chass.utoronto.ca> <4C127A0E.4000005@cognigencorp.com> <1276288316-sup-7855@pinkfloyd.chass.utoronto.ca> <1276302909-sup-4068@pinkfloyd.chass.utoronto.ca> <4C1781EC.1010603@cognigencorp.com> Message-ID: <1276625516-sup-2233@pinkfloyd.chass.utoronto.ca> Excerpts from Peter Bonivart's message of Tue Jun 15 11:32:20 -0400 2010: > This sounds like an old bug. If I look at Ben's experimental folder > the version there is 1.35, we have now released 1.37. What version > do you have installed? That's odd. The one I placed the other night was 1.37 (I intentionally left the version as is until Peter ups it for release): --snip-- # pkgparam CSWcswclassutils none / Canada/Eastern /sbin:/usr/sbin:/usr/bin:/usr/sadm/install/bin /usr/sadm/sysadm CSWcswclassutils cswclassutils - CSW class action utilities all 1.37,REV=2010.06.12 application http://www.opencsw.org packaged for CSW by Ben Walton bwalton at opencsw.org bwalton at current9x-20100612023350 http://www.opencsw.org/bugtrack/ cswclassutils 32 https://gar.svn.sourceforge.net/svnroot/gar/csw/mgar/pkg/cswclassutils/trunk at 10201 ../build-isa-i386 CSWcswclassutils /var/sadm/pkg/CSWcswclassutils/save Jun 11 2010 21:04 --snip-- Thanks -Ben -- Ben Walton Systems Programmer - CHASS University of Toronto C:416.407.5610 | W:416.978.4302 From Darin.Perusich at cognigencorp.com Tue Jun 15 20:30:39 2010 From: Darin.Perusich at cognigencorp.com (Darin Perusich) Date: Tue, 15 Jun 2010 14:30:39 -0400 Subject: [csw-maintainers] i.cswinet install script failing In-Reply-To: <1276625196-sup-5001@pinkfloyd.chass.utoronto.ca> References: <4C127792.6000608@cognigencorp.com> <1276278926-sup-163@pinkfloyd.chass.utoronto.ca> <4C127A0E.4000005@cognigencorp.com> <1276288316-sup-7855@pinkfloyd.chass.utoronto.ca> <1276302909-sup-4068@pinkfloyd.chass.utoronto.ca> <4C1781EC.1010603@cognigencorp.com> <1276625196-sup-5001@pinkfloyd.chass.utoronto.ca> Message-ID: <4C17C6CF.50906@cognigencorp.com> On 06/15/2010 02:08 PM, Ben Walton wrote: > Excerpts from Darin Perusich's message of Tue Jun 15 09:36:44 -0400 2010: > >> Installing class ... >> /opt/csw/etc/pkg/CSWamanda/inetd.conf >> /usr/sadm/install/scripts/i.cswinetd: test: unknown operator == >> pkgadd: ERROR: class action script did not complete successfully > > I did slip this in at one point[1], but that was before the release in > current. Looking at the current files that went into the package I > put in experiemental, I see: > > --snip-- > bwalton @ current9x : ~/packages/cswclassutils/trunk/files > $ grep '==' * > CSWcswclassutils.README.CSW:================ > CSWcswclassutils.i.cswetcservices: exists=`awk "\\\$1 == > \"$svcname\" && \\\$2 == \"$port_proto\" { print \"found\" }" > $svcfile` > CSWcswclassutils.i.cswinetd: exists=`awk "\\\$1 == \"$svcname\" && > \\\$2 ~ /$proto/ {print \"found\"; exit }" /etc/services` > CSWcswclassutils.i.cswinitsmf: set -- `/usr/bin/awk " \\\$1 == > \"$1\" { print \\\$2 } " "$SMF_STATE_FILE"` > CSWcswclassutils.i.cswusergroup: awk 'BEGIN { FS=":"; OFS=":" } $1 > == "'$user'" { $2 = "NP" } { print }' /etc/shadow > > /etc/shadow.$PKGINST > > --snip-- > > That looks good to me...pkgparam CSWcswclassutils ? > The version I grabbed from http://mirror.opencsw.org/opencsw/experimental/bwalton/ is cswclassutils-1.35,REV=2010.05.06. -- Darin Perusich Unix Systems Administrator Cognigen Corporation 395 Youngs Rd. Williamsville, NY 14221 Phone: 716-633-3463 Email: darinper at cognigencorp.com From bwalton at opencsw.org Tue Jun 15 20:40:58 2010 From: bwalton at opencsw.org (Ben Walton) Date: Tue, 15 Jun 2010 14:40:58 -0400 Subject: [csw-maintainers] i.cswinet install script failing In-Reply-To: <4C17C6CF.50906@cognigencorp.com> References: <4C127792.6000608@cognigencorp.com> <1276278926-sup-163@pinkfloyd.chass.utoronto.ca> <4C127A0E.4000005@cognigencorp.com> <1276288316-sup-7855@pinkfloyd.chass.utoronto.ca> <1276302909-sup-4068@pinkfloyd.chass.utoronto.ca> <4C1781EC.1010603@cognigencorp.com> <1276625196-sup-5001@pinkfloyd.chass.utoronto.ca> <4C17C6CF.50906@cognigencorp.com> Message-ID: <1276626751-sup-6874@pinkfloyd.chass.utoronto.ca> Excerpts from Darin Perusich's message of Tue Jun 15 14:30:39 -0400 2010: > The version I grabbed from > http://mirror.opencsw.org/opencsw/experimental/bwalton/ is > cswclassutils-1.35,REV=2010.05.06. Which is what I see there now too. I'm not sure what happened. Anyway, I've just placed 1.37,REV=2010.06.15 in in /home/experimental/bwalton on the buildfarm. Try this one. If it's good, let us know and Peter can kick 1.38 out. Thanks -Ben -- Ben Walton Systems Programmer - CHASS University of Toronto C:416.407.5610 | W:416.978.4302 From Darin.Perusich at cognigencorp.com Tue Jun 15 21:04:19 2010 From: Darin.Perusich at cognigencorp.com (Darin Perusich) Date: Tue, 15 Jun 2010 15:04:19 -0400 Subject: [csw-maintainers] i.cswinet install script failing In-Reply-To: <1276626751-sup-6874@pinkfloyd.chass.utoronto.ca> References: <4C127792.6000608@cognigencorp.com> <1276278926-sup-163@pinkfloyd.chass.utoronto.ca> <4C127A0E.4000005@cognigencorp.com> <1276288316-sup-7855@pinkfloyd.chass.utoronto.ca> <1276302909-sup-4068@pinkfloyd.chass.utoronto.ca> <4C1781EC.1010603@cognigencorp.com> <1276625196-sup-5001@pinkfloyd.chass.utoronto.ca> <4C17C6CF.50906@cognigencorp.com> <1276626751-sup-6874@pinkfloyd.chass.utoronto.ca> Message-ID: <4C17CEB3.5000402@cognigencorp.com> Looks good Ben, thanks! On 06/15/2010 02:40 PM, Ben Walton wrote: > Excerpts from Darin Perusich's message of Tue Jun 15 14:30:39 -0400 2010: > >> The version I grabbed from >> http://mirror.opencsw.org/opencsw/experimental/bwalton/ is >> cswclassutils-1.35,REV=2010.05.06. > > Which is what I see there now too. I'm not sure what happened. > Anyway, I've just placed 1.37,REV=2010.06.15 in in > /home/experimental/bwalton on the buildfarm. Try this one. > > If it's good, let us know and Peter can kick 1.38 out. > > Thanks > -Ben > > -- > Ben Walton > Systems Programmer - CHASS > University of Toronto > C:416.407.5610 | W:416.978.4302 > > _______________________________________________ > maintainers mailing list > maintainers at lists.opencsw.org > https://lists.opencsw.org/mailman/listinfo/maintainers > .:: This mailing list's archive is public. ::. -- Darin Perusich Unix Systems Administrator Cognigen Corporation 395 Youngs Rd. Williamsville, NY 14221 Phone: 716-633-3463 Email: darinper at cognigencorp.com From bonivart at opencsw.org Tue Jun 15 23:05:26 2010 From: bonivart at opencsw.org (Peter Bonivart) Date: Tue, 15 Jun 2010 23:05:26 +0200 Subject: [csw-maintainers] i.cswinet install script failing In-Reply-To: <4C17CEB3.5000402@cognigencorp.com> References: <4C127792.6000608@cognigencorp.com> <1276278926-sup-163@pinkfloyd.chass.utoronto.ca> <4C127A0E.4000005@cognigencorp.com> <1276288316-sup-7855@pinkfloyd.chass.utoronto.ca> <1276302909-sup-4068@pinkfloyd.chass.utoronto.ca> <4C1781EC.1010603@cognigencorp.com> <1276625196-sup-5001@pinkfloyd.chass.utoronto.ca> <4C17C6CF.50906@cognigencorp.com> <1276626751-sup-6874@pinkfloyd.chass.utoronto.ca> <4C17CEB3.5000402@cognigencorp.com> Message-ID: On Tue, Jun 15, 2010 at 9:04 PM, Darin Perusich wrote: > Looks good Ben, thanks! > > On 06/15/2010 02:40 PM, Ben Walton wrote: >> Excerpts from Darin Perusich's message of Tue Jun 15 14:30:39 -0400 2010: >> >>> The version I grabbed from >>> http://mirror.opencsw.org/opencsw/experimental/bwalton/ is >>> cswclassutils-1.35,REV=2010.05.06. >> >> Which is what I see there now too. ?I'm not sure what happened. >> Anyway, I've just placed 1.37,REV=2010.06.15 in in >> /home/experimental/bwalton on the buildfarm. ?Try this one. >> >> If it's good, let us know and Peter can kick 1.38 out. I have put 1.38 in my experimental folder. Please verify that it's correct. http://mirror.opencsw.org/experimental.html#bonivart -- /peter From william at wbonnet.net Wed Jun 16 00:47:55 2010 From: william at wbonnet.net (William Bonnet) Date: Wed, 16 Jun 2010 00:47:55 +0200 Subject: [csw-maintainers] New website is online In-Reply-To: <1276562266-sup-8646@pinkfloyd.chass.utoronto.ca> References: <4C14F558.9040003@wbonnet.net> <4C154141.7070504@opencsw.org> <4C16B35F.3010800@wbonnet.net> <1276562266-sup-8646@pinkfloyd.chass.utoronto.ca> Message-ID: <4C18031B.4050503@wbonnet.net> Hi Ben > Thanks to you and everyone else that spent cycles on this task. > Thanks to you. I had the pleasure to announce the new website, but i'm far from being the only one which worked on this web site ! In fact, i only worked to produce php code in order to generate new bugs ;) and it tooks me a loooonnnngggg time to reach this goal :D Several members were involved in the new web site, since end of 2009. And thanks should go to the whole team ! Especially regarding to the design, testing and kicking me to finish the side :) cheers W. -- William http://www.wbonnet.net http://www.sunwizard.net Le site fran?ais des amateurs de stations Unix http://www.opencsw.org Community SoftWare for Solaris http://www.guses.org French speaking Solaris User Group From skayser at opencsw.org Wed Jun 16 12:36:04 2010 From: skayser at opencsw.org (Sebastian Kayser) Date: Wed, 16 Jun 2010 12:36:04 +0200 Subject: [csw-maintainers] website issues In-Reply-To: References: Message-ID: <20100616103604.GB31767@sebastiankayser.de> * Maciej (Matchek) Blizinski wrote: > 3. No links to the search subpage. 4. Package links to the bug tracker not always correctly determined http://www.opencsw.org/packages/pm_paramsvalidate/ -> http://www.opencsw.org/mantis/set_project.php?project_id= (empty project_id) Sebastian From maciej at opencsw.org Wed Jun 16 14:19:45 2010 From: maciej at opencsw.org (Maciej (Matchek) Blizinski) Date: Wed, 16 Jun 2010 13:19:45 +0100 Subject: [csw-maintainers] libev and libevent In-Reply-To: <7EE3CF2E-4D6C-41B6-AD4B-6F9228F0E358@opencsw.org> References: <54DDD943-FE7E-418F-8FB9-FDC5C91B88FC@baltic-online.de> <7EE3CF2E-4D6C-41B6-AD4B-6F9228F0E358@opencsw.org> Message-ID: No dia 12 de Maio de 2010 12:39, Dagobert Michelsen escreveu: > because of the collision of event.h from libev and libevent I am > deinstalling CSWlibev for now as tmux which I am building right > now does not work with the event.h from libev. I have not looked > into the proposed solution from > ?http://www.opencsw.org/mantis/view.php?id=4376 > but may do so soon :) Debian renames event.h to ev-event.h for libev. If people think this is acceptable, I'll send the updated package for release. From phil at bolthole.com Wed Jun 16 15:03:55 2010 From: phil at bolthole.com (Philip Brown) Date: Wed, 16 Jun 2010 06:03:55 -0700 Subject: [csw-maintainers] libev and libevent In-Reply-To: References: <54DDD943-FE7E-418F-8FB9-FDC5C91B88FC@baltic-online.de> <7EE3CF2E-4D6C-41B6-AD4B-6F9228F0E358@opencsw.org> Message-ID: okay On Wednesday, June 16, 2010, Maciej (Matchek) Blizinski wrote: > No dia 12 de Maio de 2010 12:39, Dagobert Michelsen escreveu: >> because of the collision of event.h from libev and libevent I am >> deinstalling CSWlibev for now as tmux which I am building right >> now does not work with the event.h from libev. I have not looked >> into the proposed solution from >> ?http://www.opencsw.org/mantis/view.php?id=4376 >> but may do so soon :) > > Debian renames event.h to ev-event.h for libev. ?If people think this > is acceptable, I'll send the updated package for release. > _______________________________________________ > maintainers mailing list > maintainers at lists.opencsw.org > https://lists.opencsw.org/mailman/listinfo/maintainers > .:: This mailing list's archive is public. ::. From skayser at opencsw.org Wed Jun 16 15:05:14 2010 From: skayser at opencsw.org (Sebastian Kayser) Date: Wed, 16 Jun 2010 15:05:14 +0200 Subject: [csw-maintainers] website issues In-Reply-To: <20100616103604.GB31767@sebastiankayser.de> References: <20100616103604.GB31767@sebastiankayser.de> Message-ID: <20100616130513.GC31767@sebastiankayser.de> * Sebastian Kayser wrote: > * Maciej (Matchek) Blizinski wrote: > > 3. No links to the search subpage. > > 4. Package links to the bug tracker not always correctly determined > http://www.opencsw.org/packages/pm_paramsvalidate/ > -> http://www.opencsw.org/mantis/set_project.php?project_id= > (empty project_id) 5. GD jpeg handling and thereby Mantis captcha generation for signups is broken https://www.opencsw.org/mantis/signup_page.php https://www.opencsw.org/mantis/make_captcha_img.php?public_key=93738 Apache error log holds: gd-jpeg: JPEG library reports unrecoverable error: Wrong JPEG library version: library is 62, caller expects 70 Sebastian From phil at bolthole.com Wed Jun 16 16:04:39 2010 From: phil at bolthole.com (Philip Brown) Date: Wed, 16 Jun 2010 07:04:39 -0700 Subject: [csw-maintainers] web site issue: email Message-ID: <20100616140439.GA19594@bolthole.com> http://www.opencsw.org/sendEmail "Server error :( Cannot submit form.Please try again later. " From maciej at opencsw.org Wed Jun 16 16:32:33 2010 From: maciej at opencsw.org (Maciej (Matchek) Blizinski) Date: Wed, 16 Jun 2010 15:32:33 +0100 Subject: [csw-maintainers] cswclassutils: location of template init scripts In-Reply-To: References: Message-ID: No dia 11 de Junho de 2010 17:22, Philip Brown escreveu: > On Fri, Jun 11, 2010 at 6:38 AM, Maciej (Matchek) Blizinski >> In any case, I'm weary of throwing away or significantly modifying >> what we have right now, just because of the NFS case. ?I'd rather add >> whatever NFS setup needs, without disturbing the existing solution. >> > > I believe the above proposal meets that criteria. The only > "significant modification" is a one-time alteration of the cswinitsmf > CAS package. True, it's a one time modification, assuming there are no bugs. The modification has some drawbacks though; it doesn't use the system mechanisms in an intuitive way. If I were to debug a problem with such solution, soon enough I'd start brandishing blade weapons in the author's general direction. As a side note, I've just discovered that the same issue concerns our alternatives script, so our sudo package is broken too and I need to get it fixed relatively soon. Our discussion so far looked like: Person A: "I think we should use solution A1, because of reasons R1 and R2." Person B: "I think we should use solution B1 or B2 because of reasons R3 and R4." At no point there's enough information on one page to make an informed decision. We probably need to do a complete assessment similar to the one that we did for cswmigrateconf[1]. It'll require some time to write, but we can start a stub pretty quickly. Thoughts? [1] http://wiki.opencsw.org/configuration-directory-migration From maciej at opencsw.org Wed Jun 16 17:36:55 2010 From: maciej at opencsw.org (Maciej (Matchek) Blizinski) Date: Wed, 16 Jun 2010 16:36:55 +0100 Subject: [csw-maintainers] cswclassutils: location of template init scripts In-Reply-To: References: Message-ID: No dia 16 de Junho de 2010 15:32, Maciej (Matchek) Blizinski escreveu: > At no point there's enough information on one page to make an informed > decision. ?We probably need to do a complete assessment similar to the > one that we did for cswmigrateconf[1]. ?It'll require some time to > write, but we can start a stub pretty quickly. Here's a sketch of the comparison table: http://wiki.opencsw.org/cas-and-shared-nfs-support The criteria are phrased such that "Yes" can be always green and "No" can be always red. Maciej From phil at bolthole.com Wed Jun 16 18:12:31 2010 From: phil at bolthole.com (Philip Brown) Date: Wed, 16 Jun 2010 09:12:31 -0700 Subject: [csw-maintainers] sudo, and alternatives implementation Message-ID: On Wed, Jun 16, 2010 at 7:32 AM, Maciej (Matchek) Blizinski wrote: > >.... > > As a side note, I've just discovered that the same issue concerns our > alternatives script, so our sudo package is broken too and I need to > get it fixed relatively soon. I started a separate thread on this, because I think it will probably lead to a different solution set. Could you give specifics of the problem, please? I thought I designed our csw implementation of alternatives, to specifically be nfs friendly. From phil at bolthole.com Wed Jun 16 18:16:22 2010 From: phil at bolthole.com (Philip Brown) Date: Wed, 16 Jun 2010 09:16:22 -0700 Subject: [csw-maintainers] cswclassutils: location of template init scripts In-Reply-To: References: Message-ID: On Wed, Jun 16, 2010 at 8:36 AM, Maciej (Matchek) Blizinski wrote: > No dia 16 de Junho de 2010 15:32, Maciej (Matchek) Blizinski > escreveu: >> At no point there's enough information on one page to make an informed >> decision. ?We probably need to do a complete assessment similar to the >> one that we did for cswmigrateconf[1]. ?It'll require some time to >> write, but we can start a stub pretty quickly. > > Here's a sketch of the comparison table: > http://wiki.opencsw.org/cas-and-shared-nfs-support > > The criteria are phrased such that "Yes" can be always green and "No" > can be always red. > i dont see coverage of the more complex to explain case, of [have the CAS script handle things in a uniform manner, reguardless of whether script is shipped in /etc or opt] I also dont understand the purpose of the column labeled, "No modification to cswinitsmf (no bugs)" From maciej at opencsw.org Wed Jun 16 19:10:50 2010 From: maciej at opencsw.org (Maciej (Matchek) Blizinski) Date: Wed, 16 Jun 2010 18:10:50 +0100 Subject: [csw-maintainers] sudo, and alternatives implementation In-Reply-To: References: Message-ID: No dia 16 de Junho de 2010 17:12, Philip Brown escreveu: > On Wed, Jun 16, 2010 at 7:32 AM, Maciej (Matchek) Blizinski > wrote: >> >.... >> >> As a side note, I've just discovered that the same issue concerns our >> alternatives script, so our sudo package is broken too and I need to >> get it fixed relatively soon. > > > I started a separate thread on this, because I think it will probably > lead to a different solution set. > > Could you give specifics of the problem, please? > I thought I designed our csw implementation of alternatives, to > specifically be nfs friendly. The alternatives file gets installed under /opt, its class is 'cswalternatives'. It's the same case as with the SMF integration, the CAS script is called, but doesn't get to handle the alternatives file in the sparse non-global zones. The effect is that the symlinks aren't created in the non-global zones. It's not really the problem of the alternatives per se, it's the CAS integration issue. This is a very similar case to the SMF one: The installed file serves as data for the CAS script to do some work. We need the work to be done in all the zones, but we happened to make the trigger file global-zone-only. From phil at bolthole.com Wed Jun 16 19:22:59 2010 From: phil at bolthole.com (Philip Brown) Date: Wed, 16 Jun 2010 10:22:59 -0700 Subject: [csw-maintainers] sudo, and alternatives implementation In-Reply-To: References: Message-ID: On Wed, Jun 16, 2010 at 10:10 AM, Maciej (Matchek) Blizinski wrote: > No dia 16 de Junho de 2010 17:12, Philip Brown escreveu: > >> Could you give specifics of the problem, please? >> I thought I designed our csw implementation of alternatives, to >> specifically be nfs friendly. > > The alternatives file gets installed under /opt, its class is > 'cswalternatives'. ?It's the same case as with the SMF integration, > the CAS script is called, but doesn't get to handle the alternatives > file in the sparse non-global zones. ?The effect is that the symlinks > aren't created in the non-global zones. errr.. but that shouldnt be a problem, because ideally, we should only be creating symlinks with it under /opt/csw in the first place. and so they should already be set up in sparse non-global zones. This was part of my implicit design assumptions. If someone is using alternatives, to set up symlinks ELSEWHERE... then I'd like to know why. becuase that would make no sense to me (and in that case we'd need to set up some explicit docs for "dont use 'alternatives' outside of /opt/csw" :-} ) From maciej at opencsw.org Wed Jun 16 19:26:09 2010 From: maciej at opencsw.org (Maciej (Matchek) Blizinski) Date: Wed, 16 Jun 2010 18:26:09 +0100 Subject: [csw-maintainers] cswclassutils: location of template init scripts In-Reply-To: References: Message-ID: No dia 16 de Junho de 2010 17:16, Philip Brown escreveu: > i dont see coverage of the more complex to explain case, of > [have the CAS script handle things in a uniform manner, reguardless of > whether script is shipped in /etc or opt] It's an ingenious idea, no question about it. The definition of CAS is that they handle the installation of files, with potential additional processing. Having CAS process other files is something that CAS isn't intended for. I know that you can hack it, but it might cross the fine line between use and abuse. Perhaps CAS are not the right way to solve this problem? If you just want to run a script upon package installation, this is what pre- and postinstall are for. > I also dont understand the purpose of the column labeled, > "No modification to cswinitsmf (no bugs)" The purpose is similar to other criterion columns: point out any flaws in described approaches. What is the specific issue with this column? From phil at bolthole.com Wed Jun 16 19:38:15 2010 From: phil at bolthole.com (Philip Brown) Date: Wed, 16 Jun 2010 10:38:15 -0700 Subject: [csw-maintainers] cswclassutils: location of template init scripts In-Reply-To: References: Message-ID: On Wed, Jun 16, 2010 at 10:26 AM, Maciej (Matchek) Blizinski wrote: > ... ?The definition of CAS > is that they handle the installation of files, with potential > additional processing. ?Having CAS process other files is something > that CAS isn't intended for. ?I know that you can hack it, but it > might cross the fine line between use and abuse. I disagree. Sun has used it in this way for years. almost a decade? one of the "standard, documented(but often ignored)" ways, used it in this sort of way. the cron, sed, or awk class I think. dont remember which. by definition, the sun cron class is an example of, "take one file, and use a CAS to modify another file (crontab)" > > Perhaps CAS are not the right way to solve this problem? ?If you just > want to run a script upon package installation, this is what pre- and > postinstall are for. but "just run a script" is not the full design goal. the full design goal is, "and not bug users about prompts for 'do you wish to allow this unknown script to run?' for common , known, shared framework scripts, while at the same time STILL prompting (if the user desires) for truly unknown one-shot scripts" The CAS scripts may eventually be shared by hundreds of our packages. The site admin should be able to evaluate the safety of the script(s) one time, and then have trust in them. The CAS approach allows this. The regular postinstall script approach does not. > >> I also dont understand the purpose of the column labeled, >> "No modification to cswinitsmf (no bugs)" > > The purpose is similar to other criterion columns: point out any flaws > in described approaches. ?What is the specific issue with this column? let me rephrase: I dont understand the column :) From phil at bolthole.com Wed Jun 16 19:41:41 2010 From: phil at bolthole.com (Philip Brown) Date: Wed, 16 Jun 2010 10:41:41 -0700 Subject: [csw-maintainers] cswclassutils: location of template init scripts In-Reply-To: References: Message-ID: On Wed, Jun 16, 2010 at 10:38 AM, Philip Brown wrote: > > The CAS scripts ?may eventually be shared by hundreds of our packages. > The site admin should be able to evaluate the safety of the script(s) > one time, and then have trust in them. > The CAS approach allows this. The regular postinstall script approach does not. > Speaking of which.... I'm not happy with the frequency of which our cswclassutils package keeps getting updated. I'm beginning to think we need to split the stuff out into separate packages for each action. That way, the impact on our end users when one, potentially new script is changed, is thus minimized. From bwalton at opencsw.org Thu Jun 17 03:18:05 2010 From: bwalton at opencsw.org (Ben Walton) Date: Wed, 16 Jun 2010 21:18:05 -0400 Subject: [csw-maintainers] mantis issue Message-ID: <1276737446-sup-1588@pinkfloyd.chass.utoronto.ca> Hi All, I'm seeing the following while filing bugs in mantis. Is this already known or due to a recent change? SYSTEM WARNING: error_log(/var/opt/csw/mantis/logfile) [function.error-log]: failed to open stream: No such file or directory Thanks -Ben -- Ben Walton Systems Programmer - CHASS University of Toronto C:416.407.5610 | W:416.978.4302 From bwalton at opencsw.org Thu Jun 17 03:24:17 2010 From: bwalton at opencsw.org (Ben Walton) Date: Wed, 16 Jun 2010 21:24:17 -0400 Subject: [csw-maintainers] gnulinks bugs Message-ID: <1276737765-sup-8053@pinkfloyd.chass.utoronto.ca> Hi All, I've just filed 'packaging' bugs against each package that is included in the set for which CSWgnulinks provides symlinks from /opt/csw/gnu -> /opt/csw/bin. When you get a change, please rebuild your package with added symlinks for each g* binary. Before releasing it, let me know that you're ready so I can provide an updated CSWgnulinks with your package(s) removed. Thanks -Ben -- Ben Walton Systems Programmer - CHASS University of Toronto C:416.407.5610 | W:416.978.4302 From skayser at opencsw.org Thu Jun 17 10:42:26 2010 From: skayser at opencsw.org (Sebastian Kayser) Date: Thu, 17 Jun 2010 10:42:26 +0200 Subject: [csw-maintainers] mantis issue In-Reply-To: <1276737446-sup-1588@pinkfloyd.chass.utoronto.ca> References: <1276737446-sup-1588@pinkfloyd.chass.utoronto.ca> Message-ID: <20100617084226.GD31767@sebastiankayser.de> * Ben Walton wrote: > I'm seeing the following while filing bugs in mantis. Is this already > known or due to a recent change? > > SYSTEM WARNING: error_log(/var/opt/csw/mantis/logfile) > [function.error-log]: failed to open stream: No such file or directory No Mantis configuration change. The installation was moved to another system, Ihsan needs to create the directory and make it writable for the Mantis/CGI user. Sebastian From phil at bolthole.com Fri Jun 18 03:53:33 2010 From: phil at bolthole.com (Philip Brown) Date: Thu, 17 Jun 2010 18:53:33 -0700 Subject: [csw-maintainers] packaging humour Message-ID: <20100618015333.GA30950@bolthole.com> http://www.xkcd.com/754/ Then check the "hidden" mouseover comment for the strip. (or click "properties" for the image, and expand the window, or view source, or whatever tickles your fancy) From phil at bolthole.com Fri Jun 18 06:17:43 2010 From: phil at bolthole.com (Philip Brown) Date: Thu, 17 Jun 2010 21:17:43 -0700 Subject: [csw-maintainers] web site: admins??? Message-ID: <20100618041743.GA39761@bolthole.com> Guys... what's going on with the web site, and who has the power to fix it? it's been broken for DAYS now ... ? (as far as the packages listing pages) I'd try to fix it myself, but I cant even get IN to www.opencsw.org any more, let alone have the power to fix it. From maciej at opencsw.org Fri Jun 18 11:03:42 2010 From: maciej at opencsw.org (Maciej (Matchek) Blizinski) Date: Fri, 18 Jun 2010 10:03:42 +0100 Subject: [csw-maintainers] packaging humour In-Reply-To: <20100618015333.GA30950@bolthole.com> References: <20100618015333.GA30950@bolthole.com> Message-ID: No dia 18 de Junho de 2010 02:53, Philip Brown escreveu: > > > http://www.xkcd.com/754/ > > Then check the "hidden" mouseover comment for the strip. > (or click "properties" for the image, and expand the window, > ?or view source, or whatever tickles your fancy) What are you saying? Click? Expand the window? I did it the proper way, as follows: echo -ne "GET /754/ HTTP/1.1\r\nHost: xkcd.com\r\n\r\n" \ | nc xckd.com 80 | grep "^ I think it's James who runs our mirror status page but I'm not sure so I'm also addressing the maintainers list. http://www.canoedissent.org.uk/mirror/status/ As you can see at the bottom there's a few mirrors that have not worked for quite a while and they are still listed. However, my mirror of choice - http://ftp.df.lth.se/pub/csw/, is not listed any more. They suffered a raid hardware failure two weeks ago and everything was lost. Since it's based at a university they haven't prepared for this in a professional way so they had to fix the hardware and start mirroring all over again. They expect the CSW mirror to be up again sometime this weekend so I would like to have it listed again. http://ftp.df.lth.se/pub/csw ftp://ftp.df.lth.se/pub/csw Thank you. :-) -- /peter From phil at bolthole.com Fri Jun 18 17:00:44 2010 From: phil at bolthole.com (Philip Brown) Date: Fri, 18 Jun 2010 08:00:44 -0700 Subject: [csw-maintainers] web site: admins??? In-Reply-To: <20100618041743.GA39761@bolthole.com> References: <20100618041743.GA39761@bolthole.com> Message-ID: Thank you to whoever fixed the package page display. (you're just not going to admit you fixed it,so you dont get bugged in the future...? :-) On Thu, Jun 17, 2010 at 9:17 PM, Philip Brown wrote: > Guys... what's going on with the web site, and who has the power to fix it? > > it's been broken for DAYS now ... ? > > (as far as the packages listing pages) > > I'd try to fix it myself, but I cant even get IN to www.opencsw.org any > more, let alone have the power to fix it. > _______________________________________________ > maintainers mailing list > maintainers at lists.opencsw.org > https://lists.opencsw.org/mailman/listinfo/maintainers > .:: This mailing list's archive is public. ::. > From phil at bolthole.com Fri Jun 18 17:02:04 2010 From: phil at bolthole.com (Philip Brown) Date: Fri, 18 Jun 2010 08:02:04 -0700 Subject: [csw-maintainers] packaging humour In-Reply-To: References: <20100618015333.GA30950@bolthole.com> Message-ID: On Fri, Jun 18, 2010 at 2:03 AM, Maciej (Matchek) Blizinski wrote: > No dia 18 de Junho de 2010 02:53, Philip Brown escreveu: >> >> >> http://www.xkcd.com/754/ >> >> Then check the "hidden" mouseover comment for the strip. >> (or click "properties" for the image, and expand the window, >> or view source, or whatever tickles your fancy) > > What are you saying? Click? Expand the window? I did it the proper > way, as follows: > > echo -ne "GET /754/ HTTP/1.1\r\nHost: xkcd.com\r\n\r\n" \ > | nc xckd.com 80 | grep "^ > ;-) > thats.... thats just.... well you were just lucky there were no line wraps :) From phil at bolthole.com Fri Jun 18 19:16:41 2010 From: phil at bolthole.com (Philip Brown) Date: Fri, 18 Jun 2010 10:16:41 -0700 Subject: [csw-maintainers] comments and info on new gtk, etc Message-ID: Some info for interested people: So, we have the "new" gtk, cairo, and pango, officially in our public current tree now. The bad news is, our existing firefox3.0 package is kinda slow with it. The worse news is, getting a fresh compile, out of what's in our gar tree, is somewhat painful. To the point where I havent been successful with it. The GOOD news is.... the "sun beijing" contributed packages of firefox 3.6.3, work **really great** with it! So at least we know we're on the right track. For those interested in the 3.6 binaries, do the following: Go to http://releases.mozilla.org/pub/mozilla.org/firefox/releases/latest-3.6/contrib/solaris_pkgadd/ download the appropriate fcs pkg.bz2 file bunzip2 it. use pkgtrans on it( pkgtrans firefox*-pkg /tmp) **DO NOT** say "all". pick the specific firefox package. Or, I guess the short cut is pkgtrans firefox*-pkg /tmp SFWfirefox pkgadd -d /tmp/SFWfirefox EDIT /opt/sfw/lib/firefox/run-mozilla.sh They take a shortcut, and force LD_LIBRARY_PATH near the end of the script. you need to re-force it to be LD_LIBRARY_PATH=/opt/csw/lib:/opt/sfw/lib All done! For the curious, their full package set provides updated versions of: GNOME2 SFWatk ATK - Accesibility Toolkit Libraries GNOME2 SFWcairo Vector graphics library GNOME2 SFWfirefox Mozilla Firefox Web browser GNOME2 SFWglib2 Low level core compatibility library for GTK+ and GNOME GNOME2 SFWgtk2 GTK+ - GIMP Toolkit Library for creation of graphical user interfaces GNOME2 SFWpango Library for layout and rendering of internationalized text GNOME2 SFWpixman Vector graphics library Installed space is: 66M for SFWfirefox, 23 megs for all the other stuff. From james at opencsw.org Fri Jun 18 20:05:33 2010 From: james at opencsw.org (James Lee) Date: Fri, 18 Jun 2010 18:05:33 GMT Subject: [csw-maintainers] Mirror status page In-Reply-To: References: Message-ID: <20100618.18053300.2273110464@gyor.oxdrove.co.uk> On 18/06/10, 15:56:44, Peter Bonivart wrote regarding Mirror status page: > As you can see at the bottom there's a few mirrors that have not > worked for quite a while and they are still listed. Those mirrors are working well but they are offering an old file. > However, my mirror > of choice - http://ftp.df.lth.se/pub/csw/, is not listed any more. > They suffered a raid hardware failure two weeks ago and everything was > lost. Since it's based at a university they haven't prepared for this > in a professional way so they had to fix the hardware and start > mirroring all over again. They expect the CSW mirror to be up again > sometime this weekend so I would like to have it listed again. > http://ftp.df.lth.se/pub/csw > ftp://ftp.df.lth.se/pub/csw Relisted. You can DIY add here: http://www.canoedissent.org.uk/mirror/add/ They last worked at 2010-06-03 16:21. After repeated failures the probe stops asking. James. From james at opencsw.org Fri Jun 18 20:41:41 2010 From: james at opencsw.org (James Lee) Date: Fri, 18 Jun 2010 18:41:41 GMT Subject: [csw-maintainers] Mirror status page In-Reply-To: <20100618.18053300.2273110464@gyor.oxdrove.co.uk> References: <20100618.18053300.2273110464@gyor.oxdrove.co.uk> Message-ID: <20100618.18414100.1996401438@gyor.oxdrove.co.uk> On 18/06/10, 19:05:33, James Lee wrote regarding Re: [csw-maintainers] Mirror status page: > > mirroring all over again. They expect the CSW mirror to be up again > > sometime this weekend so I would like to have it listed again. > > http://ftp.df.lth.se/pub/csw > > ftp://ftp.df.lth.se/pub/csw > Relisted. And they promptly delisted on the first probe because they are not actually working. $ curl -I http://ftp.df.lth.se/pub/csw/current/sparc/5.10/catalog HTTP/1.1 503 Service Temporarily Unavailable Date: Fri, 18 Jun 2010 18:41:17 GMT Server: Apache/2.2.9 (Debian) PHP/5.2.6-1+lenny4 with Suhosin-Patch mod_ssl/2.2.9 OpenSSL/0.9.8n Last-Modified: Thu, 03 Jun 2010 19:25:02 GMT ETag: "df903-117-4882529b8ff80" Accept-Ranges: bytes Content-Length: 279 Connection: close Content-Type: text/html; charset=iso-8859-1 I've just enhanced the relist procedure so it won't relist non-working mirrors, they will have to wait until they are running. James. From bonivart at opencsw.org Fri Jun 18 20:46:14 2010 From: bonivart at opencsw.org (Peter Bonivart) Date: Fri, 18 Jun 2010 20:46:14 +0200 Subject: [csw-maintainers] Mirror status page In-Reply-To: <20100618.18414100.1996401438@gyor.oxdrove.co.uk> References: <20100618.18053300.2273110464@gyor.oxdrove.co.uk> <20100618.18414100.1996401438@gyor.oxdrove.co.uk> Message-ID: On Fri, Jun 18, 2010 at 8:41 PM, James Lee wrote: > I've just enhanced the relist procedure so it won't relist non-working > mirrors, they will have to wait until they are running. Thank you for the answers. :-) -- /peter From phil at bolthole.com Fri Jun 18 23:46:49 2010 From: phil at bolthole.com (Philip Brown) Date: Fri, 18 Jun 2010 14:46:49 -0700 Subject: [csw-maintainers] web site issue Message-ID: <20100618214649.GA47640@bolthole.com> (First issue is: there needs to be a "webmaster" contact!) The mirrors page, is very old. it's missing the ibiblio conversion. the www.ibiblio.org url, needs to be converted to the newer mirrors.ibiblio.org url. (that happened around january, I think!) http://mirrors.ibiblio.org/pub/mirrors/opencsw From phil at bolthole.com Sat Jun 19 00:08:29 2010 From: phil at bolthole.com (Philip Brown) Date: Fri, 18 Jun 2010 15:08:29 -0700 Subject: [csw-maintainers] a question of gamin Message-ID: I've hit something rather odd, about a "new" package, called gamin. its not exactly new: its a replacement for "fam". gamin is favored by GNOME peoples. and sun. The principle is that its a background demon that gets spawned, watching for "file alterations". http://people.gnome.org/~veillard/gamin/ The trouble is: its horribly horribly written. Unless you have the linux specific(of course) kernel calls.. OR some magical opensolaris hooks... it goes into a huge cpu-stealing polling loop. sucking up 50% of cpu or worse. I've tried tweaking its config file to dial down polling interfaces. It doesnt seem to work. (Interestingly, i've seen complaints on other systems about it too) Leaving it as-is, doesnt seem to really be viable. I see a few options, and I'm soliciting opinions on them: (option 0: someone else finds a fix for it ;-) option: Just ship the client-side lib, but disable/not ship the backend demon. On OpenSolaris, it will actually integrate with an already running /usr/lib/gam_server, and work great. Otherwise, it ("it" being gtk/glib) will whine a few times to stdout/err... and then shut up with no harm done. Hmm... CORRECTION. while gimp and firefox handle this nicely.. our old-old 'gedit', crashes badly, unless it can actually FIND it. arrg. So, this option does not seem viable option: leave as-is. yuk. option: Completely disable "FAM" support from glib, and other things we have, such as gnome-vfs I'm honestly not sure what exactly we "lose" this way. or rather, not sure its that important. See http://www.opencsw.org/packages/fam/ for dependant packages. side note: the original "libfam" is an orphaned project from sgi.com. hasnt been touched in 6 years? and requires modifying inetd.conf. for a user-level tool. yuk. gamin seems to be the only replacement going forward. From phil at bolthole.com Sat Jun 19 00:15:23 2010 From: phil at bolthole.com (Philip Brown) Date: Fri, 18 Jun 2010 15:15:23 -0700 Subject: [csw-maintainers] a question of gamin In-Reply-To: References: Message-ID: On Fri, Jun 18, 2010 at 3:08 PM, Philip Brown wrote: > option: Completely disable "FAM" support from glib, and other things > we have, such as gnome-vfs > ? I'm honestly not sure what exactly we "lose" this way. or rather, > not sure its that important. PS: I did a little test. locally compiled gnome-vfs2, WITHOUT fam support. then removed "libgiofam" from glib2. when I opened a file in gedit, then tweaked it in another window before saving, gedit still detected that the file had changed. SO.... I dont think this is a particularly big deal to lose. From bwalton at opencsw.org Sat Jun 19 01:42:24 2010 From: bwalton at opencsw.org (Ben Walton) Date: Fri, 18 Jun 2010 19:42:24 -0400 Subject: [csw-maintainers] a question of gamin In-Reply-To: References: Message-ID: <1276904462-sup-9663@pinkfloyd.chass.utoronto.ca> Excerpts from Philip Brown's message of Fri Jun 18 18:15:23 -0400 2010: > SO.... I dont think this is a particularly big deal to lose. You may see more in the way of issues when (and this is well in the future, I think), you add remove files in a directory viewed by an open nautilus window or something. Just a guess, but that's where I'd expect to see the issues more than with concurrent file edits. (Obviously this doesn't get you closer to an answer...) Thanks -Ben -- Ben Walton Systems Programmer - CHASS University of Toronto C:416.407.5610 | W:416.978.4302 From phil at bolthole.com Sat Jun 19 03:05:53 2010 From: phil at bolthole.com (Philip Brown) Date: Fri, 18 Jun 2010 18:05:53 -0700 Subject: [csw-maintainers] a question of gamin In-Reply-To: <1276904462-sup-9663@pinkfloyd.chass.utoronto.ca> References: <1276904462-sup-9663@pinkfloyd.chass.utoronto.ca> Message-ID: On Friday, June 18, 2010, Ben Walton wrote: > > You may see more in the way of issues when (and this is well in the > future, I think), you add remove files in a directory viewed by an > open nautilus window or something. that would make sense. bur I view that as just a minor inconvenience, easily worked around From bwalton at opencsw.org Sat Jun 19 03:09:22 2010 From: bwalton at opencsw.org (Ben Walton) Date: Fri, 18 Jun 2010 21:09:22 -0400 Subject: [csw-maintainers] a question of gamin In-Reply-To: References: <1276904462-sup-9663@pinkfloyd.chass.utoronto.ca> Message-ID: <1276909735-sup-1509@pinkfloyd.chass.utoronto.ca> Excerpts from Philip Brown's message of Fri Jun 18 21:05:53 -0400 2010: > that would make sense. bur I view that as just a minor inconvenience, > easily worked around Agreed. It's "nice to have" but not essential. Ctrl+R works just fine. -Ben -- Ben Walton Systems Programmer - CHASS University of Toronto C:416.407.5610 | W:416.978.4302 From bwalton at opencsw.org Sat Jun 19 05:36:25 2010 From: bwalton at opencsw.org (Ben Walton) Date: Fri, 18 Jun 2010 23:36:25 -0400 Subject: [csw-maintainers] bad interactions with CSWcommon Message-ID: <1276917687-sup-6307@pinkfloyd.chass.utoronto.ca> Hi All, A bug was filed against CSWcoreutils that I've tracked down to a bad interaction with CSWcommon. Consider: # grep csw/share/locale/es /var/sadm/install/contents | egrep "(CSWcommon|CSWcoreutils)" /opt/csw/share/locale/es d none 0755 root bin CSWbinutils CSWgmake CSWmeld CSWfindutils CSWgconf2 CSWbonobo2 CSWcommon CSWgawk /opt/csw/share/locale/es/LC_MESSAGES d none 0755 root bin CSWbinutils CSWgmake CSWmeld CSWfindutils CSWgconf2 CSWbonobo2 CSWcommon CSWgawk /opt/csw/share/locale/es/LC_MESSAGES/coreutils.mo f none 0644 root bin 113352 38429 1274322259 CSWcoreutils /opt/csw/share/locale/es/LC_TIME=LC_MESSAGES s none CSWcommon /opt/csw/share/locale/es/LC_TIME/coreutils.mo=../LC_MESSAGES/coreutils.mo s none CSWcoreutils CSWcoreutils wants to provide: $ grep locale/es CSWcoreutils.prototype f none /opt/csw/share/locale/es/LC_MESSAGES/coreutils.mo 0644 root bin s none /opt/csw/share/locale/es/LC_TIME/coreutils.mo=../LC_MESSAGES/coreutils.mo So, when CSWcoreutils is installed, the LC_TIME/coreutils.mo symlink is delivered as LC_MESSAGES/coreutils.mo -> ../LC_MESSAGES/coreutils.mo (self referential symlink)... I thought initially that I could work around this by simply filtering out any LC_TIME entries from the coreutils prototype, but not every LC_TIME is a symlink to the corresponding LC_MESSAGES (eg: locale/vi). Am I missing an obvious fix here? I'm happy to settle for a non-obvious fix too, but I don't see a good one right now. Have other packages encountered this? For reference, the open bug is here: https://www.opencsw.org/mantis/view.php?id=4454 Thanks -Ben -- Ben Walton Systems Programmer - CHASS University of Toronto C:416.407.5610 | W:416.978.4302 From william at wbonnet.net Sat Jun 19 12:43:25 2010 From: william at wbonnet.net (William Bonnet) Date: Sat, 19 Jun 2010 12:43:25 +0200 Subject: [csw-maintainers] web site: admins??? In-Reply-To: References: <20100618041743.GA39761@bolthole.com> Message-ID: <4C1C9F4D.707@wbonnet.net> Hi > (you're just not going to admit you fixed it,so you dont get bugged in > the future...? :-) > It has been fixed. Sorry for this issue. It looks like concurrent modifications broke it :( cheers W. -- William http://www.wbonnet.net http://www.sunwizard.net Le site fran?ais des amateurs de stations Unix http://www.opencsw.org Community SoftWare for Solaris http://www.guses.org French speaking Solaris User Group From william at wbonnet.net Sat Jun 19 12:53:22 2010 From: william at wbonnet.net (William Bonnet) Date: Sat, 19 Jun 2010 12:53:22 +0200 Subject: [csw-maintainers] web site issue In-Reply-To: <20100618214649.GA47640@bolthole.com> References: <20100618214649.GA47640@bolthole.com> Message-ID: <4C1CA1A2.5090305@wbonnet.net> Hi Phil > (First issue is: there needs to be a "webmaster" contact!) > Will do it this week end > http://mirrors.ibiblio.org/pub/mirrors/opencsw > I think it's done now. Please can you check i modified it correctly ( please tell if i forgot links outside of the mirrors page) cheers W. -- William http://www.wbonnet.net http://www.sunwizard.net Le site fran?ais des amateurs de stations Unix http://www.opencsw.org Community SoftWare for Solaris http://www.guses.org French speaking Solaris User Group From rupert at opencsw.org Sat Jun 19 12:53:29 2010 From: rupert at opencsw.org (rupert THURNER) Date: Sat, 19 Jun 2010 12:53:29 +0200 Subject: [csw-maintainers] apr-util - searches for libiconv.la - how to use .so ? In-Reply-To: <1274531028-sup-9935@pinkfloyd.chass.utoronto.ca> References: <1274531028-sup-9935@pinkfloyd.chass.utoronto.ca> Message-ID: On Sat, May 22, 2010 at 14:26, Ben Walton wrote: > Excerpts from rupert THURNER's message of Sat May 22 07:41:14 -0400 2010: > > Hi Rupert, > >> how could one convince apr-util to not search for ?libiconv.la, but >> use what is existing and working? > > Have you turned on the 'STRIP_LIBTOOL = 1' option in your GAR recipe? > If not, try that first. ?If it still doesn't work nicely, you'll need > a bigger hammer...I've got one recipe where the trick about didn't > work and I had to correct it through other means. ?See the Openjade > build recipe here: > https://gar.svn.sourceforge.net/svnroot/gar/csw/mgar/pkg/openjade/trunk/Makefile ben, you meen this code? echo 'ARGS=`echo "$$@" | gsed "s#/opt/csw/lib/libosp.la#/opt/csw/lib/libosp.so#g"`' >> libtool; i tried to grep for libiconv ... but is seems to be nowhere? $ ggrep -r libiconv trunk/work/solaris9-sparc/build-isa-sparcv8/apr-util-1.3.9/ Binary file trunk/work/solaris9-sparc/build-isa-sparcv8/apr-util-1.3.9/xlate/.libs/xlate.o matches Binary file trunk/work/solaris9-sparc/build-isa-sparcv8/apr-util-1.3.9/xlate/xlate.o matches Binary file trunk/work/solaris9-sparc/build-isa-sparcv8/apr-util-1.3.9/.libs/libaprutil-1.so matches Binary file trunk/work/solaris9-sparc/build-isa-sparcv8/apr-util-1.3.9/.libs/libaprutil-1.a matches Binary file trunk/work/solaris9-sparc/build-isa-sparcv8/apr-util-1.3.9/.libs/libaprutil-1.so.0.3.9 matches Binary file trunk/work/solaris9-sparc/build-isa-sparcv8/apr-util-1.3.9/.libs/libaprutil-1.so.0 matches trunk/work/solaris9-sparc/build-isa-sparcv8/apr-util-1.3.9/CHANGES: as required by missing libiconv. [William Rowe] trunk/work/solaris9-sparc/build-isa-sparcv8/apr-util-1.3.9/config.log:libiconv_open conftest.o trunk/work/solaris9-sparc/build-isa-sparcv8/apr-util-1.3.9/config.log:libiconv conftest.o currently i only see two possibilities: 1. somebody with more knowledge and/or more time takes that up 2. we remove apr for the moment from unstable and the buildserver to not block the new apache build unecessarily. rupert. From william at wbonnet.net Sat Jun 19 13:03:18 2010 From: william at wbonnet.net (William Bonnet) Date: Sat, 19 Jun 2010 13:03:18 +0200 Subject: [csw-maintainers] mantis issue In-Reply-To: <20100617084226.GD31767@sebastiankayser.de> References: <1276737446-sup-1588@pinkfloyd.chass.utoronto.ca> <20100617084226.GD31767@sebastiankayser.de> Message-ID: <4C1CA3F6.9000204@wbonnet.net> Hi >> SYSTEM WARNING: error_log(/var/opt/csw/mantis/logfile) >> [function.error-log]: failed to open stream: No such file or directory >> > > No Mantis configuration change. The installation was moved to another > system, Ihsan needs to create the directory and make it writable for the > Mantis/CGI user. > It is now fixed. I have created successfully a new issue. cheers W. -- William http://www.wbonnet.net http://www.sunwizard.net Le site fran?ais des amateurs de stations Unix http://www.opencsw.org Community SoftWare for Solaris http://www.guses.org French speaking Solaris User Group From phil at bolthole.com Sat Jun 19 14:53:04 2010 From: phil at bolthole.com (Philip Brown) Date: Sat, 19 Jun 2010 05:53:04 -0700 Subject: [csw-maintainers] web site issue In-Reply-To: <4C1CA1A2.5090305@wbonnet.net> References: <20100618214649.GA47640@bolthole.com> <4C1CA1A2.5090305@wbonnet.net> Message-ID: looks fine, thanks On Saturday, June 19, 2010, William Bonnet wrote: > Hi Phil > > (First issue is: there needs to be a "webmaster" contact!) > > > Will do it this week end > > http://mirrors.ibiblio.org/pub/mirrors/opencsw > > > I think it's done now. Please can you check i modified it correctly ( please tell if i forgot links outside of the mirrors page) > > cheers > W. > > -- > William ? ? ? ? ? ? ? ? ?http://www.wbonnet.net > > http://www.sunwizard.net Le site fran?ais des amateurs de stations Unix > http://www.opencsw.org ? Community SoftWare for Solaris > http://www.guses.org ? ? French speaking Solaris User Group > > > _______________________________________________ > maintainers mailing list > maintainers at lists.opencsw.org > https://lists.opencsw.org/mailman/listinfo/maintainers > .:: This mailing list's archive is public. ::. > From phil at bolthole.com Sat Jun 19 14:58:40 2010 From: phil at bolthole.com (Philip Brown) Date: Sat, 19 Jun 2010 05:58:40 -0700 Subject: [csw-maintainers] bad interactions with CSWcommon In-Reply-To: <1276917687-sup-6307@pinkfloyd.chass.utoronto.ca> References: <1276917687-sup-6307@pinkfloyd.chass.utoronto.ca> Message-ID: hmmm. I don't remember any more why the lnks were put in. but I suspect the missing ones arefor locales that got added later On Friday, June 18, 2010, Ben Walton wrote: > > Hi All, > > A bug was filed against CSWcoreutils that I've tracked down to a bad > interaction with CSWcommon. > > Consider: > > # grep csw/share/locale/es /var/sadm/install/contents | egrep "(CSWcommon|CSWcoreutils)" > /opt/csw/share/locale/es d none 0755 root bin CSWbinutils CSWgmake CSWmeld CSWfindutils CSWgconf2 CSWbonobo2 CSWcommon CSWgawk > /opt/csw/share/locale/es/LC_MESSAGES d none 0755 root bin CSWbinutils CSWgmake CSWmeld CSWfindutils CSWgconf2 CSWbonobo2 CSWcommon CSWgawk > /opt/csw/share/locale/es/LC_MESSAGES/coreutils.mo f none 0644 root bin 113352 38429 1274322259 CSWcoreutils > /opt/csw/share/locale/es/LC_TIME=LC_MESSAGES s none CSWcommon > /opt/csw/share/locale/es/LC_TIME/coreutils.mo=../LC_MESSAGES/coreutils.mo s none CSWcoreutils > > > CSWcoreutils wants to provide: > $ grep locale/es CSWcoreutils.prototype > f none /opt/csw/share/locale/es/LC_MESSAGES/coreutils.mo 0644 root bin > s none /opt/csw/share/locale/es/LC_TIME/coreutils.mo=../LC_MESSAGES/coreutils.mo > > So, when CSWcoreutils is installed, the LC_TIME/coreutils.mo symlink > is delivered as LC_MESSAGES/coreutils.mo -> > ../LC_MESSAGES/coreutils.mo (self referential symlink)... > > I thought initially that I could work around this by simply filtering > out any LC_TIME entries from the coreutils prototype, but not every > LC_TIME is a symlink to the corresponding LC_MESSAGES (eg: > locale/vi). > > Am I missing an obvious fix here? ?I'm happy to settle for a > non-obvious fix too, but I don't see a good one right now. Have other > packages encountered this? > > For reference, the open bug is here: > https://www.opencsw.org/mantis/view.php?id=4454 > > Thanks > -Ben > -- > Ben Walton > Systems Programmer - CHASS > University of Toronto > C:416.407.5610 | W:416.978.4302 > > _______________________________________________ > maintainers mailing list > maintainers at lists.opencsw.org > https://lists.opencsw.org/mailman/listinfo/maintainers > .:: This mailing list's archive is public. ::. > From bwalton at opencsw.org Sat Jun 19 18:30:42 2010 From: bwalton at opencsw.org (Ben Walton) Date: Sat, 19 Jun 2010 12:30:42 -0400 Subject: [csw-maintainers] apr-util - searches for libiconv.la - how to use .so ? In-Reply-To: References: <1274531028-sup-9935@pinkfloyd.chass.utoronto.ca> Message-ID: <1276964728-sup-8610@pinkfloyd.chass.utoronto.ca> Excerpts from rupert THURNER's message of Sat Jun 19 06:53:29 -0400 2010: Hi Rupert, > ben, you meen this code? > echo 'ARGS=`echo "$$@" | gsed > "s#/opt/csw/lib/libosp.la#/opt/csw/lib/libosp.so#g"`' >> libtool; Yes, but it is the whole pre-configure-modulated that makes it work. The above is just one piece of it. I dug into this a bit for you and think I've found a resolution. I copied in /opt/csw/share/build-1/libtool to the build directory and added set -x at the top of the script. I then manually ran the failing command and looked for the first occurrence of libiconv.la. This turned out to be from the use of -lsysdb (pulling in /opt/csw/lib/libsysdb.la (CSWfreetds). If you add the configure option --without-freetds, the build completes successfully (at least on current9x). So, you've got two options here: 1. Rebuild CSWfreetds; it currently belongs to Alex Moore who is retired, I think). This would clean up the libtool .la files it provides and hopefully allow you to then use it. 2. Build apr-util without freetds support. I have no idea whether this is useful functionality or not as it's the first I've ever heard of it. HTH. -Ben -- Ben Walton Systems Programmer - CHASS University of Toronto C:416.407.5610 | W:416.978.4302 From dam at opencsw.org Sat Jun 19 18:39:13 2010 From: dam at opencsw.org (Dagobert Michelsen) Date: Sat, 19 Jun 2010 18:39:13 +0200 Subject: [csw-maintainers] apr-util - searches for libiconv.la - how to use .so ? In-Reply-To: <1276964728-sup-8610@pinkfloyd.chass.utoronto.ca> References: <1274531028-sup-9935@pinkfloyd.chass.utoronto.ca> <1276964728-sup-8610@pinkfloyd.chass.utoronto.ca> Message-ID: Hi, Am 19.06.2010 um 18:30 schrieb Ben Walton: > Excerpts from rupert THURNER's message of Sat Jun 19 06:53:29 -0400 2010: > 1. Rebuild CSWfreetds; it currently belongs to Alex Moore who is > retired, I think). This would clean up the libtool .la files it > provides and hopefully allow you to then use it. I'll give freetds a try. Best regards -- Dago From rupert at opencsw.org Sat Jun 19 19:55:06 2010 From: rupert at opencsw.org (rupert THURNER) Date: Sat, 19 Jun 2010 19:55:06 +0200 Subject: [csw-maintainers] apr-util - searches for libiconv.la - how to use .so ? In-Reply-To: References: <1274531028-sup-9935@pinkfloyd.chass.utoronto.ca> <1276964728-sup-8610@pinkfloyd.chass.utoronto.ca> Message-ID: On Sat, Jun 19, 2010 at 18:39, Dagobert Michelsen wrote: > Hi, > > Am 19.06.2010 um 18:30 schrieb Ben Walton: >> Excerpts from rupert THURNER's message of Sat Jun 19 06:53:29 -0400 2010: >> 1. Rebuild CSWfreetds; it currently belongs to Alex Moore who is >> ? retired, I think). ?This would clean up the libtool .la files it >> ? provides and hopefully allow you to then use it. > > I'll give freetds a try. thanks so much ben and dago for the support! i excluded freetds until we have the new version. btw, is the following an error in gar? $ gmake platforms-reinstall platforms-remerge platforms-repackage ... gmake[1]: Entering directory `/home/rupert/mgar/pkg/apr-util/trunk' [ Reset install state for modulation isa-amd64: ISA=amd64 ] gmake[1]: Leaving directory `/home/rupert/mgar/pkg/apr-util/trunk' gmake[1]: Entering directory `/home/rupert/mgar/pkg/apr-util/trunk' gar/gar.conf.mk:427: *** The ISA 'amd64' can not be build on this kernel with the arch 'i386'. Stop. gmake[1]: Leaving directory `/home/rupert/mgar/pkg/apr-util/trunk' gmake: *** [reset-install-isa-amd64] Error 2 gmake: Leaving directory `/home/rupert/mgar/pkg/apr-util/trunk' Connection to current9x closed. gmake: *** [platforms-reinstall] Error 2 rupert at current9s ~/mgar/pkg/apr-util/trunk From dam at opencsw.org Sat Jun 19 20:17:14 2010 From: dam at opencsw.org (Dagobert Michelsen) Date: Sat, 19 Jun 2010 20:17:14 +0200 Subject: [csw-maintainers] apr-util - searches for libiconv.la - how to use .so ? In-Reply-To: References: <1274531028-sup-9935@pinkfloyd.chass.utoronto.ca> <1276964728-sup-8610@pinkfloyd.chass.utoronto.ca> Message-ID: Hi Rupert, Am 19.06.2010 um 19:55 schrieb rupert THURNER: > On Sat, Jun 19, 2010 at 18:39, Dagobert Michelsen wrote: >> Hi, >> >> Am 19.06.2010 um 18:30 schrieb Ben Walton: >>> Excerpts from rupert THURNER's message of Sat Jun 19 06:53:29 -0400 2010: >>> 1. Rebuild CSWfreetds; it currently belongs to Alex Moore who is >>> retired, I think). This would clean up the libtool .la files it >>> provides and hopefully allow you to then use it. >> >> I'll give freetds a try. > > thanks so much ben and dago for the support! i excluded freetds until > we have the new version. > > btw, is the following an error in gar? > > $ gmake platforms-reinstall platforms-remerge platforms-repackage > ... > gmake[1]: Entering directory `/home/rupert/mgar/pkg/apr-util/trunk' > [ Reset install state for modulation isa-amd64: ISA=amd64 ] > gmake[1]: Leaving directory `/home/rupert/mgar/pkg/apr-util/trunk' > gmake[1]: Entering directory `/home/rupert/mgar/pkg/apr-util/trunk' > gar/gar.conf.mk:427: *** The ISA 'amd64' can not be build on this > kernel with the arch 'i386'. Stop. > gmake[1]: Leaving directory `/home/rupert/mgar/pkg/apr-util/trunk' > gmake: *** [reset-install-isa-amd64] Error 2 > gmake: Leaving directory `/home/rupert/mgar/pkg/apr-util/trunk' > Connection to current9x closed. > gmake: *** [platforms-reinstall] Error 2 > rupert at current9s ~/mgar/pkg/apr-util/trunk It means GAR tries to reset the amd64 install state on current9x, which seems strange to me. Is this reproducable from a clean environment? Best regards -- Dago From rupert at opencsw.org Sat Jun 19 21:20:43 2010 From: rupert at opencsw.org (rupert THURNER) Date: Sat, 19 Jun 2010 21:20:43 +0200 Subject: [csw-maintainers] apr-util - searches for libiconv.la - how to use .so ? In-Reply-To: References: <1274531028-sup-9935@pinkfloyd.chass.utoronto.ca> <1276964728-sup-8610@pinkfloyd.chass.utoronto.ca> Message-ID: On Sat, Jun 19, 2010 at 20:17, Dagobert Michelsen wrote: > Hi Rupert, > > Am 19.06.2010 um 19:55 schrieb rupert THURNER: >> On Sat, Jun 19, 2010 at 18:39, Dagobert Michelsen wrote: >>> Hi, >>> >>> Am 19.06.2010 um 18:30 schrieb Ben Walton: >>>> Excerpts from rupert THURNER's message of Sat Jun 19 06:53:29 -0400 2010: >>>> 1. Rebuild CSWfreetds; it currently belongs to Alex Moore who is >>>> ? retired, I think). ?This would clean up the libtool .la files it >>>> ? provides and hopefully allow you to then use it. >>> >>> I'll give freetds a try. >> >> thanks so much ben and dago for the support! i excluded freetds until >> we have the new version. >> >> btw, is the following an error in gar? >> >> $ gmake platforms-reinstall platforms-remerge platforms-repackage >> ... >> gmake[1]: Entering directory `/home/rupert/mgar/pkg/apr-util/trunk' >> [ Reset install state for modulation isa-amd64: ISA=amd64 ] >> gmake[1]: Leaving directory `/home/rupert/mgar/pkg/apr-util/trunk' >> gmake[1]: Entering directory `/home/rupert/mgar/pkg/apr-util/trunk' >> gar/gar.conf.mk:427: *** The ISA 'amd64' can not be build on this >> kernel with the arch 'i386'. ?Stop. >> gmake[1]: Leaving directory `/home/rupert/mgar/pkg/apr-util/trunk' >> gmake: *** [reset-install-isa-amd64] Error 2 >> gmake: Leaving directory `/home/rupert/mgar/pkg/apr-util/trunk' >> Connection to current9x closed. >> gmake: *** [platforms-reinstall] Error 2 >> rupert at current9s ~/mgar/pkg/apr-util/trunk > > It means GAR tries to reset the amd64 install state on current9x, which > seems strange to me. Is this reproducable from a clean environment? logging in at build9s and build9x and doing: gmake reinstall remerge repackage works at least. rupert. From bwalton at opencsw.org Sun Jun 20 00:37:25 2010 From: bwalton at opencsw.org (Ben Walton) Date: Sat, 19 Jun 2010 18:37:25 -0400 Subject: [csw-maintainers] bad interactions with CSWcommon In-Reply-To: References: <1276917687-sup-6307@pinkfloyd.chass.utoronto.ca> Message-ID: <1276976909-sup-6233@pinkfloyd.chass.utoronto.ca> Excerpts from Philip Brown's message of Sat Jun 19 08:58:40 -0400 2010: > hmmm. I don't remember any more why the lnks were put in. but I > suspect the missing ones arefor locales that got added later So the next question then is whether or not having CSWcommon provide those links is correct? A cursory inspection of a few boxes seems to highlight only coreutils as being affected (broken symlink check). Looking at which packages deliver content to csw/share/.../LC_TIME, I see only CSWcommon and CSWcoreutils. This accounts for why the conflict hasn't been discovered previously. As only two packages are in play, there should be some reasonably easy fix. I'm going to read some more about this stuff before I propose anything though. Thanks -Ben -- Ben Walton Systems Programmer - CHASS University of Toronto C:416.407.5610 | W:416.978.4302 From bwalton at opencsw.org Sun Jun 20 03:18:59 2010 From: bwalton at opencsw.org (Ben Walton) Date: Sat, 19 Jun 2010 21:18:59 -0400 Subject: [csw-maintainers] bad interactions with CSWcommon In-Reply-To: References: <1276917687-sup-6307@pinkfloyd.chass.utoronto.ca> Message-ID: <1276994755-sup-5700@pinkfloyd.chass.utoronto.ca> Excerpts from Philip Brown's message of Sat Jun 19 08:58:40 -0400 2010: > hmmm. I don't remember any more why the lnks were put in. but I > suspect the missing ones arefor locales that got added later Why does CSWcommon provide anything deeper than /opt/csw/share? There may be a good reason...I'm wondering, though, if dropping locale/ (or maybe just everything below it) isn't the best solution? It's not currently delivering a complete set[2] of locales and it doesn't place any .mo files. The other option is to modify coreutils to see it deliver LC_TIME as a symlink instead of a directory containing a symlink. This feels less correct to me, even though it would be functionally equivalent in this case. Having read through the gettext docs[2], It's not inconceivable that some package out there would deliver .mo files for LC_MESSAGES and separate/distinct .mo files for LC_TIME. (It's discouraged but not invalid.) At the very least, we should come to some consensus on how packages should deliver files into /opt/csw/share/locale/ in order to have something to point at in the future if something like this occurs. Thoughts? Thanks -Ben [1] Different than what sol10 provides, augmented already by other apps. [2] http://bit.ly/9WPqjK -- Ben Walton Systems Programmer - CHASS University of Toronto C:416.407.5610 | W:416.978.4302 From phil at bolthole.com Sun Jun 20 04:12:48 2010 From: phil at bolthole.com (Philip Brown) Date: Sat, 19 Jun 2010 19:12:48 -0700 Subject: [csw-maintainers] bad interactions with CSWcommon In-Reply-To: <1276994755-sup-5700@pinkfloyd.chass.utoronto.ca> References: <1276917687-sup-6307@pinkfloyd.chass.utoronto.ca> <1276994755-sup-5700@pinkfloyd.chass.utoronto.ca> Message-ID: > At the very least, we should come to some consensus on how packages > should deliver files into /opt/csw/share/locale/ in order to have > something to point at in the future if something like this occurs. > I agree. unfortunately I have lost context of why we dis it. I do believe that back in the day, "certain things" did not work unless files were delivered in BOTH those subdirs, but many applications only delivered to one, hence the symlink. don't know if that was a sol8 specific problem or not though. test case: set locale to one of those, remove the symlink, see what breaks. on sol8 if it does, then we have our test to see if 9 is equally broken From bwalton at opencsw.org Sun Jun 20 15:27:23 2010 From: bwalton at opencsw.org (Ben Walton) Date: Sun, 20 Jun 2010 09:27:23 -0400 Subject: [csw-maintainers] bad interactions with CSWcommon In-Reply-To: References: <1276917687-sup-6307@pinkfloyd.chass.utoronto.ca> <1276994755-sup-5700@pinkfloyd.chass.utoronto.ca> Message-ID: <1277040073-sup-8018@pinkfloyd.chass.utoronto.ca> Excerpts from Philip Brown's message of Sat Jun 19 22:12:48 -0400 2010: > set locale to one of those, remove the symlink, see what breaks. on sol8 > if it does, then we have our test to see if 9 is equally broken If something were broken with a missing LC_TIME directory, that would be either a) the app or b) the gettext/libintl in play at the time. CSWgettext is much newer at this point. Unless you can recall an app that was affected, this is a needle in a haystack issue. I'd vote that the best solution is to remove the directories from CSWcommon and then fix up individual packages as required. The fix is actually very simple and won't require a rebuild. The affected packages can be unrolled and the the LC_TIME directory and LC_TIME/$domain.mo -> ../LC_MESSAGES/$domain.mo links added, the protoype updated, etc. This allows us to correct even very old packages without necessarily rebuilding them. Thoughts? Thanks -Ben -- Ben Walton Systems Programmer - CHASS University of Toronto C:416.407.5610 | W:416.978.4302 From phil at bolthole.com Sun Jun 20 16:00:21 2010 From: phil at bolthole.com (Philip Brown) Date: Sun, 20 Jun 2010 07:00:21 -0700 Subject: [csw-maintainers] bad interactions with CSWcommon In-Reply-To: <1277040073-sup-8018@pinkfloyd.chass.utoronto.ca> References: <1276917687-sup-6307@pinkfloyd.chass.utoronto.ca> <1276994755-sup-5700@pinkfloyd.chass.utoronto.ca> <1277040073-sup-8018@pinkfloyd.chass.utoronto.ca> Message-ID: cswcommon is too critical a package to just change and cross our fingers . I'd want to see a specific case found before I approve a change like this. if not, then I'd say do the weorkaround in your one specific package From bwalton at opencsw.org Sun Jun 20 16:05:55 2010 From: bwalton at opencsw.org (Ben Walton) Date: Sun, 20 Jun 2010 10:05:55 -0400 Subject: [csw-maintainers] bad interactions with CSWcommon In-Reply-To: References: <1276917687-sup-6307@pinkfloyd.chass.utoronto.ca> <1276994755-sup-5700@pinkfloyd.chass.utoronto.ca> <1277040073-sup-8018@pinkfloyd.chass.utoronto.ca> Message-ID: <1277042624-sup-3749@pinkfloyd.chass.utoronto.ca> Excerpts from Philip Brown's message of Sun Jun 20 10:00:21 -0400 2010: > cswcommon is too critical a package to just change and cross our > fingers . I'd want to see a specific case found before I approve a > change like this. if not, then I'd say do the weorkaround in your one > specific package I'm willing to test a package as you say...what I'm not willing to do is to dig through 227 packages depending on CSWggettext in order to find one that might break. You don't have any email traces of the original issue as a starting point? The current solution _is_ harmful to a known package. It _may_ be beneficial to other _broken_ packages. -Ben -- Ben Walton Systems Programmer - CHASS University of Toronto C:416.407.5610 | W:416.978.4302 From phil at bolthole.com Sun Jun 20 17:03:42 2010 From: phil at bolthole.com (Philip Brown) Date: Sun, 20 Jun 2010 08:03:42 -0700 Subject: [csw-maintainers] bad interactions with CSWcommon In-Reply-To: <1277042624-sup-3749@pinkfloyd.chass.utoronto.ca> References: <1276917687-sup-6307@pinkfloyd.chass.utoronto.ca> <1276994755-sup-5700@pinkfloyd.chass.utoronto.ca> <1277040073-sup-8018@pinkfloyd.chass.utoronto.ca> <1277042624-sup-3749@pinkfloyd.chass.utoronto.ca> Message-ID: unfortunately this is some truly ancient mojo back from the first months of csw. maaaybe Ihsan has an archive somewhere. or James. there are no public ones I just reread you original email though. it sounds to me, as though coreutils is attempting to do at an individual program level, exactly the same thing that is being done at the system level with cswcommon! in which case, cswcommon is basically doing the right thing and you should remove the coreutils symlinks From phil at bolthole.com Sun Jun 20 17:06:43 2010 From: phil at bolthole.com (Philip Brown) Date: Sun, 20 Jun 2010 08:06:43 -0700 Subject: [csw-maintainers] bad interactions with CSWcommon In-Reply-To: <1277042624-sup-3749@pinkfloyd.chass.utoronto.ca> References: <1276917687-sup-6307@pinkfloyd.chass.utoronto.ca> <1276994755-sup-5700@pinkfloyd.chass.utoronto.ca> <1277040073-sup-8018@pinkfloyd.chass.utoronto.ca> <1277042624-sup-3749@pinkfloyd.chass.utoronto.ca> Message-ID: it's interesting to note, that the example you give where coreutils does deliver two individual, separate files...(in locale/vi) does NOT have the common level symlink for that locale ! From bwalton at opencsw.org Sun Jun 20 17:11:04 2010 From: bwalton at opencsw.org (Ben Walton) Date: Sun, 20 Jun 2010 11:11:04 -0400 Subject: [csw-maintainers] bad interactions with CSWcommon In-Reply-To: References: <1276917687-sup-6307@pinkfloyd.chass.utoronto.ca> <1276994755-sup-5700@pinkfloyd.chass.utoronto.ca> <1277040073-sup-8018@pinkfloyd.chass.utoronto.ca> <1277042624-sup-3749@pinkfloyd.chass.utoronto.ca> Message-ID: <1277046453-sup-8542@pinkfloyd.chass.utoronto.ca> Excerpts from Philip Brown's message of Sun Jun 20 11:03:42 -0400 2010: > it sounds to me, as though coreutils is attempting to do at an > individual program level, exactly the same thing that is being done > at the system level with cswcommon! In this case, the outcome is the same, but it may not always be that way. Also, as cswcommon isn't providing every locale, it's doing only a partial job to circumvent something that should be fixed at the package level. > in which case, cswcommon is basically doing the right thing and you > should remove the coreutils symlinks cswcommon is only doing the right thing when LC_MESSAGES == LC_TIME. It is possible in the gettext universe for the .mo files delivered to be separate. The symlink of LC_TIME to LC_MESSAGES (while it will _mostly_ work) is incorrect. I understand that mucking with cswcommon shouldn't be taken lightly, but in this case, it's fairly clearly not the correct solution to the problem. Thanks -Ben -- Ben Walton Systems Programmer - CHASS University of Toronto C:416.407.5610 | W:416.978.4302 From bwalton at opencsw.org Sun Jun 20 17:18:21 2010 From: bwalton at opencsw.org (Ben Walton) Date: Sun, 20 Jun 2010 11:18:21 -0400 Subject: [csw-maintainers] bad interactions with CSWcommon In-Reply-To: References: <1276917687-sup-6307@pinkfloyd.chass.utoronto.ca> <1276994755-sup-5700@pinkfloyd.chass.utoronto.ca> <1277040073-sup-8018@pinkfloyd.chass.utoronto.ca> <1277042624-sup-3749@pinkfloyd.chass.utoronto.ca> Message-ID: <1277047002-sup-479@pinkfloyd.chass.utoronto.ca> Excerpts from Philip Brown's message of Sun Jun 20 11:06:43 -0400 2010: > it's interesting to note, that the example you give where coreutils > does deliver two individual, separate files...(in locale/vi) does NOT > have the common level symlink for that locale ! Huh? # grep locale/vi /var/sadm/install/contents | grep coreutils /opt/csw/share/locale/vi/LC_MESSAGES/coreutils.mo f none 0644 root bin 327678 49246 1274322262 CSWcoreutils /opt/csw/share/locale/vi/LC_TIME d none 0755 root bin CSWcoreutils /opt/csw/share/locale/vi/LC_TIME/coreutils.mo=../LC_MESSAGES/coreutils.mo s none CSWcoreutils # find . -name coreutils.mo -ls 186123 328 -rw-r--r-- 1 root bin 327678 May 19 22:24 ./LC_MESSAGES/coreutils.mo 599924 0 lrwxrwxrwx 1 root root 27 May 20 19:25 ./LC_TIME/coreutils.mo -> ../LC_MESSAGES/coreutils.mo This one doesn't get broken as cswcommon doesn't provide any locale files in vi. However, the point remains that while symlinks will work in this case as coreutils delivers all .mo data as a single file, it is _valid_ for a package to provide separate files. Thanks -Ben -- Ben Walton Systems Programmer - CHASS University of Toronto C:416.407.5610 | W:416.978.4302 From dam at opencsw.org Sun Jun 20 22:11:12 2010 From: dam at opencsw.org (Dagobert Michelsen) Date: Sun, 20 Jun 2010 22:11:12 +0200 Subject: [csw-maintainers] apr-util - searches for libiconv.la - how to use .so ? In-Reply-To: References: <1274531028-sup-9935@pinkfloyd.chass.utoronto.ca> <1276964728-sup-8610@pinkfloyd.chass.utoronto.ca> Message-ID: Hi Rupert, Am 19.06.2010 um 19:55 schrieb rupert THURNER: > On Sat, Jun 19, 2010 at 18:39, Dagobert Michelsen > wrote: >> Am 19.06.2010 um 18:30 schrieb Ben Walton: >>> Excerpts from rupert THURNER's message of Sat Jun 19 06:53:29 >>> -0400 2010: >>> 1. Rebuild CSWfreetds; it currently belongs to Alex Moore who is >>> retired, I think). This would clean up the libtool .la files it >>> provides and hopefully allow you to then use it. >> >> I'll give freetds a try. > > thanks so much ben and dago for the support! i excluded freetds until > we have the new version. I have now build an updated freetds 32/64 bit in my experimental: http://mirror.opencsw.org/experimental.html#dam It is installed on the testing build machines. Best regards -- Dago From phil at bolthole.com Mon Jun 21 08:46:14 2010 From: phil at bolthole.com (Philip Brown) Date: Sun, 20 Jun 2010 23:46:14 -0700 Subject: [csw-maintainers] interested in package tools? Message-ID: <20100621064614.GA39566@bolthole.com> btw, for those who are curious or interested in writing SVR4pkg tools.. i just learned that there is actually a LIBRARY for that sort of thing. libpkg. static on sol9, dynamic on sol10. with actual header files. such as: current9s$ grep pkg /var/sadm/install/contents|grep include /usr/include/pkgdev.h f none 0644 root bin 722 54095 1018120352 SUNWhea /usr/include/pkginfo.h f none 0644 root bin 958 6957 1018120352 SUNWhea /usr/include/pkglocs.h f none 0644 root bin 637 48042 1018120352 SUNWhea /usr/include/pkgstrct.h f none 0644 root bin 1595 56208 1244683066 SUNWhea /usr/include/pkgtrans.h f none 0644 root bin 612 44640 1018120352 SUNWhea There also allegedly is the full source code on opensolaris.org: http://hub.opensolaris.org/bin/view/Project+svr4_packaging/ claims The commands are in [svr4pkg] and the library is under [libpkg] I personally think it could be beneficial to us to package our own compilation of thse tools, but then take out artificial limitations such as "must be root to do [xyz]", and just rely on filesystem perms to determine that sort of thing. From skayser at opencsw.org Mon Jun 21 12:16:16 2010 From: skayser at opencsw.org (Sebastian Kayser) Date: Mon, 21 Jun 2010 12:16:16 +0200 Subject: [csw-maintainers] PHP + GD + JPEG broken -> Mantis captchas deactivated (was: Re: website issues) In-Reply-To: <20100616130513.GC31767@sebastiankayser.de> References: <20100616103604.GB31767@sebastiankayser.de> <20100616130513.GC31767@sebastiankayser.de> Message-ID: <4C1F3BF0.5060105@opencsw.org> Sebastian Kayser wrote on 16.06.2010 15:05: > 5. GD jpeg handling and thereby Mantis captcha generation for signups is broken > https://www.opencsw.org/mantis/signup_page.php > https://www.opencsw.org/mantis/make_captcha_img.php?public_key=93738 > > Apache error log holds: > gd-jpeg: JPEG library reports unrecoverable error: Wrong JPEG library > version: library is 62, caller expects 70 PHP+GD+JPEG is still broken on the www box (and likely in current/), thus I've deactivated Mantis signup captchas for now. Looking at the libs, the PHP gd module was built against libjpeg.so.62 while libgd itself is built against libjpeg.so.7. The gd module pulls in 62 first and gd complains because it expects 70. $ ldd /opt/csw/php5/lib/php/extensions/no-debug-non-zts-20060613/gd.so | grep jpeg libjpeg.so.62 => /opt/csw/lib/libjpeg.so.62 libjpeg.so.7 => /opt/csw/lib/sparcv8/libjpeg.so.7 $ ldd -s /opt/csw/php5/lib/php/extensions/no-debug-non-zts-20060613/gd.so | grep object=libjpeg find object=libjpeg.so.62; required by /opt/csw/php5/lib/php/extensions/no-debug-non-zts-20060613/gd.so find object=libjpeg.so.7; required by /opt/csw/lib/libgd.so.2 Is there anything short of a PHP rebuild that we can do about it? Sebastian From skayser at opencsw.org Mon Jun 21 12:27:00 2010 From: skayser at opencsw.org (Sebastian Kayser) Date: Mon, 21 Jun 2010 12:27:00 +0200 Subject: [csw-maintainers] www.opencsw.org: PHP + sendmail broken (/opt/csw/sbin/sendmail missing) Message-ID: <4C1F3E74.7020308@opencsw.org> Hi, the new www box doesn't send out Mantis emails and is either missing CSW sendmail or needs a PHP config adjustment. Could someone with admin right take care of it please? $ tail /var/log/apache/error_log sh: /opt/csw/sbin/sendmail: not found sh: /opt/csw/sbin/sendmail: not found sh: /opt/csw/sbin/sendmail: not found sh: /opt/csw/sbin/sendmail: not found sh: /opt/csw/sbin/sendmail: not found sh: /opt/csw/sbin/sendmail: not found sh: /opt/csw/sbin/sendmail: not found sh: /opt/csw/sbin/sendmail: not found sh: /opt/csw/sbin/sendmail: not found Use of uninitialized value in concatenation (.) or string at /var/www/www.opencsw.org/cgi-buglist/maintainer.cgi line 10. $ grep sendmail_path /opt/csw/php5/lib/php.ini sendmail_path = /opt/csw/sbin/sendmail -t -i $ file /opt/csw/sbin/sendmail /opt/csw/sbin/sendmail: cannot open: No such file or directory Mantis itself doesn't report those errors :/ thus I can't exactly say since when the bug notification emails are missing. The last one on the bug notifications list is from the 13th of June, so that's a likely bet. Anyone who pinged others via bug notes in the meantime needs to re-ping them once the mail setup is fixed. Sebastian From maciej at opencsw.org Mon Jun 21 13:40:40 2010 From: maciej at opencsw.org (Maciej (Matchek) Blizinski) Date: Mon, 21 Jun 2010 12:40:40 +0100 Subject: [csw-maintainers] cswclassutils: location of template init scripts In-Reply-To: References: Message-ID: No dia 16 de Junho de 2010 18:38, Philip Brown escreveu: > On Wed, Jun 16, 2010 at 10:26 AM, Maciej (Matchek) Blizinski > wrote: >> ... ?The definition of CAS >> is that they handle the installation of files, with potential >> additional processing. ?Having CAS process other files is something >> that CAS isn't intended for. ?I know that you can hack it, but it >> might cross the fine line between use and abuse. > > I disagree. Sun has used it in this way for years. almost a decade? one of the > "standard, documented(but often ignored)" ways, used it in this sort of way. > the cron, sed, or awk class I think. dont remember which. There are i.awk and i.sed action scripts, but none of them processes files or locations outside the given list. Your two above paragraphs sound as if you thought that I'm questioning the use of classutils, which I'm not. I'm saying that classutils are designed to process given lists of files. Even though you could write a CAS that processes additional files under certain circumstances (empty input), I think this is a bad idea, because: 1. It relies on undocumented behavior which might change (CAS being called with empty input vs CAS not being called at all) 2. It makes the whole solution harder to debug > by definition, the ?sun cron class is an example of, "take one file, > and use a CAS to modify another file (crontab)" Yes, and let CAS stay this way. >> Perhaps CAS are not the right way to solve this problem? ?If you just >> want to run a script upon package installation, this is what pre- and >> postinstall are for. > > but "just run a script" is not the full design goal. the full design goal is, > "and not bug users about prompts for 'do you wish to allow this > unknown script to run?' for common , known, shared framework scripts, > while at the same time STILL prompting (if the user desires) for truly > unknown one-shot scripts" > > The CAS scripts ?may eventually be shared by hundreds of our packages. > The site admin should be able to evaluate the safety of the script(s) > one time, and then have trust in them. > The CAS approach allows this. The regular postinstall script approach does not. The way of shipping the script is a separate issue from script contents. The trust you're bringing up has to do with the script contents, while the discussion above is about the way of shipping. >>> I also dont understand the purpose of the column labeled, >>> "No modification to cswinitsmf (no bugs)" >> >> The purpose is similar to other criterion columns: point out any flaws >> in described approaches. ?What is the specific issue with this column? > > > let me rephrase: I dont understand the column :) There are 2 issues in 2 columns: one is whether you need to modify the script at all (and potentially introduce bugs). The other is whether the introduced change will make the script harder to debug. From dam at opencsw.org Mon Jun 21 15:10:28 2010 From: dam at opencsw.org (Dagobert Michelsen) Date: Mon, 21 Jun 2010 15:10:28 +0200 Subject: [csw-maintainers] interested in package tools? In-Reply-To: <20100621064614.GA39566@bolthole.com> References: <20100621064614.GA39566@bolthole.com> Message-ID: <25C93BF8-4D38-4796-9BF4-46AEF8EA0499@opencsw.org> Hi Phil, Am 21.06.2010 um 08:46 schrieb Philip Brown: > btw, for those who are curious or interested in writing SVR4pkg > tools.. > i just learned that there is actually a LIBRARY for that sort of > thing. > libpkg. > static on sol9, dynamic on sol10. > > with actual header files. such as: > current9s$ grep pkg /var/sadm/install/contents|grep include > /usr/include/pkgdev.h f none 0644 root bin 722 54095 1018120352 > SUNWhea > /usr/include/pkginfo.h f none 0644 root bin 958 6957 1018120352 > SUNWhea > /usr/include/pkglocs.h f none 0644 root bin 637 48042 1018120352 > SUNWhea > /usr/include/pkgstrct.h f none 0644 root bin 1595 56208 1244683066 > SUNWhea > /usr/include/pkgtrans.h f none 0644 root bin 612 44640 1018120352 > SUNWhea > > There also allegedly is the full source code on opensolaris.org: > http://hub.opensolaris.org/bin/view/Project+svr4_packaging/ claims > > The commands are in [svr4pkg] and the library is under [libpkg] > > I personally think it could be beneficial to us to package our own > compilation of thse tools, but then take out artificial limitations > such as > "must be root to do [xyz]", and just rely on filesystem perms to > determine > that sort of thing. What tools do you have in mind? There is also the heirloom project which ported the SYSV packaging tools to Linux from OpenSolaris. But what specifically would we gain from it? Best regards -- Dago From phil at bolthole.com Mon Jun 21 18:05:09 2010 From: phil at bolthole.com (Philip Brown) Date: Mon, 21 Jun 2010 09:05:09 -0700 Subject: [csw-maintainers] interested in package tools? In-Reply-To: <25C93BF8-4D38-4796-9BF4-46AEF8EA0499@opencsw.org> References: <20100621064614.GA39566@bolthole.com> <25C93BF8-4D38-4796-9BF4-46AEF8EA0499@opencsw.org> Message-ID: On Mon, Jun 21, 2010 at 6:10 AM, Dagobert Michelsen wrote: > Hi Phil, > > Am 21.06.2010 um 08:46 schrieb Philip Brown: > >> I personally think it could be beneficial to us to package our own >> compilation of thse tools, but then take out artificial limitations such >> as >> "must be root to do [xyz]", and just rely on filesystem perms to determine >> that sort of thing. > > What tools do you have in mind?.... But what > specifically would we gain from it? > I can think of two things off the top of my head: 1. I could take the internal pkgtrans routines out of pkg-get. They are there because of a stupid bug in pkgtrans. if we have our own, we could fix the bug, and I could ditch that hack 2. It allows for non-root people to keep their own csw repository. This is one of the "IPS is better" gripes: non-root people can supposedly use it. So this would address a documented user complaint. (well, okay, SOME things are hardcoded to reference /opt/csw paths. but a lot of things will work fine without it; the most common issue is really an RPATH issue, which can be solved via LD_LIBRARY_PATH.) From phil at bolthole.com Mon Jun 21 18:14:55 2010 From: phil at bolthole.com (Philip Brown) Date: Mon, 21 Jun 2010 09:14:55 -0700 Subject: [csw-maintainers] PHP + GD + JPEG broken -> Mantis captchas deactivated (was: Re: website issues) In-Reply-To: <4C1F3BF0.5060105@opencsw.org> References: <20100616103604.GB31767@sebastiankayser.de> <20100616130513.GC31767@sebastiankayser.de> <4C1F3BF0.5060105@opencsw.org> Message-ID: On Mon, Jun 21, 2010 at 3:16 AM, Sebastian Kayser wrote: > Sebastian Kayser wrote on 16.06.2010 15:05: >> > PHP+GD+JPEG is still broken on the www box (and likely in current/), thus > I've deactivated Mantis signup captchas for now. Looking at the libs, the > PHP gd module was built against libjpeg.so.62 while libgd itself is built > against libjpeg.so.7. The gd module pulls in 62 first and gd complains > because it expects 70. > > $ ldd /opt/csw/php5/lib/php/extensions/no-debug-non-zts-20060613/gd.so | grep jpeg > ? ? ? ?libjpeg.so.62 => ? ? ? ? /opt/csw/lib/libjpeg.so.62 > ? ? ? ?libjpeg.so.7 => ?/opt/csw/lib/sparcv8/libjpeg.so.7 > > > Is there anything short of a PHP rebuild that we can do about it? > sure... IFF the libraries are backwords compatible, force load of 7. try it out with LD_PRELOAD, or force the .62 symlink to point to 7, temporarily? Two comments though: 1. if it is nicely backwards compatible, i would think we wouldnt need the two separate versions 2. we survived without captcha just fine for years. leave it off :) but that being said... we do need a php rebuild.... From bwalton at opencsw.org Tue Jun 22 01:47:19 2010 From: bwalton at opencsw.org (Ben Walton) Date: Mon, 21 Jun 2010 19:47:19 -0400 Subject: [csw-maintainers] bad interactions with CSWcommon In-Reply-To: <1277047002-sup-479@pinkfloyd.chass.utoronto.ca> References: <1276917687-sup-6307@pinkfloyd.chass.utoronto.ca> <1276994755-sup-5700@pinkfloyd.chass.utoronto.ca> <1277040073-sup-8018@pinkfloyd.chass.utoronto.ca> <1277042624-sup-3749@pinkfloyd.chass.utoronto.ca> <1277047002-sup-479@pinkfloyd.chass.utoronto.ca> Message-ID: <1277163839-sup-3934@pinkfloyd.chass.utoronto.ca> Excerpts from Ben Walton's message of Sun Jun 20 11:18:21 -0400 2010: > Excerpts from Philip Brown's message of Sun Jun 20 11:06:43 -0400 2010: > > it's interesting to note, that the example you give where coreutils > > does deliver two individual, separate files...(in locale/vi) does NOT > > have the common level symlink for that locale ! Sorry, misread this yesterday. This is correct. The locale/vi files for CSWcoreutils actually work, since they're not affected by the 'global' workaround provided in CSWcommon. CSWcommon is 'fixing' things the wrong way, by papering over bugs in other packages[1]...it's cancelling an intended feature of gettext/libintl and doing only half the job at that. Let's fix this _properly_ instead of adding another layer of papier macher. Thanks -Ben [1] James/Ihsan: Do either of you have more info on this in your mail archives? -- Ben Walton Systems Programmer - CHASS University of Toronto C:416.407.5610 | W:416.978.4302 From phil at bolthole.com Tue Jun 22 02:53:53 2010 From: phil at bolthole.com (Philip Brown) Date: Mon, 21 Jun 2010 17:53:53 -0700 Subject: [csw-maintainers] users of fbpanel? Message-ID: BTW: I've just compiled fbpanel. After some initial problems with my own dumb setup of gtk2, I now have it mostly running. and I like it! fbpanel + openbox seems like a nice simple friendly way to go. The trouble is... It could use some fine tuning. fbpanel could use a little fix with a missing icon, and it would be nice to have an actual "session" or something, rather than having to start up things by hand right now. The "ultimate" fix would be running lxde. But even that is a bit overkill. Does anyone currently run fbpanel and/or openbox, and would be interested in contributing to this effort? From jeff at cjsa.com Tue Jun 22 19:09:03 2010 From: jeff at cjsa.com (Jeffery Small) Date: Tue, 22 Jun 2010 17:09:03 GMT Subject: [csw-maintainers] Problem with gam_server and gimp Message-ID: I recently upgraded in the "current" tree on my SPARC Solaris 10 system. This included the new gtk2 packages. I do not recall if the gamin package was updated recently, but it is at the current 0.1.9,REV=2010.06.15 rev. level. I am now experiencing a serious problem when I start up gimp 2.4.3. With gimp sitting idle, I am seeing a huge CPU load my the gimp and the gam_server processes. After sitting quiet overnight, they were burning 40-50% of the CPU this morning, and immediately after restarting gimp, I see similar load number in the range of 32% for gam_server and 42% for gimp. Prior to the CSW upgrade I was having no problems. The next closest process load is firefox-bin with around 1.25%. Since nothing in gimp has changed, I have to assume the problem is in the gam_server or is an artifact of the new gtk2 package. By the way, visiting the CSW site shows the new web pages. However the file search feature does not appear to be indexed into the search field on these pages. I was only able to do a file search on the pages still at www.canoedissent.org.uk. This is a facility I often use and it needs to be integrated into the main pages - or made easier to find if it is already there and I missed it. Regards, -- Jeff C. Jeffery Small CJSA LLC 206-232-3338 jeff at cjsa.com 7000 E Mercer Way, Mercer Island, WA 98040 From jeff at cjsa.com Tue Jun 22 19:22:57 2010 From: jeff at cjsa.com (Jeffery Small) Date: Tue, 22 Jun 2010 17:22:57 GMT Subject: [csw-maintainers] Problem with gam_server and gimp Message-ID: Following up to my own post, I just realized that the fam package was also recently upgraded, and this may also be a source for the new troubles. Other people on the internet have reported similar problems with high gam_server loads. Here is a link to a discussion which may be of some use. Personally, I'm completely unfamiliar with the operation of any of these packages, so I have no immediated suggestions to offer until I do more investigation. http://www.alekz.net/archives/16 There does not appear to be a default gaminrc configuration file installed here in any of the etc directories. Regards, -- Jeff C. Jeffery Small CJSA LLC 206-232-3338 jeff at cjsa.com 7000 E Mercer Way, Mercer Island, WA 98040 From phil at bolthole.com Tue Jun 22 19:55:34 2010 From: phil at bolthole.com (Philip Brown) Date: Tue, 22 Jun 2010 10:55:34 -0700 Subject: [csw-maintainers] Problem with gam_server and gimp In-Reply-To: References: Message-ID: On Tue, Jun 22, 2010 at 10:22 AM, Jeffery Small wrote: > Following up to my own post, I just realized that the fam package was also > recently upgraded, and this may also be a source for the new troubles. > > Other people on the internet have reported similar problems with high > gam_server loads. ?Here is a link to a discussion which may be of some use. > Personally, I'm completely unfamiliar with the operation of any of these > packages, so I have no immediated suggestions to offer until I do more > investigation. > > http://www.alekz.net/archives/16 > > There does not appear to be a default gaminrc configuration file installed > here in any of the etc directories. > Errm... Jeff.. you ARE still on the maintainers list, right? I just posted to the list, a coupla days ago, about this very problem, and asked for suggestions :-} btw, if you DO find some kind of gaminrc that helps this situation, please let me know and i'd be happy to update gamin with it. From phil at bolthole.com Tue Jun 22 20:16:43 2010 From: phil at bolthole.com (Philip Brown) Date: Tue, 22 Jun 2010 11:16:43 -0700 Subject: [csw-maintainers] desktop team, looking for members Message-ID: Hi folks, we are potentially now revving up our packages to the level where we're going to have "desktop issues". That is to say, "issues", in the sense of "what should our strategy be in area x, y, and z?" These are issues that dont neccessarily have a "right" or "wrong" answer... they just need to be consistent. This sort of thing benefits from having a person, or persons, who have a long-term commitment or interest in making sure things stay sane. I would personally NOT like to be heavily involved in this. Unfortunately, with the path I'm treading on, (packaging wise), I'm running into things where I'm having to make decisions on this sort of thing, but I'd much rather someone else be making the decisions :-} Anyone interested? We can have more than one person involved (in fact, I HOPE we will have more). Qualifications are minimal: 1. Must have an interest in having OpenCSW desktop visible stuff look nice 2. Must have an interest, or at least a commitment, to doing a little research when neccessary. In other words, prior experience not required... but a willingness to OBTAIN experience, is ;-) And you do NOT have to do all the research yourself. it is perfectly valid to ask for help, etc. Any takers? Please? Pretty-please? A reminder: This is not a position "to package all desktop programs". I'm thinking more like a simple "desktop theme and design committee". You may end up being a "committee of 1". If you're okay with that, all the better. From phil at bolthole.com Tue Jun 22 20:34:51 2010 From: phil at bolthole.com (Philip Brown) Date: Tue, 22 Jun 2010 11:34:51 -0700 Subject: [csw-maintainers] cswclassutils: location of template init scripts In-Reply-To: References: Message-ID: Maciej, you seem to wish to go down the pedantic path. So okay, lets get pedantic :-) On Mon, Jun 21, 2010 at 4:40 AM, Maciej (Matchek) Blizinski wrote: > No dia 16 de Junho de 2010 18:38, Philip Brown escreveu: >> On Wed, Jun 16, 2010 at 10:26 AM, Maciej (Matchek) Blizinski >> wrote: >>> ... ?The definition of CAS >>> is that they handle the installation of files, with potential >>> additional processing. ?Having CAS process other files is something >>> that CAS isn't intended for. ?I know that you can hack it, but it >>> might cross the fine line between use and abuse. >> >>.... > > Your two above paragraphs sound as if you thought that I'm questioning > the use of classutils, which I'm not. ?I'm saying that classutils are > designed to process given lists of files. http://docs.sun.com/app/docs/doc/806-7008/6jftmsc38?a=view "Object classes allow a series of actions to be performed on a group of package objects at installation or removal." This is ambiguous. It doesnt say that renaming the objects, or duplicating, or removing, objects, is either explicitly allowed, or explicitly denied. It merely says "actions", on "objects" with the only limitation that the objects be associated with the package. My view of this is: Since the entire purpose of system standard utilities such as "installf" and "removef", are to DYNAMICALLY create and remove files associated with the package.... files that, **by definition**, are not in the initial prototype/pkgmap file... then the ambiguity is strongly resolved in favor of "it is allowed.". > ?Even though you could write > a CAS that processes additional files under certain circumstances > (empty input), I think this is a bad idea, because: > > 1. It relies on undocumented behavior which might change (CAS being > called with empty input vs CAS not being called at all) Actually, it is *explicitly documented* behaviour. again, from the above url: "Note ? Even if there are no regular files of this class anywhere in the package, the class action script will be called at least once with an empty list and the ENDOFCLASS argument. " From jeff at cjsa.com Tue Jun 22 22:14:21 2010 From: jeff at cjsa.com (Jeffery Small) Date: Tue, 22 Jun 2010 20:14:21 GMT Subject: [csw-maintainers] Problem with gam_server and gimp References: Message-ID: Philip Brown writes: >On Tue, Jun 22, 2010 at 10:22 AM, Jeffery Small wrote: >> Following up to my own post, I just realized that the fam package was also >> recently upgraded, and this may also be a source for the new troubles. >> Other people on the internet have reported similar problems with high >> gam_server loads. >Errm... Jeff.. you ARE still on the maintainers list, right? >I just posted to the list, a coupla days ago, about this very problem, >and asked for suggestions :-} Sorry, I was away for a long weekend and probably missed this discussion as I plowed through many days worth of messages. >btw, if you DO find some kind of gaminrc that helps this situation, >please let me know and i'd be happy to update gamin with it. I'll report any positive information I uncover. Can the maintainer of the gam_server package verify the proper location of the gaminrc file. I would have assumed /opt/etc/gaminrc or /opt/etc/gamin/gaminrc, but using strings on the binary only reveals /etc/gamin/gaminrc, /etc/gamin/mandatory_gaminrc and /.gminrc. Regards, -- Jeff C. Jeffery Small CJSA LLC 206-232-3338 jeff at cjsa.com 7000 E Mercer Way, Mercer Island, WA 98040 From phil at bolthole.com Tue Jun 22 23:15:00 2010 From: phil at bolthole.com (Philip Brown) Date: Tue, 22 Jun 2010 14:15:00 -0700 Subject: [csw-maintainers] Problem with gam_server and gimp In-Reply-To: References: Message-ID: On Tuesday, June 22, 2010, Jeffery Small wrote: > > > I'll report any positive information I uncover. ?Can the maintainer of the > gam_server package verify the proper location of the gaminrc file. ?I would > have assumed /opt/etc/gaminrc or /opt/etc/gamin/gaminrc, but using strings > on the binary only reveals /etc/gamin/gaminrc, /etc/gamin/mandatory_gaminrc > and /.gminrc. that would be me. it was a quick hack. tell me what works and I'll update it From phil at bolthole.com Tue Jun 22 23:50:25 2010 From: phil at bolthole.com (Philip Brown) Date: Tue, 22 Jun 2010 14:50:25 -0700 Subject: [csw-maintainers] subversion and binaries Message-ID: Does subversion support binaries? Can I dump a binary in our svn repository, and expect that it NOT get tweaked with any kind of magic expansion? is there something I need to do, to ensure no expansion or substitution? From bwalton at opencsw.org Wed Jun 23 01:49:06 2010 From: bwalton at opencsw.org (Ben Walton) Date: Tue, 22 Jun 2010 19:49:06 -0400 Subject: [csw-maintainers] subversion and binaries In-Reply-To: References: Message-ID: <1277250480-sup-5709@pinkfloyd.chass.utoronto.ca> Excerpts from Philip Brown's message of Tue Jun 22 17:50:25 -0400 2010: > Does subversion support binaries? It should handle this just fine. I'm fairly certain there is already a recipe somewhere with a legacy .so file being carried. HTH. -Ben -- Ben Walton Systems Programmer - CHASS University of Toronto C:416.407.5610 | W:416.978.4302 From bwalton at opencsw.org Wed Jun 23 02:12:07 2010 From: bwalton at opencsw.org (Ben Walton) Date: Tue, 22 Jun 2010 20:12:07 -0400 Subject: [csw-maintainers] problems with symbol version script Message-ID: <1277251662-sup-5229@pinkfloyd.chass.utoronto.ca> Hi All, I'm trying to build an updated libxslt which has, since the last release, added a symbol version script. The build fails with: libtool: link: /opt/studio/SOS12/SUNWspro/bin/cc -G -h libxslt.so.1 -o .libs/libxslt.so.1.1.26 .libs/attrvt.o .libs/xslt.o .libs/xsltlocale.o .libs/xsltutils.o .libs/pattern.o .libs/templates.o .libs/variables.o .libs/keys.o .libs/numbers.o .libs/extensions.o .libs/extra.o .libs/functions.o .libs/namespaces.o .libs/imports.o .libs/attributes.o .libs/documents.o .libs/preproc.o .libs/transform.o .libs/security.o -R/opt/csw/lib -L/opt/csw/lib -lxml2 -lz -lpthread -liconv -lsocket -lnsl -lm -lc -m32 -xarch=v8 -Wl,-M -Wl,./libxslt.syms -m32 -xarch=v8 Undefined first referenced symbol in file _fini /opt/studio/SOS12/SUNWspro/prod/lib/crti.o (symbol has no version assigned) _init /opt/studio/SOS12/SUNWspro/prod/lib/crti.o (symbol has no version assigned) xsltPointerListFree ./libxslt.syms xsltCopyTree .libs/transform.o (symbol has no version assigned) xsltTransStorageAdd ./libxslt.syms xsltPointerListCreate ./libxslt.syms xsltPointerListAddSize ./libxslt.syms xsltParseSequenceConstructor ./libxslt.syms xsltComputingGlobalVarMarker .libs/variables.o (symbol has no version assigned) xsltPointerListClear ./libxslt.syms xsltParseAnyXSLTElem ./libxslt.syms xsltRestoreDocumentNamespaces ./libxslt.syms xsltXSLTAttrMarker ./libxslt.syms xsltTransStorageRemove ./libxslt.syms _lib_version /opt/studio/SOS12/SUNWspro/prod/lib/values-xa.o (symbol has no version assigned) xsltDefaultTrace .libs/transform.o (symbol has no version assigned) xsltStyleStylesheetLevelGetExtData ./libxslt.syms xsltConstNamespaceNameXSLT ./libxslt.syms xsltMatchPattern ./libxslt.syms ld: fatal: Symbol referencing errors. No output written to .libs/libxslt.so.1.1.26 gmake: *** [libxslt.la] Error 1 So far, I've been able to make the build work by either removing the -Wl,-M -Wl,-./libxslt.syms parameters, thus disabling the symbol versioning completely or by added -Wl,-B -Wl,local to the arguments, thus changing the intended behaviour of the symbol map (all symbols in the map are defined in the global scope). The symbol version script looks ok[1] to me. Has anyone hit this before? I'd prefer to enable the functionality if possible since that makes for better .so files... Any and all help is appreciated. Thanks -Ben [1] I'll be submitting a patch upstream to correct the names...looks like a copy error from libxml2. They build the .syms file from an xml document and an xslt transform file where the transform file contains the LIBXML2 tags presumably as they were pulled in from that twin project. -- Ben Walton Systems Programmer - CHASS University of Toronto C:416.407.5610 | W:416.978.4302 From phil at bolthole.com Wed Jun 23 02:55:23 2010 From: phil at bolthole.com (Philip Brown) Date: Tue, 22 Jun 2010 17:55:23 -0700 Subject: [csw-maintainers] problems with symbol version script In-Reply-To: <1277251662-sup-5229@pinkfloyd.chass.utoronto.ca> References: <1277251662-sup-5229@pinkfloyd.chass.utoronto.ca> Message-ID: that's odd. stuff like fini and init are very basic cprogram hooks. that's sort of equivalent to saying "no main() found" it shouldn't be linking CRT.o I think From bwalton at opencsw.org Wed Jun 23 03:04:06 2010 From: bwalton at opencsw.org (Ben Walton) Date: Tue, 22 Jun 2010 21:04:06 -0400 Subject: [csw-maintainers] problems with symbol version script In-Reply-To: References: <1277251662-sup-5229@pinkfloyd.chass.utoronto.ca> Message-ID: <1277254959-sup-1559@pinkfloyd.chass.utoronto.ca> Excerpts from Philip Brown's message of Tue Jun 22 20:55:23 -0400 2010: > that's odd. stuff like fini and init are very basic cprogram hooks. > that's sort of equivalent to saying "no main() found" I think those ones are just warnings (non-versioned symbols) whereas the others are hard errors...emphasis on _I think_. I think studio itself is adding crti. Thanks -Ben -- Ben Walton Systems Programmer - CHASS University of Toronto C:416.407.5610 | W:416.978.4302 From phil at bolthole.com Wed Jun 23 05:53:11 2010 From: phil at bolthole.com (Philip Brown) Date: Tue, 22 Jun 2010 20:53:11 -0700 Subject: [csw-maintainers] problems with symbol version script In-Reply-To: <1277254959-sup-1559@pinkfloyd.chass.utoronto.ca> References: <1277251662-sup-5229@pinkfloyd.chass.utoronto.ca> <1277254959-sup-1559@pinkfloyd.chass.utoronto.ca> Message-ID: On Tuesday, June 22, 2010, Ben Walton wrote: > Excerpts from Philip Brown's message of Tue Jun 22 20:55:23 -0400 2010: > >> that's odd. stuff like fini and init are very basic cprogram hooks. >> that's sort of equivalent to saying "no main() found" > > I think those ones are just warnings (non-versioned symbols) whereas > the others are hard errors...emphasis on _I think_. ?I think studio > itself is adding crti. > the thing is, it shouldn't be. that's not for use by shared libraries AFAIK From maciej at opencsw.org Wed Jun 23 11:04:01 2010 From: maciej at opencsw.org (Maciej (Matchek) Blizinski) Date: Wed, 23 Jun 2010 10:04:01 +0100 Subject: [csw-maintainers] desktop team, looking for members In-Reply-To: References: Message-ID: No dia 22 de Junho de 2010 19:16, Philip Brown escreveu: > Anyone interested? I'm not in, sorry. I have been using Solaris 10 desktop for educational purposes for some time. It was when I was contributing X11-related packages. I've moved away from it since, hence no recent X11 updates from me. I've decided I'll focus on checkpkg and the handful of server packages I currently maintain. From ihsan at opencsw.org Wed Jun 23 13:35:59 2010 From: ihsan at opencsw.org (Ihsan Dogan) Date: Wed, 23 Jun 2010 14:35:59 +0300 Subject: [csw-maintainers] website issues In-Reply-To: <20100616130513.GC31767@sebastiankayser.de> References: <20100616103604.GB31767@sebastiankayser.de> <20100616130513.GC31767@sebastiankayser.de> Message-ID: <4C21F19F.9070804@opencsw.org> Hello, Am 16.06.10 16:05, schrieb Sebastian Kayser: > 5. GD jpeg handling and thereby Mantis captcha generation for signups is broken > https://www.opencsw.org/mantis/signup_page.php > https://www.opencsw.org/mantis/make_captcha_img.php?public_key=93738 > > Apache error log holds: > gd-jpeg: JPEG library reports unrecoverable error: Wrong JPEG library > version: library is 62, caller expects 70 Which PHP modules are needed for Mantis? Ihsan -- ihsan at dogan.ch http://blog.dogan.ch/ From bwalton at opencsw.org Wed Jun 23 14:30:01 2010 From: bwalton at opencsw.org (Ben Walton) Date: Wed, 23 Jun 2010 08:30:01 -0400 Subject: [csw-maintainers] problems with symbol version script In-Reply-To: References: <1277251662-sup-5229@pinkfloyd.chass.utoronto.ca> <1277254959-sup-1559@pinkfloyd.chass.utoronto.ca> Message-ID: <1277296175-sup-9822@pinkfloyd.chass.utoronto.ca> Excerpts from Philip Brown's message of Tue Jun 22 23:53:11 -0400 2010: > the thing is, it shouldn't be. that's not for use by shared > libraries AFAIK Ok, that's interesting. I'll start digging around in that area then. Thanks for the pointer. -Ben -- Ben Walton Systems Programmer - CHASS University of Toronto C:416.407.5610 | W:416.978.4302 From dam at opencsw.org Thu Jun 24 09:21:59 2010 From: dam at opencsw.org (Dagobert Michelsen) Date: Thu, 24 Jun 2010 09:21:59 +0200 Subject: [csw-maintainers] Website errors Message-ID: Hi, I just wanted to take a look at my buglist-scripts as they didn't seem to be found at https://www.opencsw.org/buglist/maintainer.cgi?maintainer='dam' but when I looked there are massive errors: > [Thu Jun 24 09:19:26 2010] [error] [client 86.103.232.250] FastCGI: > server "/var/www/fcgi-bin/php5-fcgi-starter" stderr: PHP Warning: > mysql_error(): 5 is not a valid MySQL-Link resource in /var/www/www.opencsw.org/htdocs/wp-includes/wp-db.php > on line 715 > [Thu Jun 24 09:19:26 2010] [error] [client 86.103.232.250] FastCGI: > server "/var/www/fcgi-bin/php5-fcgi-starter" stderr: PHP Warning: > mysql_error(): 5 is not a valid MySQL-Link resource in /var/www/www.opencsw.org/htdocs/wp-includes/wp-db.php > on line 719 > [Thu Jun 24 09:19:29 2010] [error] [client 79.211.83.244] FastCGI: > server "/var/www/fcgi-bin/php5-fcgi-starter" stderr: PHP Warning: > mysql_error(): 5 is not a valid MySQL-Link resource in /var/www/www.opencsw.org/htdocs/wp-includes/wp-db.php > on line 715, referer: http://www.opencsw.org/about/ > [Thu Jun 24 09:19:29 2010] [error] [client 79.211.83.244] FastCGI: > server "/var/www/fcgi-bin/php5-fcgi-starter" stderr: PHP Warning: > mysql_error(): 5 is not a valid MySQL-Link resource in /var/www/www.opencsw.org/htdocs/wp-includes/wp-db.php > on line 719, referer: http://www.opencsw.org/about/ > [Thu Jun 24 09:19:35 2010] [error] [client 77.242.201.52] FastCGI: > server "/var/www/fcgi-bin/php5-fcgi-starter" stderr: PHP Warning: > mysql_error(): 5 is not a valid MySQL-Link resource in /var/www/www.opencsw.org/htdocs/wp-includes/wp-db.php > on line 715 > [Thu Jun 24 09:19:35 2010] [error] [client 77.242.201.52] FastCGI: > server "/var/www/fcgi-bin/php5-fcgi-starter" stderr: PHP Warning: > mysql_error(): 5 is not a valid MySQL-Link resource in /var/www/www.opencsw.org/htdocs/wp-includes/wp-db.php > on line 719 > [Thu Jun 24 09:19:39 2010] [error] [client 79.211.83.244] FastCGI: > server "/var/www/fcgi-bin/php5-fcgi-starter" stderr: PHP Warning: > mysql_error(): 5 is not a valid MySQL-Link resource in /var/www/www.opencsw.org/htdocs/wp-includes/wp-db.php > on line 715, referer: http://www.opencsw.org/about/core-principles/ > [Thu Jun 24 09:19:39 2010] [error] [client 79.211.83.244] FastCGI: > server "/var/www/fcgi-bin/php5-fcgi-starter" stderr: PHP Warning: > mysql_error(): 5 is not a valid MySQL-Link resource in /var/www/www.opencsw.org/htdocs/wp-includes/wp-db.php > on line 719, referer: http://www.opencsw.org/about/core-principles/ William, Ihsan, Sebastian: Could please someone take a look at this issue? Best regards -- Dago From bwalton at opencsw.org Thu Jun 24 14:59:12 2010 From: bwalton at opencsw.org (Ben Walton) Date: Thu, 24 Jun 2010 08:59:12 -0400 Subject: [csw-maintainers] addition to CSWcommon Message-ID: <1277384284-sup-8472@pinkfloyd.chass.utoronto.ca> Hi Phil, I think CSWcommon should include a directory entry for /opt/csw/gnu in the prototype. This will work around the bug discovered in CSWgawk last night where it doesn't provide that directory but does provide a symlink inside it. It's only triggered when no other packages on the system deliver some gnulinks first. Thoughts? Thanks -Ben -- Ben Walton Systems Programmer - CHASS University of Toronto C:416.407.5610 | W:416.978.4302 From dam at opencsw.org Thu Jun 24 15:13:10 2010 From: dam at opencsw.org (Dagobert Michelsen) Date: Thu, 24 Jun 2010 15:13:10 +0200 Subject: [csw-maintainers] addition to CSWcommon In-Reply-To: <1277384284-sup-8472@pinkfloyd.chass.utoronto.ca> References: <1277384284-sup-8472@pinkfloyd.chass.utoronto.ca> Message-ID: <9AF254A4-68FF-447A-AA43-FF090C2FA375@opencsw.org> Hi Ben, Am 24.06.2010 um 14:59 schrieb Ben Walton: > I think CSWcommon should include a directory entry for /opt/csw/gnu > in the prototype. This will work around the bug discovered in CSWgawk > last night where it doesn't provide that directory but does provide a > symlink inside it. It's only triggered when no other packages on the > system deliver some gnulinks first. > > Thoughts? Is this really necessary? What's wrong if the packages bring the directory themselves? Besides that it is possible to forget the directory if you are not using GAR ;-) Best regards -- Dago From bwalton at opencsw.org Thu Jun 24 15:26:33 2010 From: bwalton at opencsw.org (Ben Walton) Date: Thu, 24 Jun 2010 09:26:33 -0400 Subject: [csw-maintainers] addition to CSWcommon In-Reply-To: <9AF254A4-68FF-447A-AA43-FF090C2FA375@opencsw.org> References: <1277384284-sup-8472@pinkfloyd.chass.utoronto.ca> <9AF254A4-68FF-447A-AA43-FF090C2FA375@opencsw.org> Message-ID: <1277385817-sup-7407@pinkfloyd.chass.utoronto.ca> Excerpts from Dagobert Michelsen's message of Thu Jun 24 09:13:10 -0400 2010: Hi Dago, > Is this really necessary? What's wrong if the packages bring the > directory themselves? Besides that it is possible to forget the > directory if you are not using GAR ;-) It's not a problem if everyone also includes it with their packages (the default when using GAR). This just provides extra protection. :) I thought it would be a good addition to the common directory layout provided by the CSWcommon package. Thanks -Ben -- Ben Walton Systems Programmer - CHASS University of Toronto C:416.407.5610 | W:416.978.4302 From ihsan at opencsw.org Thu Jun 24 15:44:19 2010 From: ihsan at opencsw.org (Ihsan Dogan) Date: Thu, 24 Jun 2010 16:44:19 +0300 Subject: [csw-maintainers] www.opencsw.org: PHP + sendmail broken (/opt/csw/sbin/sendmail missing) In-Reply-To: <4C1F3E74.7020308@opencsw.org> References: <4C1F3E74.7020308@opencsw.org> Message-ID: <4C236133.4030307@opencsw.org> Am 21.06.10 13:27, schrieb Sebastian Kayser: > the new www box doesn't send out Mantis emails and is either missing CSW > sendmail or needs a PHP config adjustment. Could someone with admin > right take care of it please? This is fixed now and should work again. Ihsan -- ihsan at dogan.ch http://blog.dogan.ch/ From bwalton at opencsw.org Thu Jun 24 15:52:08 2010 From: bwalton at opencsw.org (Ben Walton) Date: Thu, 24 Jun 2010 09:52:08 -0400 Subject: [csw-maintainers] www.opencsw.org: PHP + sendmail broken (/opt/csw/sbin/sendmail missing) In-Reply-To: <4C236133.4030307@opencsw.org> References: <4C1F3E74.7020308@opencsw.org> <4C236133.4030307@opencsw.org> Message-ID: <1277387491-sup-679@pinkfloyd.chass.utoronto.ca> Excerpts from Ihsan Dogan's message of Thu Jun 24 09:44:19 -0400 2010: > > the new www box doesn't send out Mantis emails and is either missing CSW > > sendmail or needs a PHP config adjustment. Could someone with admin > > right take care of it please? > > This is fixed now and should work again. Given the deluge I got from mantis this morning, I can confirm that it is indeed working again! :) Thanks Ihsan! -Ben -- Ben Walton Systems Programmer - CHASS University of Toronto C:416.407.5610 | W:416.978.4302 From phil at bolthole.com Thu Jun 24 15:56:26 2010 From: phil at bolthole.com (Philip Brown) Date: Thu, 24 Jun 2010 06:56:26 -0700 Subject: [csw-maintainers] addition to CSWcommon In-Reply-To: <1277385817-sup-7407@pinkfloyd.chass.utoronto.ca> References: <1277384284-sup-8472@pinkfloyd.chass.utoronto.ca> <9AF254A4-68FF-447A-AA43-FF090C2FA375@opencsw.org> <1277385817-sup-7407@pinkfloyd.chass.utoronto.ca> Message-ID: I think it makes sense if I update it. I'm just not convinced about the update yet :-) btw it's wierd that regular files autocreate subdirs but symlinks do not On Thursday, June 24, 2010, Ben Walton wrote: > Excerpts from Dagobert Michelsen's message of Thu Jun 24 09:13:10 -0400 2010: > > Hi Dago, > >> Is this really necessary? What's wrong if the packages bring the >> directory themselves? Besides that it is possible to forget the >> directory if you are not using GAR ;-) > > It's not a problem if everyone also includes it with their packages > (the default when using GAR). ?This just provides extra protection. :) > I thought it would be a good addition to the common directory layout > provided by the CSWcommon package. > > Thanks > -Ben > -- > Ben Walton > Systems Programmer - CHASS > University of Toronto > C:416.407.5610 | W:416.978.4302 > > _______________________________________________ > maintainers mailing list > maintainers at lists.opencsw.org > https://lists.opencsw.org/mailman/listinfo/maintainers > .:: This mailing list's archive is public. ::. > From phil at bolthole.com Sat Jun 26 00:53:09 2010 From: phil at bolthole.com (Philip Brown) Date: Fri, 25 Jun 2010 15:53:09 -0700 Subject: [csw-maintainers] gar suggestion Message-ID: i'm pulling out an old garified dir, libgnomecanvas. I updated the GARVERSION, and did a make. it fetched the file. and then bombed out because didnt know how to [...] I had to manually do 'make makesums', and then the make again. so I'd like to suggest that a fetch, automatically calculate the sums. From bwalton at opencsw.org Sat Jun 26 01:37:08 2010 From: bwalton at opencsw.org (Ben Walton) Date: Fri, 25 Jun 2010 19:37:08 -0400 Subject: [csw-maintainers] gar suggestion In-Reply-To: References: Message-ID: <1277508911-sup-531@pinkfloyd.chass.utoronto.ca> Excerpts from Philip Brown's message of Fri Jun 25 18:53:09 -0400 2010: > so I'd like to suggest that a fetch, automatically calculate the > sums. I've thought about this same solution before and reconsidered. If you want to update the makesums automatically, there is no point in having them at all. If you want to rephrase the suggestion to "lets get rid of makesums" I'd likely jump on that with you. -Ben -- Ben Walton Systems Programmer - CHASS University of Toronto C:416.407.5610 | W:416.978.4302 From phil at bolthole.com Sat Jun 26 01:53:00 2010 From: phil at bolthole.com (Philip Brown) Date: Fri, 25 Jun 2010 16:53:00 -0700 Subject: [csw-maintainers] gar suggestion In-Reply-To: <1277508911-sup-531@pinkfloyd.chass.utoronto.ca> References: <1277508911-sup-531@pinkfloyd.chass.utoronto.ca> Message-ID: On Fri, Jun 25, 2010 at 4:37 PM, Ben Walton wrote: > Excerpts from Philip Brown's message of Fri Jun 25 18:53:09 -0400 2010: > >> so I'd like to suggest that a fetch, automatically calculate the >> sums. > > I've thought about this same solution before and reconsidered. ?If you > want to update the makesums automatically, there is no point in having > them at all. This is not quite the same thing as I am saying. I am saying that IF you are in "download the source code" mode, AND IF there is no existing checksum in the file, THEN auto-generate one. > If you want to rephrase the suggestion to "lets get rid > of [checksums] [(entirely)] I'd likely jump on that with you. I'd second that though :-) From bwalton at opencsw.org Sat Jun 26 02:06:57 2010 From: bwalton at opencsw.org (Ben Walton) Date: Fri, 25 Jun 2010 20:06:57 -0400 Subject: [csw-maintainers] gar suggestion In-Reply-To: References: <1277508911-sup-531@pinkfloyd.chass.utoronto.ca> Message-ID: <1277510745-sup-838@pinkfloyd.chass.utoronto.ca> Excerpts from Philip Brown's message of Fri Jun 25 19:53:00 -0400 2010: > I am saying that IF you are in "download the source code" mode, AND > IF there is no existing checksum in the file, THEN auto-generate > one. Ok, so autogenerate iff ! -f makesums. That would be a fair addition. > > If you want to rephrase the suggestion to "lets get rid of > > [checksums] [(entirely)] I'd likely jump on that with you. > > I'd second that though :-) Does anyone else find the makesums helpful? Thanks -Ben -- Ben Walton Systems Programmer - CHASS University of Toronto C:416.407.5610 | W:416.978.4302 From bonivart at opencsw.org Sat Jun 26 17:29:59 2010 From: bonivart at opencsw.org (Peter Bonivart) Date: Sat, 26 Jun 2010 17:29:59 +0200 Subject: [csw-maintainers] Define supported use cases and work from there Message-ID: For as long as I can remember we have had these issues with NFS, zones, SMF, OpenSolaris and so on. Usually someone finds something that doesn't work and we jump into a technical discussion of how we can fix that case. Often seemingly small changes then break stuff, i.e. makes stuff worse. The big problem, as I see it, is that the fixes we try to provide only help very few but complicate/break stuff for many. It also takes a long time to "discuss" this since what people want is already locked and their flavor of solution is "simple" (to them). During this time, we have broken packages and/or a halted release process. Right now we have both with the init script placement issue. I suggest that we define which use cases we should support and in which priority. I think it will be much easier to solve technical issues if we have agreed on the goals. It will also increase the chance of a maintainer to do the right thing from the start instead of personal interpretations of what we have lacked to document. Not to mention that end users can actually know what to expect. -- /peter From phil at bolthole.com Sat Jun 26 17:53:51 2010 From: phil at bolthole.com (Philip Brown) Date: Sat, 26 Jun 2010 08:53:51 -0700 Subject: [csw-maintainers] Define supported use cases and work from there In-Reply-To: References: Message-ID: On Saturday, June 26, 2010, Peter Bonivart wrote: .... > I suggest that we define which use cases we should support and in > which priority. You make it sound as though this was not previously defined. But it has already been defined, and documented, for over 4 years, approximately. We have officially supported NFS use for at least that long, AFAIR. Problems arise when people forget that, and implement something that then has to be reworked later when someone else points that out to them. From bonivart at opencsw.org Sat Jun 26 17:57:06 2010 From: bonivart at opencsw.org (Peter Bonivart) Date: Sat, 26 Jun 2010 17:57:06 +0200 Subject: [csw-maintainers] Define supported use cases and work from there In-Reply-To: References: Message-ID: On Sat, Jun 26, 2010 at 5:53 PM, Philip Brown wrote: > On Saturday, June 26, 2010, Peter Bonivart wrote: > .... >> I suggest that we define which use cases we should support and in >> which priority. > > You make it sound as though this was not previously defined. But it > has already been defined, and documented, for over 4 years, > approximately. We have officially supported NFS use for at least that > long, AFAIR. Problems arise when people forget that, and implement > something that then has to be reworked later when someone else points > that out to them. Is this documentet? Where? Should we support client utilities only or full server daemons? Maybe someone else than you should come forward for those "big enterprises" using NFS since I can only remember you promoting them. I've never seen any active maintainer other than you be interested in this use case. However, it's about more cases than the dreaded NFS. -- /peter From trygvis at opencsw.org Sat Jun 26 21:09:14 2010 From: trygvis at opencsw.org (=?ISO-8859-1?Q?Trygve_Laugst=F8l?=) Date: Sat, 26 Jun 2010 21:09:14 +0200 Subject: [csw-maintainers] gar suggestion In-Reply-To: <1277510745-sup-838@pinkfloyd.chass.utoronto.ca> References: <1277508911-sup-531@pinkfloyd.chass.utoronto.ca> <1277510745-sup-838@pinkfloyd.chass.utoronto.ca> Message-ID: <4C26505A.4060109@opencsw.org> On 6/26/10 2:06 AM, Ben Walton wrote: > Excerpts from Philip Brown's message of Fri Jun 25 19:53:00 -0400 2010: > >> I am saying that IF you are in "download the source code" mode, AND >> IF there is no existing checksum in the file, THEN auto-generate >> one. > > Ok, so autogenerate iff ! -f makesums. That would be a fair addition. > >>> If you want to rephrase the suggestion to "lets get rid of >>> [checksums] [(entirely)] I'd likely jump on that with you. >> >> I'd second that though :-) > > Does anyone else find the makesums helpful? Not me at least. -- Trygve From rupert at opencsw.org Sun Jun 27 12:07:38 2010 From: rupert at opencsw.org (rupert THURNER) Date: Sun, 27 Jun 2010 12:07:38 +0200 Subject: [csw-maintainers] website: experimental? latest package releases? Message-ID: hi, how one is supposed to reach experimental and a guide how to install with the new website? * http://www.opencsw.org/get-it/packages/ contains a dead link "testing". * http://www.opencsw.org/get-it/testing-packages/ is outdated (newest pkg from 2008) * feeds contains http://mirror.opencsw.org/cgi-bin/experimental.atom and where does one see the latest upgraded packages? rupert From maciej at opencsw.org Sun Jun 27 14:46:29 2010 From: maciej at opencsw.org (Maciej (Matchek) Blizinski) Date: Sun, 27 Jun 2010 13:46:29 +0100 Subject: [csw-maintainers] website: experimental? latest package releases? In-Reply-To: References: Message-ID: No dia 27 de Junho de 2010 11:07, rupert THURNER escreveu: > hi, > > how one is supposed to reach experimental and a guide how to install > with the new website? > > * http://www.opencsw.org/get-it/packages/ contains a dead link "testing". This one is not directly editable. > * http://www.opencsw.org/get-it/testing-packages/ is outdated (newest > pkg from 2008) The list of packages was not live; it was a static HTML stored in the database. I've replaced it with a descriptive paragraph and added a link to experimental. Feel free to further improve it. I only did the minimal amount of work to prevent this page from being misleading. > * feeds contains http://mirror.opencsw.org/cgi-bin/experimental.atom > and where does one see the latest upgraded packages? Wasn't this feed not up yet? There recently was a thread about it, if you look up the archives. From phil at bolthole.com Mon Jun 28 05:09:40 2010 From: phil at bolthole.com (Philip Brown) Date: Sun, 27 Jun 2010 20:09:40 -0700 Subject: [csw-maintainers] When is ARCH=all not the same on all platforms? Message-ID: <20100628030940.GA29262@bolthole.com> When is ARCH=all not actually the same? Apparently, When you build it on x86, vs sparc. current9x$ ls -l *doc*gz -rw-r--r-- 1 phil csw 942656 Jun 28 04:43 gimplibs_doc-2.6.8,REV=2010.06.12-SunOS5.9-all-CSW.pkg.gz current9s$ ls -l *doc*gz -rw-r--r-- 1 phil csw 1008932 Jun 28 04:40 gimplibs_doc-2.6.8,REV=2010.06.12-SunOS5.9-all-CSW.pkg.gz same contents in package. really, really wierd. oh wait... package is the same size when gunzipped. gzip works "better" on current9x by default!??? From william at wbonnet.net Mon Jun 28 07:47:38 2010 From: william at wbonnet.net (William Bonnet) Date: Mon, 28 Jun 2010 07:47:38 +0200 Subject: [csw-maintainers] Summer '10 will be held on 14 and 15 of August Message-ID: <4C28377A.3060103@wbonnet.net> Hi all, The next summer camp will be held on the 14 and the 15 of aughust in Paris France. These two days match the choice of the the voters from the doodle poll ( http://www.doodle.com/vsaehg5rxyrcgxcq ). Please can you start to update the attendee section in the wiki page ( http://wiki.opencsw.org/summercamp-2010 ). People needing an hotel reservation can ask me for prebooking. I will provide some hotel addresses next to the meeting place. But before doing so, i'd like to have on idea of how many people will need a room (please indicates single or double room). thanks in advance cheers W. -- William http://www.wbonnet.net http://www.sunwizard.net Le site fran?ais des amateurs de stations Unix http://www.opencsw.org Community SoftWare for Solaris http://www.guses.org French speaking Solaris User Group From dam at opencsw.org Mon Jun 28 09:23:31 2010 From: dam at opencsw.org (Dagobert Michelsen) Date: Mon, 28 Jun 2010 09:23:31 +0200 Subject: [csw-maintainers] Summer '10 will be held on 14 and 15 of August In-Reply-To: <4C28377A.3060103@wbonnet.net> References: <4C28377A.3060103@wbonnet.net> Message-ID: <6E97216E-80EC-4290-A38D-7449962BB429@opencsw.org> Hi William, Am 28.06.2010 um 07:47 schrieb William Bonnet: > The next summer camp will be held on the 14 and the 15 of aughust in > Paris France. > > These two days match the choice of the the voters from the doodle > poll ( http://www.doodle.com/vsaehg5rxyrcgxcq ). > > Please can you start to update the attendee section in the wiki page > ( http://wiki.opencsw.org/summercamp-2010 ). > > People needing an hotel reservation can ask me for prebooking. I > will provide some hotel addresses next to the meeting place. But > before doing so, i'd like to have on idea of how many people will > need a room (please indicates single or double room). This is great! Please arrange a hotel room from 13. to 16. for me. I really look forward to this :-) Best regards -- Dago From dam at opencsw.org Mon Jun 28 09:38:24 2010 From: dam at opencsw.org (Dagobert Michelsen) Date: Mon, 28 Jun 2010 09:38:24 +0200 Subject: [csw-maintainers] gar suggestion In-Reply-To: <4C26505A.4060109@opencsw.org> References: <1277508911-sup-531@pinkfloyd.chass.utoronto.ca> <1277510745-sup-838@pinkfloyd.chass.utoronto.ca> <4C26505A.4060109@opencsw.org> Message-ID: Hi, Am 26.06.2010 um 21:09 schrieb Trygve Laugst?l: > On 6/26/10 2:06 AM, Ben Walton wrote: >> Excerpts from Philip Brown's message of Fri Jun 25 19:53:00 -0400 >> 2010: >> >>> I am saying that IF you are in "download the source code" mode, AND >>> IF there is no existing checksum in the file, THEN auto-generate >>> one. >> >> Ok, so autogenerate iff ! -f makesums. That would be a fair >> addition. >> >>>> If you want to rephrase the suggestion to "lets get rid of >>>> [checksums] [(entirely)] I'd likely jump on that with you. >>> >>> I'd second that though :-) >> >> Does anyone else find the makesums helpful? > > Not me at least. Well, ok then. Let's disable this checksum-thing. Best regards -- Dago From nicols at opencsw.org Mon Jun 28 09:43:43 2010 From: nicols at opencsw.org (Andrew Robert Nicols) Date: Mon, 28 Jun 2010 08:43:43 +0100 Subject: [csw-maintainers] gar suggestion In-Reply-To: References: <1277508911-sup-531@pinkfloyd.chass.utoronto.ca> <1277510745-sup-838@pinkfloyd.chass.utoronto.ca> <4C26505A.4060109@opencsw.org> Message-ID: <20100628074343.GB20026@banoffee.lancs.ac.uk> On Fri, Jun 25, 2010 at 08:06:57PM -0400, Ben Walton wrote: > Excerpts from Philip Brown's message of Fri Jun 25 19:53:00 -0400 2010: > > > I am saying that IF you are in "download the source code" mode, AND > > IF there is no existing checksum in the file, THEN auto-generate > > one. > > Ok, so autogenerate iff ! -f makesums. That would be a fair addition. To me that seems a good solution. Though I think that it should be additive. I.e. If an individual file is missing, add it rather than removing the whole file and re-generating it. > > > If you want to rephrase the suggestion to "lets get rid of > > > [checksums] [(entirely)] I'd likely jump on that with you. > > > > I'd second that though :-) > > Does anyone else find the makesums helpful? I can see the use for them. I think that they do have their place but perhaps the use-case is a little rare. For example, a package whose source has a generic download URL with no version numbering in the file. If you want to re-build the package without updating the version, then you may want to ensure that the version you've downloaded hasn't changed. Andrew -- Systems Developer e: andrew.nicols at luns.net.uk im: a.nicols at jabber.lancs.ac.uk t: +44 (0)1524 5 10147 Lancaster University Network Services is a limited company registered in England and Wales. Registered number: 04311892. Registered office: University House, Lancaster University, Lancaster, LA1 4YW -------------- next part -------------- A non-text attachment was scrubbed... Name: not available Type: application/pgp-signature Size: 197 bytes Desc: Digital signature URL: From dam at opencsw.org Mon Jun 28 09:47:43 2010 From: dam at opencsw.org (Dagobert Michelsen) Date: Mon, 28 Jun 2010 09:47:43 +0200 Subject: [csw-maintainers] website: experimental? latest package releases? In-Reply-To: References: Message-ID: Hi Rupert, Am 27.06.2010 um 14:46 schrieb Maciej (Matchek) Blizinski: > No dia 27 de Junho de 2010 11:07, rupert THURNER > escreveu: >> hi, >> >> how one is supposed to reach experimental and a guide how to install >> with the new website? >> >> * http://www.opencsw.org/get-it/packages/ contains a dead link >> "testing". > > This one is not directly editable. Just fwd' William. >> * http://www.opencsw.org/get-it/testing-packages/ is outdated (newest >> pkg from 2008) > > The list of packages was not live; it was a static HTML stored in the > database. I've replaced it with a descriptive paragraph and added a > link to experimental. Feel free to further improve it. I only did > the minimal amount of work to prevent this page from being misleading. Looks good. For Paris I would like to have a webpage-workshop to straighten out all open issues. >> * feeds contains http://mirror.opencsw.org/cgi-bin/experimental.atom >> and where does one see the latest upgraded packages? > > Wasn't this feed not up yet? There recently was a thread about it, if > you look up the archives. You mean the "released packages feed"? Unfortunately that one is down at the moment - historically www and the release machines where the same in the past, we need to have some interface here. Best regards -- Dago From dam at opencsw.org Mon Jun 28 09:50:21 2010 From: dam at opencsw.org (Dagobert Michelsen) Date: Mon, 28 Jun 2010 09:50:21 +0200 Subject: [csw-maintainers] When is ARCH=all not the same on all platforms? In-Reply-To: <20100628030940.GA29262@bolthole.com> References: <20100628030940.GA29262@bolthole.com> Message-ID: <38515B69-8C64-4CBF-A8FE-541222ED9DCE@opencsw.org> Hi Phil, Am 28.06.2010 um 05:09 schrieb Philip Brown: > When is ARCH=all not actually the same? > > Apparently, When you build it on x86, vs sparc. > > current9x$ ls -l *doc*gz > -rw-r--r-- 1 phil csw 942656 Jun 28 04:43 > gimplibs_doc-2.6.8,REV=2010.06.12-SunOS5.9-all-CSW.pkg.gz > > > current9s$ ls -l *doc*gz > -rw-r--r-- 1 phil csw 1008932 Jun 28 04:40 > gimplibs_doc-2.6.8,REV=2010.06.12-SunOS5.9-all-CSW.pkg.gz > > > same contents in package. really, really wierd. > oh wait... package is the same size when gunzipped. > > gzip works "better" on current9x by default!??? Maybe this is a path issue - there are different gzip version in /usr/ bin and /opt/csw/bin. Best rerads -- Dago From Cyrus.MEHTA at sungard.com Thu Jun 17 00:38:17 2010 From: Cyrus.MEHTA at sungard.com (Cyrus.MEHTA at sungard.com) Date: Wed, 16 Jun 2010 18:38:17 -0400 Subject: [csw-maintainers] website issues In-Reply-To: <20100616103604.GB31767@sebastiankayser.de> References: <20100616103604.GB31767@sebastiankayser.de> Message-ID: Might be related: When I click on any individual package links like: http://www.opencsw.org/packages/CSWa52dec/ Its mostly blank with this error at the top (only visible if your text size is large enough): Invalid query : Table 'CSW.mantis_project_table' doesn't exist Query used : select id from mantis_project_table where name = 'a52dec' Cyrus -----Original Message----- From: maintainers-bounces+ckmehta1=opencsw.org at lists.opencsw.org [mailto:maintainers-bounces+ckmehta1=opencsw.org at lists.opencsw.org] On Behalf Of Sebastian Kayser Sent: Wednesday, June 16, 2010 6:36 AM To: List for OpenCSW maintainers Subject: Re: [csw-maintainers] website issues * Maciej (Matchek) Blizinski wrote: > 3. No links to the search subpage. 4. Package links to the bug tracker not always correctly determined http://www.opencsw.org/packages/pm_paramsvalidate/ -> http://www.opencsw.org/mantis/set_project.php?project_id= (empty project_id) Sebastian _______________________________________________ maintainers mailing list maintainers at lists.opencsw.org https://lists.opencsw.org/mailman/listinfo/maintainers .:: This mailing list's archive is public. ::. From maciej at opencsw.org Mon Jun 28 17:31:51 2010 From: maciej at opencsw.org (Maciej (Matchek) Blizinski) Date: Mon, 28 Jun 2010 16:31:51 +0100 Subject: [csw-maintainers] cswclassutils: location of template init scripts In-Reply-To: References: Message-ID: No dia 22 de Junho de 2010 19:34, Philip Brown escreveu: > Maciej, you seem to wish to go down the pedantic path. So okay, lets > get pedantic :-) Me, pedantic? Since when? ;-) > On Mon, Jun 21, 2010 at 4:40 AM, Maciej (Matchek) Blizinski > wrote: >> No dia 16 de Junho de 2010 18:38, Philip Brown escreveu: >>> On Wed, Jun 16, 2010 at 10:26 AM, Maciej (Matchek) Blizinski >>> wrote: >>>> ... ?The definition of CAS >>>> is that they handle the installation of files, with potential >>>> additional processing. ?Having CAS process other files is something >>>> that CAS isn't intended for. ?I know that you can hack it, but it >>>> might cross the fine line between use and abuse. >>> >>>.... >> >> Your two above paragraphs sound as if you thought that I'm questioning >> the use of classutils, which I'm not. ?I'm saying that classutils are >> designed to process given lists of files. > > http://docs.sun.com/app/docs/doc/806-7008/6jftmsc38?a=view > > "Object classes allow a series of actions to be performed on a group > of package objects at installation or removal." > > This is ambiguous. It doesnt say that renaming the objects, or > duplicating, or removing, objects, is either explicitly allowed, or > explicitly denied. It merely says "actions", on "objects" with the > only limitation that the objects be associated with the package. > > > My view of this is: > ?Since the entire purpose of system standard utilities such as > "installf" and "removef", are to DYNAMICALLY create and remove files > associated with the package.... files that, **by definition**, are not > in the initial prototype/pkgmap file... then the ambiguity is strongly > resolved in favor of "it is allowed.". > > > > > >> ?Even though you could write >> a CAS that processes additional files under certain circumstances >> (empty input), I think this is a bad idea, because: >> >> 1. It relies on undocumented behavior which might change (CAS being >> called with empty input vs CAS not being called at all) > > Actually, it is *explicitly documented* behaviour. > > again, from the above url: > > "Note ? > > Even if there are no regular files of this class anywhere in the > package, the class action script will be called at least once with an > empty list and the ENDOFCLASS argument. > " *drumroll* /me is convinced Cool, if it's documented, I only see minor reasons why not do it that way, so the balance shifts towards the solution you propose. If it's implemented in a sane way and documented on our end, it should be okay. The above link should be added to the source in a comments section, for future reference. Maciej From phil at bolthole.com Mon Jun 28 19:19:40 2010 From: phil at bolthole.com (Philip Brown) Date: Mon, 28 Jun 2010 10:19:40 -0700 Subject: [csw-maintainers] When is ARCH=all not the same on all platforms? In-Reply-To: <38515B69-8C64-4CBF-A8FE-541222ED9DCE@opencsw.org> References: <20100628030940.GA29262@bolthole.com> <38515B69-8C64-4CBF-A8FE-541222ED9DCE@opencsw.org> Message-ID: On Mon, Jun 28, 2010 at 12:50 AM, Dagobert Michelsen wrote: > Hi Phil, > > Am 28.06.2010 um 05:09 schrieb Philip Brown: > >> gzip works "better" on current9x by default!??? > > Maybe this is a path issue - there are different gzip version in /usr/bin > and /opt/csw/bin. > > interesting idea, but nope thats not it. From phil at bolthole.com Mon Jun 28 21:48:09 2010 From: phil at bolthole.com (Philip Brown) Date: Mon, 28 Jun 2010 12:48:09 -0700 Subject: [csw-maintainers] website issues In-Reply-To: References: <20100616103604.GB31767@sebastiankayser.de> Message-ID: On Wed, Jun 16, 2010 at 3:38 PM, wrote: > Might be related: > > When I click on any individual package links like: > ? ? ? ?http://www.opencsw.org/packages/CSWa52dec/ > Its mostly blank with this error at the top (only visible if your text > size is large enough): > ? ? ? ?Invalid query : Table 'CSW.mantis_project_table' doesn't exist > Query used : select id from mantis_project_table where name = 'a52dec' > seems to be fine now. From phil at bolthole.com Mon Jun 28 22:45:37 2010 From: phil at bolthole.com (Philip Brown) Date: Mon, 28 Jun 2010 13:45:37 -0700 Subject: [csw-maintainers] gimp. with python problems Message-ID: On Thu, Jun 3, 2010 at 8:12 PM, Jeffery Small wrote: > Philip Brown writes: > >>Which in theory means, we could have a new GNOME environment after this, too! > > Could you work on updating gimp before tackling the much more involved Gnome > project? > fyi: you got your wish. sort of. "proper" gimp 2.6 just being batched out now. Oddly, it doesnt use gnome. the gnomevfs funtionality seems to be missing. and it dropped its own help browser, in favor of just integrating into whatever browser you have already (uses firefox if you have it) But otherwise, fully featured. WIth the exception that it crashed and burned attempting to compiile python modules. Maciej, mr python, you should try a gimp configure yourself, and see if you can figure out why it fails you dont have to compile the thing. just attempt to configure. From phil at bolthole.com Tue Jun 29 00:14:53 2010 From: phil at bolthole.com (Philip Brown) Date: Mon, 28 Jun 2010 15:14:53 -0700 Subject: [csw-maintainers] gar suggestion In-Reply-To: References: <1277508911-sup-531@pinkfloyd.chass.utoronto.ca> <1277510745-sup-838@pinkfloyd.chass.utoronto.ca> <4C26505A.4060109@opencsw.org> Message-ID: On Mon, Jun 28, 2010 at 12:38 AM, Dagobert Michelsen wrote: > > Well, ok then. Let's disable this checksum-thing. > Cool! while you're busy disabling things... It would be nice if you disabled automatic "git-ification(?)", unless the person specifically says "I am interested in generating a patch set". Most of the time, its just an annoying waste of time. I've just come across a situation where it WOULD potentially be useful.. so please note I'm not saying it's garbage :) I'm just saying dont bother me with it, unless I tell the build system I really want it. (I'm envisioning a per-directory make target. ie: "gmake git" or something) From bwalton at opencsw.org Tue Jun 29 02:24:04 2010 From: bwalton at opencsw.org (Ben Walton) Date: Mon, 28 Jun 2010 20:24:04 -0400 Subject: [csw-maintainers] gar suggestion In-Reply-To: References: <1277508911-sup-531@pinkfloyd.chass.utoronto.ca> <1277510745-sup-838@pinkfloyd.chass.utoronto.ca> <4C26505A.4060109@opencsw.org> Message-ID: <1277770549-sup-3354@pinkfloyd.chass.utoronto.ca> Excerpts from Philip Brown's message of Mon Jun 28 18:14:53 -0400 2010: > while you're busy disabling things... It would be nice if you > disabled automatic "git-ification(?)", unless the person > specifically says "I am interested in generating a patch set". Most > of the time, its just an annoying waste of time. Sebastian and I discussed this too but I decided against it. Here's the reason: If you don't do the initial git snapshots at extract and then patch time, you don't have a repo to work from. It _is_ possible to simply go in and snapshot it after the fact, but that's not useful in the 'makepatch' sense. You typically want to: 1. find out $foo doesn't build. 2. hack the source until it does. 3. capture the changes that 'made it go.' If you don't have the git repo setup ahead of time, this workflow suffers. So, while it's easy to add a knob for this, it makes it annoying to subsequently have to re-'download', re-extract, etc in order to turn it on when it is needed. The re-extract step would blow away your changes too, so you'd need a way to maintain your own manual diff stuff. Personally, I'm finding that for small projects, the git-ification isn't overly burdensome, while for larger projects, it's a small price (time-wise) in comparison to the extract step... > really want it. (I'm envisioning a per-directory make target. ie: > "gmake git" or something) I'm willing to change it as makes the most people happy with it...that being said, much of the current benefit of the way it's setup is due to the multi-step integration (extract, patch). Adding a toggle will make it worse, imo. Anyone else have a preference one way or another? Thanks -Ben -- Ben Walton Systems Programmer - CHASS University of Toronto C:416.407.5610 | W:416.978.4302 From phil at bolthole.com Tue Jun 29 02:45:48 2010 From: phil at bolthole.com (Philip Brown) Date: Mon, 28 Jun 2010 17:45:48 -0700 Subject: [csw-maintainers] gar suggestion In-Reply-To: <1277770549-sup-3354@pinkfloyd.chass.utoronto.ca> References: <1277508911-sup-531@pinkfloyd.chass.utoronto.ca> <1277510745-sup-838@pinkfloyd.chass.utoronto.ca> <4C26505A.4060109@opencsw.org> <1277770549-sup-3354@pinkfloyd.chass.utoronto.ca> Message-ID: On Mon, Jun 28, 2010 at 5:24 PM, Ben Walton wrote: > > It _is_ possible to simply go in and snapshot it after the fact, but > that's not useful in the 'makepatch' sense. ?You typically want to: > > 1. find out $foo doesn't build. > 2. hack the source until it does. > 3. capture the changes that 'made it go.' > > If you don't have the git repo setup ahead of time, this workflow > suffers. ?? what's so horrible about adding to the above workflow 1.5 type "gmake gitify" ? 2. hack the source... Contrariwise, if you KNOW you are not interested in patching, and you KNOW you just want to "build and go"... waiting the extra time to watch the system "set up for patching" is fairly annoying. and consider if this is a non-maintainer, who just wants to built a package,and might not even have git installed? From bwalton at opencsw.org Tue Jun 29 04:08:59 2010 From: bwalton at opencsw.org (Ben Walton) Date: Mon, 28 Jun 2010 22:08:59 -0400 Subject: [csw-maintainers] gar suggestion In-Reply-To: References: <1277508911-sup-531@pinkfloyd.chass.utoronto.ca> <1277510745-sup-838@pinkfloyd.chass.utoronto.ca> <4C26505A.4060109@opencsw.org> <1277770549-sup-3354@pinkfloyd.chass.utoronto.ca> Message-ID: <1277777170-sup-9659@pinkfloyd.chass.utoronto.ca> Excerpts from Philip Brown's message of Mon Jun 28 20:45:48 -0400 2010: > what's so horrible about adding to the above workflow > > 1.5 type "gmake gitify" ? > 2. hack the source... Because it breaks the nice granularity that is had be making separate commits of the original source and individual patches. > Contrariwise, if you KNOW you are not interested in patching, and > you KNOW you just want to "build and go"... waiting the extra time > to watch the system "set up for patching" is fairly annoying. Before I spend time debating this with someone that only peripherally uses GAR, I'd prefer to hear views of others that use it daily... > and consider if this is a non-maintainer, who just wants to built a > package,and might not even have git installed? This is even less of a consideration for me. GAR is there to make the life of the maintainer easier. If someone wants to build a recipe and tweak a package, they can 'take the hit.' Git is a required package for using GAR (as defined by the list of packages looked for prior to doing anything). Thanks -Ben -- Ben Walton Systems Programmer - CHASS University of Toronto C:416.407.5610 | W:416.978.4302 From bonivart at opencsw.org Tue Jun 29 10:05:17 2010 From: bonivart at opencsw.org (Peter Bonivart) Date: Tue, 29 Jun 2010 10:05:17 +0200 Subject: [csw-maintainers] gar suggestion In-Reply-To: <1277777170-sup-9659@pinkfloyd.chass.utoronto.ca> References: <1277508911-sup-531@pinkfloyd.chass.utoronto.ca> <1277510745-sup-838@pinkfloyd.chass.utoronto.ca> <4C26505A.4060109@opencsw.org> <1277770549-sup-3354@pinkfloyd.chass.utoronto.ca> <1277777170-sup-9659@pinkfloyd.chass.utoronto.ca> Message-ID: On Tue, Jun 29, 2010 at 4:08 AM, Ben Walton wrote: > Before I spend time debating this with someone that only peripherally > uses GAR, I'd prefer to hear views of others that use it daily... I love the new patch stuff, the few seconds it takes is well worth it. Doesn't bother me at all. -- /peter From maciej at opencsw.org Tue Jun 29 10:46:53 2010 From: maciej at opencsw.org (Maciej (Matchek) Blizinski) Date: Tue, 29 Jun 2010 09:46:53 +0100 Subject: [csw-maintainers] gar suggestion In-Reply-To: References: <1277508911-sup-531@pinkfloyd.chass.utoronto.ca> <1277510745-sup-838@pinkfloyd.chass.utoronto.ca> <4C26505A.4060109@opencsw.org> <1277770549-sup-3354@pinkfloyd.chass.utoronto.ca> <1277777170-sup-9659@pinkfloyd.chass.utoronto.ca> Message-ID: No dia 29 de Junho de 2010 09:05, Peter Bonivart escreveu: > On Tue, Jun 29, 2010 at 4:08 AM, Ben Walton wrote: >> Before I spend time debating this with someone that only peripherally >> uses GAR, I'd prefer to hear views of others that use it daily... > > I love the new patch stuff, the few seconds it takes is well worth it. > Doesn't bother me at all. I like to new gar integration too. I guess it could be slowing down the process for large source trees, but in my case the benefits by far outweigh the costs. When I work with packages, most of the time it's about getting it to compile. Producing patches is a large part of it, and I really like it that I always have an initialized git repository I can use. From nicols at opencsw.org Tue Jun 29 11:21:32 2010 From: nicols at opencsw.org (Andrew Robert Nicols) Date: Tue, 29 Jun 2010 10:21:32 +0100 Subject: [csw-maintainers] gar suggestion In-Reply-To: References: <1277510745-sup-838@pinkfloyd.chass.utoronto.ca> <4C26505A.4060109@opencsw.org> <1277770549-sup-3354@pinkfloyd.chass.utoronto.ca> <1277777170-sup-9659@pinkfloyd.chass.utoronto.ca> Message-ID: <20100629092131.GF20026@banoffee.lancs.ac.uk> On Tue, Jun 29, 2010 at 09:46:53AM +0100, Maciej (Matchek) Blizinski wrote: > No dia 29 de Junho de 2010 09:05, Peter Bonivart > escreveu: > > On Tue, Jun 29, 2010 at 4:08 AM, Ben Walton wrote: > >> Before I spend time debating this with someone that only peripherally > >> uses GAR, I'd prefer to hear views of others that use it daily... > > > > I love the new patch stuff, the few seconds it takes is well worth it. > > Doesn't bother me at all. > > I like to new gar integration too. I guess it could be slowing down > the process for large source trees, but in my case the benefits by far > outweigh the costs. Same goes for me. I'm quite happy to have that small performance hit that the git init brings as part of the build process to make patching later easier. Andrew -- Systems Developer e: andrew.nicols at luns.net.uk im: a.nicols at jabber.lancs.ac.uk t: +44 (0)1524 5 10147 Lancaster University Network Services is a limited company registered in England and Wales. Registered number: 04311892. Registered office: University House, Lancaster University, Lancaster, LA1 4YW -------------- next part -------------- A non-text attachment was scrubbed... Name: not available Type: application/pgp-signature Size: 197 bytes Desc: Digital signature URL: From phil at bolthole.com Tue Jun 29 18:58:33 2010 From: phil at bolthole.com (Philip Brown) Date: Tue, 29 Jun 2010 09:58:33 -0700 Subject: [csw-maintainers] A question of naming, in our svn repository Message-ID: Maintainers, your input is required, over a matter of policy in naming directories, in our subversion repository. We have two mutually exclusive ways to go: a) facilitate finding an svn directory from an "upstream" name b) facilitate finding an svn directory from a package name. Specific example, of the general case: We have a package, "gtk_engines". Should the subversion directory for it, be named "gtk-engines", or "gtk_engines"? My thoughts on the issue are as follows: Of the people coming to our subversion repository, (both maintainers, and non-maintainers) almost no-one will care about case (a). They will mostly care about either "I want to take an existing OpenCSW package, but recompile it with my own tweaks", or "I am going to take over a package; I need to grab it from svn". its all very well to say, "yes well the specific directory information is in the OPENCSW_REPOSITORY property", but that is a hassle to look for, and most people will forget about that anyway (if they even knew it in the first place) In my view, simpler is better. not to mention being internally consistent. I think the svn directory name should match the package/catalog name. In the instance where multiple packages come out of a single svn directory, I think the directory should be named after the "main" package. (which we mostly already do anyway. eg: "curl" -> "curl_devel","curl_rt", "curl".) From bwalton at opencsw.org Tue Jun 29 19:05:19 2010 From: bwalton at opencsw.org (Ben Walton) Date: Tue, 29 Jun 2010 13:05:19 -0400 Subject: [csw-maintainers] A question of naming, in our svn repository In-Reply-To: References: Message-ID: <1277830994-sup-2098@pinkfloyd.chass.utoronto.ca> Excerpts from Philip Brown's message of Tue Jun 29 12:58:33 -0400 2010: > a) facilitate finding an svn directory from an "upstream" name > b) facilitate finding an svn directory from a package name. I think (b) makes the most sense, for the reasons you've given. It also works when you consider that there are a few oddball packages (docbook stuff, others?) that combine multiple upstreams into a single package. If nobody provides a better case for (a), we could move to standardizing the directory naming...say at the next update for a package, rename the directory if required? Thanks -Ben -- Ben Walton Systems Programmer - CHASS University of Toronto C:416.407.5610 | W:416.978.4302 From dam at opencsw.org Tue Jun 29 21:49:05 2010 From: dam at opencsw.org (Dagobert Michelsen) Date: Tue, 29 Jun 2010 21:49:05 +0200 Subject: [csw-maintainers] A question of naming, in our svn repository In-Reply-To: References: Message-ID: <9337133B-BD50-4A68-B261-2E3F2BCC779B@opencsw.org> Hi Phil, Am 29.06.2010 um 18:58 schrieb Philip Brown: > Maintainers, your input is required, over a matter of policy in naming > directories, in our subversion repository. > > We have two mutually exclusive ways to go: > a) facilitate finding an svn directory from an "upstream" name > b) facilitate finding an svn directory from a package name. > > Specific example, of the general case: > We have a package, "gtk_engines". > Should the subversion directory for it, be named "gtk-engines", or > "gtk_engines"? > > My thoughts on the issue are as follows: > > Of the people coming to our subversion repository, (both maintainers, > and non-maintainers) > almost no-one will care about case (a). On the contrary. See below. > They will mostly care about either "I want to take an existing OpenCSW > package, but recompile it with my own tweaks", or "I am going to take > over a package; I need to grab it from svn". > its all very well to say, "yes well the specific directory > information is in the > OPENCSW_REPOSITORY property", but that is a hassle to look for, and > most people will forget about that anyway (if they even knew it in the > first place) It is. And it is as simple as root at login [login]:/root > pkgparam CSWbash OPENCSW_REPOSITORY https://gar.svn.sf.net/svnroot/gar/csw/mgar/pkg/bash/trunk at 9406 Additionally, the link should be given on the package information webpage as browsable link, so it is fairly obvious where to look without any guessing at all. Additionally, there may be more than one package produced by a GAR recipe. Think of pm_subversion. Would you look in cpan/? Or where? It is in subversion/ which I personally find ok as there is no natural way to find the location of the GAR recipe besides from looking in the package. My most common usecase is to look for a recipe when the package has *not* been released and the only information I have is the upstream name as it has not been released as I said. > In my view, simpler is better. not to mention being internally > consistent. Consistency here is an illusion as there is no consistent name of a package from a recipe. > I think the svn directory name should match the > package/catalog name. In the instance where multiple packages come out > of a single svn directory, I think the directory should be named after > the "main" package. > (which we mostly already do anyway. eg: "curl" -> > "curl_devel","curl_rt", "curl".) Maybe you should outline what you would adjust. Changing gtk-stuff to gtk_stuff is no problem at all as the name is basically the same and both can be found by grepping for the components. Best regards -- Dago From phil at bolthole.com Tue Jun 29 22:24:52 2010 From: phil at bolthole.com (Philip Brown) Date: Tue, 29 Jun 2010 13:24:52 -0700 Subject: [csw-maintainers] A question of naming, in our svn repository In-Reply-To: <9337133B-BD50-4A68-B261-2E3F2BCC779B@opencsw.org> References: <9337133B-BD50-4A68-B261-2E3F2BCC779B@opencsw.org> Message-ID: On Tue, Jun 29, 2010 at 12:49 PM, Dagobert Michelsen wrote: > > Maybe you should outline what you would adjust. Changing > ?gtk-stuff > to > ?gtk_stuff > is no problem at all as the name is basically the same and both > can be found by grepping for the components. That is mostly all I'm saying. I've run across a whole lot of "legacy" entries in the gnome chain recently, that are like that, and in my opinion should be renamed gtk_stuff and gnome_stuff to match the packages they provide. Or at least *should* provide. More below. The vast majority of svn directories using "-" instead of our standard "_" would appear to be a quick run through by someone a long time ago, and mostly consist of directories like ./trunk ./trunk/legacy ./trunk/legacy/sources ./trunk/legacy/scripts ./trunk/legacy/specs ie: a bunch of "legacy" stuff, that hasnt even been packaged through svn, and does not follow our existing standards. You cant even type "make" in the directories. They did not produce the packages in our mirrors now; they are only non-functional placeholders, it would seem. From dam at opencsw.org Tue Jun 29 22:31:39 2010 From: dam at opencsw.org (Dagobert Michelsen) Date: Tue, 29 Jun 2010 22:31:39 +0200 Subject: [csw-maintainers] A question of naming, in our svn repository In-Reply-To: References: <9337133B-BD50-4A68-B261-2E3F2BCC779B@opencsw.org> Message-ID: <8FB41058-6BD3-4589-88CA-ED3C3CD349FE@opencsw.org> Hi Phil, Am 29.06.2010 um 22:24 schrieb Philip Brown: > On Tue, Jun 29, 2010 at 12:49 PM, Dagobert Michelsen > wrote: >> >> Maybe you should outline what you would adjust. Changing >> gtk-stuff >> to >> gtk_stuff >> is no problem at all as the name is basically the same and both >> can be found by grepping for the components. > > That is mostly all I'm saying. > I've run across a whole lot of "legacy" entries in the gnome chain > recently, that are like that, and in my opinion should be renamed > gtk_stuff and gnome_stuff to match the packages they provide. Or at > least *should* provide. More below. > > The vast majority of svn directories using "-" instead of our standard > "_" would appear to be > a quick run through by someone a long time ago, and mostly consist of > directories like > > ./trunk > ./trunk/legacy > ./trunk/legacy/sources > ./trunk/legacy/scripts > ./trunk/legacy/specs > > ie: a bunch of "legacy" stuff, that hasnt even been packaged through > svn, and does not follow our existing standards. You cant even type > "make" in the directories. The standard is to put non-GAR recipes in legacy, so they do conform to the standard ;-) > They did not produce the packages in our mirrors now; they are only > non-functional placeholders, it would seem. They have. Most of them are from previous (retired) maintainers who sent me their old build recipes for the packages they have released. To not loose them a second time I decided to commit them for reference if someone wants to take them over. If you start working on these please move /trunk/legacy to /tags/legacy I probably should have put it there from the start :-/ Best regards -- Dago From phil at bolthole.com Tue Jun 29 23:56:59 2010 From: phil at bolthole.com (Philip Brown) Date: Tue, 29 Jun 2010 14:56:59 -0700 Subject: [csw-maintainers] A question of naming, in our svn repository In-Reply-To: <8FB41058-6BD3-4589-88CA-ED3C3CD349FE@opencsw.org> References: <9337133B-BD50-4A68-B261-2E3F2BCC779B@opencsw.org> <8FB41058-6BD3-4589-88CA-ED3C3CD349FE@opencsw.org> Message-ID: On Tue, Jun 29, 2010 at 1:31 PM, Dagobert Michelsen wrote: > >> ie: a bunch of "legacy" stuff, that hasnt even been packaged through >> svn, and does not follow our existing standards. You cant even type >> "make" in the directories. > > The standard is to put non-GAR recipes in legacy, so they do conform > to the standard ;-) > If it's not documented, it's not a standard :) Nor is it a useful standard even if it were.. how do you use these "legacy" directories to actually make a package? From dam at opencsw.org Wed Jun 30 08:28:23 2010 From: dam at opencsw.org (Dagobert Michelsen) Date: Wed, 30 Jun 2010 08:28:23 +0200 Subject: [csw-maintainers] A question of naming, in our svn repository In-Reply-To: References: <9337133B-BD50-4A68-B261-2E3F2BCC779B@opencsw.org> <8FB41058-6BD3-4589-88CA-ED3C3CD349FE@opencsw.org> Message-ID: <4A430357-19CF-45CB-BB99-5F8E378E0A6A@opencsw.org> Hi Phil, Am 29.06.2010 um 23:56 schrieb Philip Brown: > On Tue, Jun 29, 2010 at 1:31 PM, Dagobert Michelsen > wrote: >> >>> ie: a bunch of "legacy" stuff, that hasnt even been packaged through >>> svn, and does not follow our existing standards. You cant even type >>> "make" in the directories. >> >> The standard is to put non-GAR recipes in legacy, so they do conform >> to the standard ;-) > > If it's not documented, it's not a standard :) > > Nor is it a useful standard even if it were.. how do you use these > "legacy" directories to actually make a package? You see what patches and options the previous maintainer used and then make a GAR recipe with similar options :-) I am currently struggling to rebuild x3270 against an updated libicu and my package looks different from the existing one - and there is no recipe for me to look at at all. Best regards -- Dago From maciej at opencsw.org Wed Jun 30 17:22:18 2010 From: maciej at opencsw.org (Maciej (Matchek) Blizinski) Date: Wed, 30 Jun 2010 16:22:18 +0100 Subject: [csw-maintainers] Package stats collection / retrieval in checkpkg Message-ID: [discussion moved from pkgsubmissions] No dia 29 de Junho de 2010 16:47, Philip Brown escreveu: > On Mon, Jun 28, 2010 at 3:53 PM, Maciej (Matchek) Blizinski > wrote: >> >>> What did you have in mind? >> >> I was thinking about exactly that - the package build level. ?The >> earlier a problem can be detected, the better, right? > > I thought you might :) > >> ?For example: >> >> maciej at current9s :~/src/opencsw/gar/gar-new-checks > bin/checkpkg >> ~/exp/*cups*sparc* >> Analyzing collected statistics... >> CSWcupsdev: >> CSWcupsd: >> CSWcupsclient: >> CSWlibcups: >> (...) > > And where are these statistics "collected"? I wrote it up at http://wiki.opencsw.org/checkpkg#toc17 ?-- I've just updated the 'Stats collection' section. > Might it be better to define an API for accessing the "official" > filename path repository, rather than hoping the package has been seen > by that build machine before? > > Ideally, some API that allows for a common shared local cached copy, > rather than "every maintainer has their own cache" ? I never thought of it as something that needs caching. ?The separation between stats collection and analysis exists mainly because it makes the development easier. ?Operationally it doesn't make that much difference for one maintainer. ?The positive impact in operational terms is the biggest when analyzing the whole catalog. There are also data about the system files and the catalog, which are in a sqlite database, which caches the content of /var/sadm/install/contents. ?This might be beneficial to share, but it requires careful planning and needs to armored against race conditions during updates. ?It's something I intend to do, but it'll take time. From william at wbonnet.net Wed Jun 30 22:21:51 2010 From: william at wbonnet.net (William Bonnet) Date: Wed, 30 Jun 2010 22:21:51 +0200 Subject: [csw-maintainers] [gar] RFE : display all prerequisitepkg at once Message-ID: <4C2BA75F.5090003@wbonnet.net> Hi, I would like to add for a mgar evolution. Current version checks for missing prerequisite packages before compilation and display missing packages. Example : [william at cyaegha:~/community/opencsw/mgar/pkg/seamonkey/trunk]$ gmake package [===== NOW BUILDING: seamonkey =====] ==> Verifying for installed package CSWpy-progressbar: MISSING gmake: *** [prerequisitepkg-CSWpy-progressbar] Erreur 1 Would it be possible to display *all* the missing packages instead of outputing the first missing package and exiting ? cheers W. -- William http://www.wbonnet.net http://www.sunwizard.net Le site fran?ais des amateurs de stations Unix http://www.opencsw.org Community SoftWare for Solaris http://www.guses.org French speaking Solaris User Group From bwalton at opencsw.org Wed Jun 30 22:31:10 2010 From: bwalton at opencsw.org (Ben Walton) Date: Wed, 30 Jun 2010 16:31:10 -0400 Subject: [csw-maintainers] [gar] RFE : display all prerequisitepkg at once In-Reply-To: <4C2BA75F.5090003@wbonnet.net> References: <4C2BA75F.5090003@wbonnet.net> Message-ID: <1277929832-sup-5707@pinkfloyd.chass.utoronto.ca> Excerpts from William Bonnet's message of Wed Jun 30 16:21:51 -0400 2010: Hi William, > Would it be possible to display *all* the missing packages instead > of outputing the first missing package and exiting ? I wonder if we shouldn't just have it look for CSWgar-dev, which depends on all gar build deps and leave it at that? Thanks -Ben -- Ben Walton Systems Programmer - CHASS University of Toronto C:416.407.5610 | W:416.978.4302 From phil at bolthole.com Wed Jun 30 23:03:28 2010 From: phil at bolthole.com (Philip Brown) Date: Wed, 30 Jun 2010 14:03:28 -0700 Subject: [csw-maintainers] Fwd: cups packages batched.. no wait... In-Reply-To: References: Message-ID: aaaah! you said you had changed lists but you didn't :p ---------- Forwarded message ---------- From: Philip Brown Date: Wednesday, June 30, 2010 Subject: cups packages batched.. no wait... To: "pkgsubmissions at lists.opencsw.org" it shouldn't take that much time; define the sqlite database as a publically writable location, have updates be done by "download the latest db and then mv it into place" which is atomic. only thing left is to define export interface from our official file path database. you could actually reduce it's code a lot since you won't have to extract and parse pkgs yourself :)