From maciej at opencsw.org Sun Jun 1 12:37:20 2014 From: maciej at opencsw.org (=?UTF-8?Q?Maciej_=28Matchek=29_Blizi=C5=84ski?=) Date: Sun, 1 Jun 2014 11:37:20 +0100 Subject: Rules for automatic package promotions In-Reply-To: References: <18E7D54B-29F2-4959-BE4B-F0D88A45BBB4@opencsw.org> Message-ID: New question about cross-catalog logic. Let's say we have package CSWfoo which is in 4 catalogs: sparc-SunOS5.9 sparc-SunOS5.10 i386-SunOS5.9 i386-SunOS5.10 It can be promoted from unstable to testing in sparc-SunOS5.9, but not sparc-SunOS5.10. Should it be promoted in sparc-SunOS5.9? In other words, should the automation keep package promotions in sync across catalogs? Maciej From pfelecan at opencsw.org Sun Jun 1 16:14:24 2014 From: pfelecan at opencsw.org (Peter FELECAN) Date: Sun, 01 Jun 2014 16:14:24 +0200 Subject: Rules for automatic package promotions In-Reply-To: ("Maciej (Matchek) =?utf-8?Q?Blizi=C5=84ski=22's?= message of "Sun, 1 Jun 2014 11:37:20 +0100") References: <18E7D54B-29F2-4959-BE4B-F0D88A45BBB4@opencsw.org> Message-ID: "Maciej (Matchek) Blizi?ski" writes: > New question about cross-catalog logic. Let's say we have package > CSWfoo which is in 4 catalogs: > > sparc-SunOS5.9 > sparc-SunOS5.10 > i386-SunOS5.9 > i386-SunOS5.10 > > It can be promoted from unstable to testing in sparc-SunOS5.9, but not > sparc-SunOS5.10. Should it be promoted in sparc-SunOS5.9? > > In other words, should the automation keep package promotions in sync > across catalogs? If possible yes. Especially for the same OS level. But, IMHO it's not a strong requirement as we dropped support for 5.9, isn't it? Consequently, if it's simple, keep in sync and, if it's complicated, don't. -- Peter From maciej at opencsw.org Mon Jun 2 12:27:04 2014 From: maciej at opencsw.org (=?UTF-8?Q?Maciej_=28Matchek=29_Blizi=C5=84ski?=) Date: Mon, 2 Jun 2014 11:27:04 +0100 Subject: Bugs in the package browser; user names are mismatched In-Reply-To: References: Message-ID: On Mon, Apr 28, 2014 at 9:16 AM, Maciej (Matchek) Blizi?ski wrote: > So... nobody knows? I was hoping Ben would know. For now we'll impose 8-character length limit on user names. I will try to rename the existing ones when I get around to it; I'm working on something else now. Maciej From slowfranklin at opencsw.org Mon Jun 2 12:29:38 2014 From: slowfranklin at opencsw.org (slowfranklin) Date: Mon, 2 Jun 2014 12:29:38 +0200 Subject: Bugs in the package browser; user names are mismatched In-Reply-To: References: Message-ID: <59694838-C330-4A9F-85DF-B9CB71258A4C@opencsw.org> Am 02.06.2014 um 12:27 schrieb Maciej (Matchek) Blizi?ski : > On Mon, Apr 28, 2014 at 9:16 AM, Maciej (Matchek) Blizi?ski > wrote: >> So... nobody knows? I was hoping Ben would know. > > For now we'll impose 8-character length limit on user names. I will > try to rename the existing ones when I get around to it; I'm working > on something else now. Another system where I become famously known as slowfran. :) Thanks! -slow From yann at pleiades.fr.eu.org Tue Jun 3 08:07:23 2014 From: yann at pleiades.fr.eu.org (Yann Rouillard) Date: Tue, 3 Jun 2014 08:07:23 +0200 Subject: GNUTLS needs libc 1.22.5 now !? In-Reply-To: <5382CE6D.70702@opencsw.org> References: <537A2465.3020902@opencsw.org> <5382CE6D.70702@opencsw.org> Message-ID: Any objection ? Yann 2014-05-26 7:17 GMT+02:00 Jan Holzh?ter : > Hi, > > > > > BTW, shouldn't we now upgrade the minimal libc version interface in > > OpenCSW ? > > 1.22.5 was brought in Solaris 10 update 8 which is nearly 5 years old > > now and mysql already depends on it. > > We probably should. > > Greetings > Jan > > > > -------------- next part -------------- An HTML attachment was scrubbed... URL: From laurent at opencsw.org Wed Jun 4 21:47:32 2014 From: laurent at opencsw.org (Laurent Blume) Date: Wed, 04 Jun 2014 21:47:32 +0200 Subject: Checkpkg insists that a library is surplus, I think it is not Message-ID: <538F77D4.9050900@opencsw.org> Hello, With the ImageMagick recipe, between a beta two weeks ago and now, the behaviour of checkpkg has changed, it has now decided that librsvg is not needed: * Dependency issues of CSWimagemagick: * If you don't know of any reasons to include these dependencies, you might remove them: * ? CSWlibrsvg2-2 That seems really weird: the library was properly detected and used, dump -Lv shows it, so why would checkpkg decide it is not needed? ./work/solaris10-i386/pkgroot/opt/csw/lib/ImageMagick-6.8.9/modules-Q16HDRI/coders/svg.so: **** DYNAMIC SECTION INFORMATION **** .dynamic: [INDEX] Tag Value [1] NEEDED libMagickCore-6.Q16HDRI.so.2 [2] NEEDED librsvg-2.so.2 Can somebody explain that? Thanks, Laurent From maciej at opencsw.org Thu Jun 5 11:07:24 2014 From: maciej at opencsw.org (=?UTF-8?Q?Maciej_=28Matchek=29_Blizi=C5=84ski?=) Date: Thu, 5 Jun 2014 10:07:24 +0100 Subject: Checkpkg insists that a library is surplus, I think it is not In-Reply-To: <538F77D4.9050900@opencsw.org> References: <538F77D4.9050900@opencsw.org> Message-ID: On Wed, Jun 4, 2014 at 8:47 PM, Laurent Blume wrote: > * Dependency issues of CSWimagemagick: > * If you don't know of any reasons to include these dependencies, you might > remove them: > * ? CSWlibrsvg2-2 > > That seems really weird: the library was properly detected and used, dump > -Lv shows it, so why would checkpkg decide it is not needed? > > ./work/solaris10-i386/pkgroot/opt/csw/lib/ImageMagick-6.8.9/modules-Q16HDRI/coders/svg.so: > > **** DYNAMIC SECTION INFORMATION **** > .dynamic: > [INDEX] Tag Value > [1] NEEDED libMagickCore-6.Q16HDRI.so.2 > [2] NEEDED librsvg-2.so.2 > > Can somebody explain that? I have a guess. checkpkg does the following: Q: "What does the binary need?" A: "librsvg-2.so.2" Q: "Which package provides librsvg-2.so.2?" A: "SUNWgnome-base-libs and CSW... no, actually, no CSW* package provides librsvg-2.so.2." The CSWlibrsvg2-2 package only provides librsvg-2.so.2.36.4 (!= librsvg-2.so.2). Maciej -------------- next part -------------- An HTML attachment was scrubbed... URL: From laurent at opencsw.org Thu Jun 5 11:24:10 2014 From: laurent at opencsw.org (Laurent Blume) Date: Thu, 05 Jun 2014 11:24:10 +0200 Subject: Checkpkg insists that a library is surplus, I think it is not In-Reply-To: References: <538F77D4.9050900@opencsw.org> Message-ID: <5390373A.8000608@opencsw.org> Le 2014/06/05 11:13 +0200, Matchek a ?crit: > I have a guess. > > checkpkg does the following: > > Q: "What does the binary need?" > A: "librsvg-2.so.2" > > Q: "Which package provides librsvg-2.so.2?" > A: "SUNWgnome-base-libs and CSW... no, actually, no CSW* package > provides librsvg-2.so.2." > > The CSWlibrsvg2-2 package only provides librsvg-2.so.2.36.4 (!= > librsvg-2.so.2). Err.... what? $ pkgchk -lp /opt/csw/lib/librsvg-2.so.2 NOTE: Couldn't lock the package database. Pathname: /opt/csw/lib/librsvg-2.so.2 Type: symbolic link Source of link: librsvg-2.so.2.36.4 Referenced by the following packages: CSWlibrsvg2-2 Current status: installed And that's not new, it's been like that for a while. So I don't think that explanation is the right one :-) Laurent From maciej at opencsw.org Thu Jun 5 12:23:12 2014 From: maciej at opencsw.org (=?UTF-8?Q?Maciej_=28Matchek=29_Blizi=C5=84ski?=) Date: Thu, 5 Jun 2014 11:23:12 +0100 Subject: Checkpkg insists that a library is surplus, I think it is not In-Reply-To: <5390373A.8000608@opencsw.org> References: <538F77D4.9050900@opencsw.org> <5390373A.8000608@opencsw.org> Message-ID: On Thu, Jun 5, 2014 at 10:24 AM, Laurent Blume wrote: > And that's not new, it's been like that for a while. So I don't think that explanation is the right one :-) The explanation is correct in the sense that this is what checkpkg thinks. checkpkg could be wrong for some reason. Here's how checkpkg looks at it and what it sees: $ whichpkgprovides() { curl -s "http://buildfarm.opencsw.org/pkgdb/rest/catalogs/unstable/sparc/SunOS5.10/pkgnames-and-paths-by-basename?basename=$1"; echo; } $ whichpkgprovides libz.so.1 {"/opt/csw/lib/sparcv9": ["CSWlibz1"], "/opt/csw/lib/sparcv8plus+vis": ["CSWlibz1"], "/usr/lib/sparcv9": ["SUNWzlib"], "/usr/lib": ["SUNWzlib"], "/opt/csw/lib": ["CSWlibz1"]} $ whichpkgprovides librsvg-2.so.2 {"/usr/lib": ["SUNWgnome-base-libs"]} Maybe something's wrong with indexing. Looking at http://buildfarm.opencsw.org/pkgdb/srv4/a417528454206ad899beb18e1ad1737c/ I see: Registration Level 2: False This is wrong, because the package is not registered (its files are not visible to checkpkg) and it's in a catalog. For now, could you re-push the CSWlibrsvg2-2 package? In the meantime, I'll think about where the bug can be. Maciej From maciej at opencsw.org Thu Jun 5 12:53:46 2014 From: maciej at opencsw.org (=?UTF-8?Q?Maciej_=28Matchek=29_Blizi=C5=84ski?=) Date: Thu, 5 Jun 2014 11:53:46 +0100 Subject: Checkpkg insists that a library is surplus, I think it is not In-Reply-To: References: <538F77D4.9050900@opencsw.org> <5390373A.8000608@opencsw.org> Message-ID: On Thu, Jun 5, 2014 at 11:23 AM, Maciej (Matchek) Blizi?ski wrote: > For now, could you re-push the CSWlibrsvg2-2 package? In the meantime, > I'll think about where the bug can be. I registered a few by hand to see if this has an effect: curl --netrc -X PUT http://buildfarm.opencsw.org/releases/svr4/7cb32f2d4a23841cc23277015620040a/db-level-2/; echo {"message": "Package registered to level 2"} It does have an effect. Maybe csw-upload-pkg has a bug where it forgets to make this call, or mistakenly thinks that it's not necessary. From laurent at opencsw.org Thu Jun 5 14:06:42 2014 From: laurent at opencsw.org (Laurent Blume) Date: Thu, 05 Jun 2014 14:06:42 +0200 Subject: Checkpkg insists that a library is surplus, I think it is not In-Reply-To: References: <538F77D4.9050900@opencsw.org> <5390373A.8000608@opencsw.org> Message-ID: <53905D52.4080601@opencsw.org> Le 2014/06/05 14:06 +0200, Matchek a ?crit: > I registered a few by hand to see if this has an effect: > > curl --netrc -X PUT > http://buildfarm.opencsw.org/releases/svr4/7cb32f2d4a23841cc23277015620040a/db-level-2/; > echo > {"message": "Package registered to level 2"} > > It does have an effect. Maybe csw-upload-pkg has a bug where it > forgets to make this call, or mistakenly thinks that it's not > necessary. At least now it looks all good for me - thanks! Laurent From maciej at opencsw.org Thu Jun 5 23:18:51 2014 From: maciej at opencsw.org (=?UTF-8?Q?Maciej_=28Matchek=29_Blizi=C5=84ski?=) Date: Thu, 5 Jun 2014 22:18:51 +0100 Subject: SF.net SVN: gar:[23769] csw/mgar/pkg/openssl/trunk/Makefile In-Reply-To: <3gkz1L0MHWz1M5@mail.opencsw.org> References: <3gkz1L0MHWz1M5@mail.opencsw.org> Message-ID: On Thu, Jun 5, 2014 at 9:22 PM, wrote: > +# False alert > +CHECKPKG_OVERRIDES_CSWopenssl-utils += file-needed-but-no-package-satisfies-it|/etc/opt/csw/ssl/openssl.cnf|CSWopenssl-utils|contains|symlink|/opt/csw/ssl/openssl.cnf|which|needs|the|target|file:|/etc/opt/csw/ssl/openssl.cnf Hm, I tend to think that there is some legitimacy to this alert. It's not just a spurious false alert out of nowhere. After all, the package does contain a dangling symlink, as far as pkgmap is concerned, that's a fact. :-) Maybe the class action script should create the /opt/csw/etc symlinks? That would keep all this logic in once place. Maciej From yann at pleiades.fr.eu.org Fri Jun 6 01:35:04 2014 From: yann at pleiades.fr.eu.org (Yann Rouillard) Date: Fri, 6 Jun 2014 01:35:04 +0200 Subject: SF.net SVN: gar:[23769] csw/mgar/pkg/openssl/trunk/Makefile In-Reply-To: References: <3gkz1L0MHWz1M5@mail.opencsw.org> Message-ID: Hi Maciej, We already talked about this case on IRC: the file /etc/opt/csw/ssl/openssl.cnf is provided by CSWopenssl-utils but it is registered as /etc/opt/csw/ssl/openssl.cnf.CSW in the pkgmap of the package as the real /etc/opt/csw/ssl/openssl.cnf is created at install time using the cswpreserveconf class. So it's indeed a false alarm as there is a package that provides this file (the package itself) and there will be no dangling symlink when the package is installed. That can probably be fixed in a another way (but I am not sure I understood your proposition, is it about creating a custom class action script to handle the creation of the symlink /opt/csw/ssl/openssl.cnf -> /etc/opt/csw/ssl/openssl.cnf ?). But for now I have no problem overriding this check. Anyway your comment was still useful, it made me notice that the openssl 0.9.8 recipe should simply not contain anymore the directives that create the openssl-utils and libssl-dev packages, they are generated by the openssl 1.0.1 recipe (which contains the same override). Yann 2014-06-05 23:18 GMT+02:00 Maciej (Matchek) Blizi?ski : > On Thu, Jun 5, 2014 at 9:22 PM, wrote: > > +# False alert > > +CHECKPKG_OVERRIDES_CSWopenssl-utils += > file-needed-but-no-package-satisfies-it|/etc/opt/csw/ssl/openssl.cnf|CSWopenssl-utils|contains|symlink|/opt/csw/ssl/openssl.cnf|which|needs|the|target|file:|/etc/opt/csw/ssl/openssl.cnf > > Hm, I tend to think that there is some legitimacy to this alert. It's > not just a spurious false alert out of nowhere. After all, the package > does contain a dangling symlink, as far as pkgmap is concerned, that's > a fact. :-) > > Maybe the class action script should create the /opt/csw/etc symlinks? > That would keep all this logic in once place. > > Maciej > -------------- next part -------------- An HTML attachment was scrubbed... URL: From maciej at opencsw.org Sat Jun 7 20:00:17 2014 From: maciej at opencsw.org (=?UTF-8?Q?Maciej_=28Matchek=29_Blizi=C5=84ski?=) Date: Sat, 7 Jun 2014 19:00:17 +0100 Subject: Rules for automatic package promotions In-Reply-To: References: <18E7D54B-29F2-4959-BE4B-F0D88A45BBB4@opencsw.org> Message-ID: On Sun, Jun 1, 2014 at 3:14 PM, Peter FELECAN wrote: > If possible yes. Especially for the same OS level. But, IMHO it's not a > strong requirement as we dropped support for 5.9, isn't it? > Consequently, if it's simple, keep in sync and, if it's complicated, > don't. It's not hard, it's only about how do we want to structure the data. I've made some progress. The program will output a HTML report, here's what I have so far: http://buildfarm.opencsw.org/~maciej/promote-packages.html flex is a comical example. libfl is missing from testing, but won't get promoted. Why? Because it has a major bug in mantis filed for it. What's the bug? The bug is that libfl is missing from testing. Maciej From maciej at opencsw.org Sat Jun 7 20:08:11 2014 From: maciej at opencsw.org (=?UTF-8?Q?Maciej_=28Matchek=29_Blizi=C5=84ski?=) Date: Sat, 7 Jun 2014 19:08:11 +0100 Subject: SF.net SVN: gar:[23769] csw/mgar/pkg/openssl/trunk/Makefile In-Reply-To: References: <3gkz1L0MHWz1M5@mail.opencsw.org> Message-ID: On Fri, Jun 6, 2014 at 12:35 AM, Yann Rouillard wrote: > Hi Maciej, > > We already talked about this case on IRC: the file > /etc/opt/csw/ssl/openssl.cnf is provided by CSWopenssl-utils but it is > registered as /etc/opt/csw/ssl/openssl.cnf.CSW in the pkgmap of the package > as the real /etc/opt/csw/ssl/openssl.cnf is created at install time using > the cswpreserveconf class. > So it's indeed a false alarm as there is a package that provides this file > (the package itself) and there will be no dangling symlink when the package > is installed. We can argue about what's a false alarm and what isn't, what I'm saying is that if you look at the contents of the package, that contents is bad (dangling symlink is in fact there), and that bug is corrected by the postinstall script. In an exaggerated example, you could have a package that has a totally messed up content, and then postinstall kicks in and patches and fixes everything. That would be wrong, wouldn't it? This dangling symlink is a mild example of that. > That can probably be fixed in a another way (but I am not sure I understood > your proposition, is it about creating a custom class action script to > handle the creation of the symlink /opt/csw/ssl/openssl.cnf -> > /etc/opt/csw/ssl/openssl.cnf ?). CAS creates /etc/opt/csw/ssl/openssl.cnf, right? So why shouldn't it create /opt/opt/etc/ssl/openssl.cnf too? Maciej From yann at pleiades.fr.eu.org Tue Jun 10 14:09:05 2014 From: yann at pleiades.fr.eu.org (Yann Rouillard) Date: Tue, 10 Jun 2014 14:09:05 +0200 Subject: GNUTLS needs libc 1.22.5 now !? In-Reply-To: References: <537A2465.3020902@opencsw.org> <5382CE6D.70702@opencsw.org> Message-ID: No objection, so I did it ! http://sourceforge.net/p/gar/code/23780/ Yann 2014-06-03 8:07 GMT+02:00 Yann Rouillard : > Any objection ? > > Yann > > > 2014-05-26 7:17 GMT+02:00 Jan Holzh?ter : > > Hi, >> >> > >> > BTW, shouldn't we now upgrade the minimal libc version interface in >> > OpenCSW ? >> > 1.22.5 was brought in Solaris 10 update 8 which is nearly 5 years old >> > now and mysql already depends on it. >> >> We probably should. >> >> Greetings >> Jan >> >> >> >> > -------------- next part -------------- An HTML attachment was scrubbed... URL: From dam at opencsw.org Tue Jun 10 14:10:52 2014 From: dam at opencsw.org (Dagobert Michelsen) Date: Tue, 10 Jun 2014 14:10:52 +0200 Subject: GNUTLS needs libc 1.22.5 now !? In-Reply-To: References: <537A2465.3020902@opencsw.org> <5382CE6D.70702@opencsw.org> Message-ID: <93D4C723-85A0-48F2-B7A7-8FB39ED25EA6@opencsw.org> Hi Yann, Am 10.06.2014 um 14:09 schrieb Yann Rouillard : > No objection, so I did it ! http://sourceforge.net/p/gar/code/23780/ I am currently working on Nettle 3.0, could you please track GNUTLS on nettle compatibility? There should be an update soon and I would like to coordinate the update effort. Best regards ? Dago -- "You don't become great by trying to be great, you become great by wanting to do something, and then doing it so hard that you become great in the process." - xkcd #896 -------------- next part -------------- A non-text attachment was scrubbed... Name: smime.p7s Type: application/pkcs7-signature Size: 2418 bytes Desc: not available URL: From yann at pleiades.fr.eu.org Tue Jun 10 14:17:49 2014 From: yann at pleiades.fr.eu.org (Yann Rouillard) Date: Tue, 10 Jun 2014 14:17:49 +0200 Subject: GNUTLS needs libc 1.22.5 now !? In-Reply-To: <93D4C723-85A0-48F2-B7A7-8FB39ED25EA6@opencsw.org> References: <537A2465.3020902@opencsw.org> <5382CE6D.70702@opencsw.org> <93D4C723-85A0-48F2-B7A7-8FB39ED25EA6@opencsw.org> Message-ID: No problem, I'll check that. Yann 2014-06-10 14:10 GMT+02:00 Dagobert Michelsen : > Hi Yann, > > Am 10.06.2014 um 14:09 schrieb Yann Rouillard : > > No objection, so I did it ! http://sourceforge.net/p/gar/code/23780/ > > I am currently working on Nettle 3.0, could you please track GNUTLS on > nettle compatibility? There should be an update soon and I would like to > coordinate the update effort. > > > Best regards > > ? Dago > > -- > "You don't become great by trying to be great, you become great by wanting > to do something, > and then doing it so hard that you become great in the process." - xkcd > #896 > > -------------- next part -------------- An HTML attachment was scrubbed... URL: From maciej at opencsw.org Tue Jun 10 18:36:33 2014 From: maciej at opencsw.org (=?UTF-8?Q?Maciej_=28Matchek=29_Blizi=C5=84ski?=) Date: Tue, 10 Jun 2014 17:36:33 +0100 Subject: Uploading to non-unstable catalogs Message-ID: Sometimes users ask about some packages being or not being there in testing (now: bratislava), or stable (now: kiel). I used to do this periodically, but I don't do that any more, and I don't plan on doing that again. Q: Can maintainers upload to bratislava and/or kiel? A: Yes. Q: How? A: You pass the --catalog-release flag to csw-upload-pkg. The upload procedure works as usual: checkpkg checks are run, and your packages are inserted into the chosen catalog. So if you want something in testing, don't wait. Upload it to testing yourself. Note: --catalog-release=testing will not work. You must give the script the release name. Right now it is: csw-upload-pkg --catalog-release bratislava ... Maciej -------------- next part -------------- An HTML attachment was scrubbed... URL: From rmottola at opencsw.org Wed Jun 11 08:34:24 2014 From: rmottola at opencsw.org (Riccardo Mottola) Date: Wed, 11 Jun 2014 08:34:24 +0200 Subject: first gnustep app! but segfaults In-Reply-To: <53864EE9.7050909@opencsw.org> References: <53864EE9.7050909@opencsw.org> Message-ID: <5397F870.5060706@opencsw.org> Hi, Riccardo Mottola wrote: > ON the same machine, I have a parallel GNUstep installation in another > tree, built entirely from source without packages, there FTP works! > (granted, it is svn trunk and not released, but that shouldn't justify > a crash) This is not true any longer. I updated "pkgutil -u all" and then updated my GNUstep installation which I build totally outside the mgar system, which worked before I attempted packaging, I posistively had some applications running. Instead, now, the applications crash and I get a stacktrace which is totally outside the application and gnustep's code. Look below. Do you have any hints? Riccardo [Switching to Thread 1 (LWP 1)] 0xfdb41748 in __1cDstdM__node_alloc4BiA_L_M_allocate6FI_pv_ () from /usr/lib/libaspell.so.15 (gdb) bt #0 0xfdb41748 in __1cDstdM__node_alloc4BiA_L_M_allocate6FI_pv_ () from /usr/lib/libaspell.so.15 #1 0xfdb4191c in __1cDstdMbasic_string4Ccn0ALchar_traits4Cc__n0AJallocator4Cc___2t6M_v_ () from /usr/lib/libaspell.so.15 #2 0xfdb3ef40 in __1cDstdIios_base2t6M_v_ () from /usr/lib/libaspell.so.15 #3 0xfdb42774 in __1cDstdJbasic_ios4Ccn0ALchar_traits4Cc___2t6M_v_ () from /usr/lib/libaspell.so.15 #4 0xfb4eab04 in __1cDstdNbasic_istream4Ccn0ALchar_traits4Cc___2t6Mpn0APbasic_streambuf4Ccn0B____v_ () from /usr/lib/libCstd.so.1 #5 0xfb57992c in __1cDstdIios_baseEInit2t6M_v_ () from /usr/lib/libCstd.so.1 #6 0xfb57a51c in __SUNW_init_iostreams () from /usr/lib/libCstd.so.1 #7 0xfb590a6c in ?? () from /usr/lib/libCstd.so.1 Backtrace stopped: previous frame identical to this frame (corrupt stack?) (gdb) q From dam at opencsw.org Wed Jun 11 09:57:15 2014 From: dam at opencsw.org (Dagobert Michelsen) Date: Wed, 11 Jun 2014 09:57:15 +0200 Subject: first gnustep app! but segfaults In-Reply-To: <5397F870.5060706@opencsw.org> References: <53864EE9.7050909@opencsw.org> <5397F870.5060706@opencsw.org> Message-ID: Hi Riccardo, Am 11.06.2014 um 08:34 schrieb Riccardo Mottola : > Riccardo Mottola wrote: >> ON the same machine, I have a parallel GNUstep installation in another tree, built entirely from source without packages, there FTP works! (granted, it is svn trunk and not released, but that shouldn't justify a crash) > This is not true any longer. I updated "pkgutil -u all" and then updated my GNUstep installation which I build totally outside the mgar system, which worked before I attempted packaging, I posistively had some applications running. > > Instead, now, the applications crash and I get a stacktrace which is totally outside the application and gnustep's code. Look below. > Do you have any hints? You can pipe it through /opt/SUNWspro/bin/c++filt, this unmangles the c++ symbols: [Switching to Thread 1 (LWP 1)] 0xfdb41748 in void*std::__node_alloc::_M_allocate(unsigned) () from /usr/lib/libaspell.so.15 (gdb) bt #0 0xfdb41748 in void*std::__node_alloc::_M_allocate(unsigned) () from /usr/lib/libaspell.so.15 #1 0xfdb4191c in std::string::basic_string() () from /usr/lib/libaspell.so.15 #2 0xfdb3ef40 in std::ios_base::ios_base() () from /usr/lib/libaspell.so.15 #3 0xfdb42774 in std::ios::basic_ios() () from /usr/lib/libaspell.so.15 #4 0xfb4eab04 in std::istream::basic_istream(std::streambuf *) () from /usr/lib/libCstd.so.1 #5 0xfb57992c in std::ios_base::Init::Init() () from /usr/lib/libCstd.so.1 #6 0xfb57a51c in __SUNW_init_iostreams () from /usr/lib/libCstd.so.1 #7 0xfb590a6c in ?? () from /usr/lib/libCstd.so.1 Backtrace stopped: previous frame identical to this frame (corrupt stack?) (gdb) q I am wondering if it is correct that you use /usr/lib/libaspell.so.15 instead of /opt/csw/lib/libaspell.so.15. Best regards ? Dago -- "You don't become great by trying to be great, you become great by wanting to do something, and then doing it so hard that you become great in the process." - xkcd #896 -------------- next part -------------- A non-text attachment was scrubbed... Name: smime.p7s Type: application/pkcs7-signature Size: 2418 bytes Desc: not available URL: From rmottola at opencsw.org Wed Jun 11 20:43:35 2014 From: rmottola at opencsw.org (Riccardo Mottola) Date: Wed, 11 Jun 2014 20:43:35 +0200 Subject: first gnustep app! but segfaults In-Reply-To: References: <53864EE9.7050909@opencsw.org> <5397F870.5060706@opencsw.org> Message-ID: <5398A357.2000509@opencsw.org> Hi Dagobert, I'll try filtering the error, but I wonder why it is at all happening, I will check what is linked on solars and on another platfomr and see where libaspell gets pulled in. Dagobert Michelsen wrote: > > I am wondering if it is correct that you use /usr/lib/libaspell.so.15 instead of > /opt/csw/lib/libaspell.so.15. Do you think this could cause trouble? How may I try to change the link order? The strange thing is that I got it all working before... at least outisde the mgar packages. Riccardo From grzemba at contac-dt.de Mon Jun 16 12:20:40 2014 From: grzemba at contac-dt.de (Carsten Grzemba) Date: Mon, 16 Jun 2014 12:20:40 +0200 Subject: somebody fills up our OSQA user database In-Reply-To: References: Message-ID: since our OSQA user database was filledup with a some logins which probably not interessted in OpenCSW I have change the registration options: - local registration is not possible anymore, existing logins still works, - new registration only allowed via OpenId providers like Google and OpenId as well as Facebook and Twitter. Carsten Am 13.06.14 schrieb Maciej (Matchek) Blizi?ski : > It's a losing battle. > > We should only allow logins via Facebook, Twitter and Google. > > > > On Fri, Jun 13, 2014 at 6:29 AM, Carsten Grzemba wrote: > > > Oh shit, now all how wants loose your weight are our friends. > > What is this? > > > > Am 12.06.14 schrieb Maciej (Matchek) Blizi?ski : > > > > > > You did the right thing. > > > > > > OSQA doesn't have good anti spam mechanisms. It will be increasingly difficult and painful to manage spam there. > > > Em 12/06/2014 09:24, "Carsten Grzemba" escreveu: > > > > > > > Hi, > > > > > > > > since some days we got a lot of new users on OSQA with a similar naming scheme and the loginname is never similar with the email address. But very time the email address is for 'outlook.com(http://outlook.com)' > > > > I guess this users are not interested in OpenCSW and potential spammers. > > > > > > > > In the moment I have blocked the 'outlook.com(http://outlook.com)' domain in: > > > > forum/forms/general.py > > > > > > > > Or should I not? > > > > > > > > Carsten > > > > > > > > > > > > > > > > > > > > -------------- next part -------------- An HTML attachment was scrubbed... URL: From rmottola at opencsw.org Thu Jun 19 01:58:39 2014 From: rmottola at opencsw.org (Riccardo Mottola) Date: Thu, 19 Jun 2014 01:58:39 +0200 Subject: first gnustep app! but segfaults In-Reply-To: References: <53864EE9.7050909@opencsw.org> <5397F870.5060706@opencsw.org> Message-ID: <53A227AF.2090200@opencsw.org> Hi, Dagobert Michelsen wrote: > I am wondering if it is correct that you use /usr/lib/libaspell.so.15 instead of > /opt/csw/lib/libaspell.so.15. I found an option ti disable aspell integration and thus, together with a couple of other fixes, I got my apps starting again, at least in my source tree, not yet in mgar. Could this be an LDFLAGS problem. SInce I get a crash, I think it is not a just a problem with the system aspell, but something like being compiled against one aspell and then run against the other. Riccardo From rmottola at opencsw.org Sat Jun 21 13:00:18 2014 From: rmottola at opencsw.org (Riccardo Mottola) Date: Sat, 21 Jun 2014 13:00:18 +0200 Subject: adding a patch Message-ID: <1403348416.7769.3.camel@anor> Hi, I have written my first patch to a package. Now when I try to commit with mgar, it tells me "working directory contains untracked changes" which is of course true. How do I commit the patch? it resides in "trunk/files" Riccardo From slowfranklin at opencsw.org Sat Jun 21 13:05:04 2014 From: slowfranklin at opencsw.org (slowfranklin) Date: Sat, 21 Jun 2014 13:05:04 +0200 Subject: adding a patch In-Reply-To: <1403348416.7769.3.camel@anor> References: <1403348416.7769.3.camel@anor> Message-ID: <9BA41247-2D3A-4CFF-996B-B88F579E7A3E@opencsw.org> Hi Riccardo Am 21.06.2014 um 13:00 schrieb Riccardo Mottola : > Hi, > > I have written my first patch to a package. Now when I try to commit > with mgar, it tells me "working directory contains untracked changes" > which is of course true. > > How do I commit the patch? it resides in "trunk/files" if you have *new* files in trunk/files you must first call `svn add FILE` on them. Afterwards you can commit with mgar commit which will commit the new files added before as well as any other modified file. Hth! -slow From dam at opencsw.org Wed Jun 25 10:53:17 2014 From: dam at opencsw.org (Dagobert Michelsen) Date: Wed, 25 Jun 2014 10:53:17 +0200 Subject: first gnustep app! but segfaults In-Reply-To: <53A227AF.2090200@opencsw.org> References: <53864EE9.7050909@opencsw.org> <5397F870.5060706@opencsw.org> <53A227AF.2090200@opencsw.org> Message-ID: <86D5DE3D-B36D-4BC2-A4F5-AC11ABE7A460@opencsw.org> Hi Riccardo, Am 19.06.2014 um 01:58 schrieb Riccardo Mottola : > Dagobert Michelsen wrote: >> I am wondering if it is correct that you use /usr/lib/libaspell.so.15 instead of >> /opt/csw/lib/libaspell.so.15. > I found an option ti disable aspell integration and thus, together with a couple of other fixes, I got my apps starting again, at least in my source tree, not yet in mgar. > > Could this be an LDFLAGS problem. SInce I get a crash, I think it is not a just a problem with the system aspell, but something like being compiled against one aspell and then run against the other. This may be possible. You can try by adding these flags to your manual compile: http://www.opencsw.org/manual/for-developers/index.html Best regards ? Dago -- "You don't become great by trying to be great, you become great by wanting to do something, and then doing it so hard that you become great in the process." - xkcd #896 -------------- next part -------------- A non-text attachment was scrubbed... Name: smime.p7s Type: application/pkcs7-signature Size: 2418 bytes Desc: not available URL: From rmottola at opencsw.org Wed Jun 25 23:16:53 2014 From: rmottola at opencsw.org (Riccardo Mottola) Date: Wed, 25 Jun 2014 23:16:53 +0200 Subject: first gnustep app! but segfaults In-Reply-To: <86D5DE3D-B36D-4BC2-A4F5-AC11ABE7A460@opencsw.org> References: <53864EE9.7050909@opencsw.org> <5397F870.5060706@opencsw.org> <53A227AF.2090200@opencsw.org> <86D5DE3D-B36D-4BC2-A4F5-AC11ABE7A460@opencsw.org> Message-ID: <53AB3C45.7010800@opencsw.org> Hi, Dagobert Michelsen wrote: > This may be possible. You can try by adding these flags to your manual > compile: > http://www.opencsw.org/manual/for-developers/index.html > I tried that. I configured the "gui" library this way: ./configure CPPFLAGS="-I/opt/csw/include" LDFLAGS="-L/opt/csw/lib -R/opt/csw/lib" PKG_CONFIG_PATH="/opt/csw/lib/pkgconfig" but I still get that crash and the wrong library used at runtime: 0xfdb41748 in __1cDstdM__node_alloc4BiA_L_M_allocate6FI_pv_ () from /usr/lib/libaspell.so.15 in fact ldd confirms that: bash-4.3$ ldd /opt/GNUstep/Local/Library/Libraries/libgnustep-gui.so libicui18n.so.52 => /opt/csw/lib/libicui18n.so.52 libicuuc.so.52 => /opt/csw/lib/libicuuc.so.52 libicudata.so.52 => /opt/csw/lib/libicudata.so.52 libpng12.so.0 => /opt/csw/lib/libpng12.so.0 libgnustep-base.so.1.24 => /opt/GNUstep/Local/Library/Libraries/libgnustep-base.so.1.24 libaspell.so.15 => /usr/lib/libaspell.so.15 libtiff.so.3 => /opt/csw/lib/libtiff.so.3 libz.so.1 => /opt/csw/lib/libz.so.1 libjpeg.so.62 => /opt/csw/lib/libjpeg.so.62 libm.so.2 => /lib/libm.so.2 libobjc.so.4 => /opt/csw/lib/libobjc.so.4 libgcc_s.so.1 => /opt/csw/lib/libgcc_s.so.1 libpthread.so.1 => /lib/libpthread.so.1 also, this is my manual compile, but I have the same problem in the packaged compile. Riccardo From maciej at opencsw.org Wed Jun 25 23:37:16 2014 From: maciej at opencsw.org (=?UTF-8?Q?Maciej_=28Matchek=29_Blizi=C5=84ski?=) Date: Wed, 25 Jun 2014 22:37:16 +0100 Subject: first gnustep app! but segfaults In-Reply-To: <53AB3C45.7010800@opencsw.org> References: <53864EE9.7050909@opencsw.org> <5397F870.5060706@opencsw.org> <53A227AF.2090200@opencsw.org> <86D5DE3D-B36D-4BC2-A4F5-AC11ABE7A460@opencsw.org> <53AB3C45.7010800@opencsw.org> Message-ID: On Wed, Jun 25, 2014 at 10:16 PM, Riccardo Mottola wrote: > ./configure CPPFLAGS="-I/opt/csw/include" LDFLAGS="-L/opt/csw/lib -R/opt/csw/lib" PKG_CONFIG_PATH="/opt/csw/lib/pkgconfig" Interesting, what does the configure script do with these foo=bar arguments? Is there a special shell lib to parse them? From rmottola at opencsw.org Wed Jun 25 23:47:27 2014 From: rmottola at opencsw.org (Riccardo Mottola) Date: Wed, 25 Jun 2014 23:47:27 +0200 Subject: copy file phase Message-ID: <53AB436F.8070105@opencsw.org> Hi, the gnustep packages are now quite ready! there are still some issues, like the "libspell" problem I had to disable right now, but the packaged apps do startup! that's great news. They need to be finalized somehow? Anyway, there is one important step I need to add in gnustep-make: copy a configuration file (installation-domains.conf). It needs to be put in the "etc" directory where GNUstep.conf is. With the current setup, that would be: /opt/csw/GNUstep/etc/GNUstep How can I accomplish that? how can that be packaged/done? Second, I suppose, that it would be more appropriate to put configure GNUstep.conf to be in /opt/csw/etc ? Riccardo From rmottola at opencsw.org Thu Jun 26 00:05:22 2014 From: rmottola at opencsw.org (Riccardo Mottola) Date: Thu, 26 Jun 2014 00:05:22 +0200 Subject: first gnustep app! but segfaults In-Reply-To: References: <53864EE9.7050909@opencsw.org> <5397F870.5060706@opencsw.org> <53A227AF.2090200@opencsw.org> <86D5DE3D-B36D-4BC2-A4F5-AC11ABE7A460@opencsw.org> <53AB3C45.7010800@opencsw.org> Message-ID: <53AB47A2.2070702@opencsw.org> Hi, Maciej (Matchek) Blizi?ski wrote: > On Wed, Jun 25, 2014 at 10:16 PM, Riccardo Mottola wrote: >> > ./configure CPPFLAGS="-I/opt/csw/include" LDFLAGS="-L/opt/csw/lib -R/opt/csw/lib" PKG_CONFIG_PATH="/opt/csw/lib/pkgconfig" > Interesting, what does the configure script do with these foo=bar > arguments? Is there a special shell lib to parse them? as far as I know, it is the standard way to pass those to configure one-shot, it always worked for me. This is from configure help: Usage: ./configure [OPTION]... [VAR=VALUE]... To assign environment variables (e.g., CC, CFLAGS...), specify them as VAR=VALUE. See below for descriptions of some of the useful variables. Riccardo From laurent at opencsw.org Thu Jun 26 09:52:30 2014 From: laurent at opencsw.org (Laurent Blume) Date: Thu, 26 Jun 2014 09:52:30 +0200 Subject: first gnustep app! but segfaults In-Reply-To: <53AB47A2.2070702@opencsw.org> References: <53864EE9.7050909@opencsw.org> <5397F870.5060706@opencsw.org> <53A227AF.2090200@opencsw.org> <86D5DE3D-B36D-4BC2-A4F5-AC11ABE7A460@opencsw.org> <53AB3C45.7010800@opencsw.org> <53AB47A2.2070702@opencsw.org> Message-ID: <53ABD13E.8020708@opencsw.org> Le 2014/06/26 09:45 +0200, Riccardo Mottola a ?crit: > as far as I know, it is the standard way to pass those to configure > one-shot, it always worked for me. > > This is from configure help: > > Usage: ./configure [OPTION]... [VAR=VALUE]... > > To assign environment variables (e.g., CC, CFLAGS...), specify them as > VAR=VALUE. See below for descriptions of some of the useful variables. Check what is the RPATH there: /usr/ccs/bin/dump -Lv /opt/GNUstep/Local/Library/Libraries/libgnustep-gui.so | grep PATH It surely looks like RUNPATH /usr/lib:/opt/csw/lib Save the output of the whole build: make >& output And confirm there if there are some places the arguments are in the wrong order (-R/usr/lib -R/opt/csw/lib), because configure has its own ideas on what to add. Most probably, it's now time to get out the LD_OPTIONS sledgehammer. Laurent From maciej at opencsw.org Thu Jun 26 14:27:55 2014 From: maciej at opencsw.org (=?UTF-8?Q?Maciej_=28Matchek=29_Blizi=C5=84ski?=) Date: Thu, 26 Jun 2014 13:27:55 +0100 Subject: first gnustep app! but segfaults In-Reply-To: <53ABD13E.8020708@opencsw.org> References: <53864EE9.7050909@opencsw.org> <5397F870.5060706@opencsw.org> <53A227AF.2090200@opencsw.org> <86D5DE3D-B36D-4BC2-A4F5-AC11ABE7A460@opencsw.org> <53AB3C45.7010800@opencsw.org> <53AB47A2.2070702@opencsw.org> <53ABD13E.8020708@opencsw.org> Message-ID: On Thu, Jun 26, 2014 at 8:52 AM, Laurent Blume wrote: > It surely looks like RUNPATH /usr/lib:/opt/csw/lib "/usr/lib:/opt/csw/lib" is wrong. It should be "/opt/csw/lib". -------------- next part -------------- An HTML attachment was scrubbed... URL: From dam at opencsw.org Thu Jun 26 21:39:03 2014 From: dam at opencsw.org (Dagobert Michelsen) Date: Thu, 26 Jun 2014 21:39:03 +0200 Subject: copy file phase In-Reply-To: <53AB436F.8070105@opencsw.org> References: <53AB436F.8070105@opencsw.org> Message-ID: Hi Riccarado, Am 25.06.2014 um 23:47 schrieb Riccardo Mottola: > the gnustep packages are now quite ready! there are still some issues, like the "libspell" problem I had to disable right now, but the packaged apps do startup! that's great news. > > They need to be finalized somehow? > > Anyway, there is one important step I need to add in gnustep-make: copy a configuration file (installation-domains.conf). It needs to be put in the "etc" directory where GNUstep.conf is. > > With the current setup, that would be: > /opt/csw/GNUstep/etc/GNUstep > > How can I accomplish that? how can that be packaged/done? This can be done in post-install: by copying stuff to $(PKGROOT). Then you define it as PRESERVECONF as documented here: http://wiki.opencsw.org/cswclassutils-package#toc18 > Second, I suppose, that it would be more appropriate to put configure GNUstep.conf to be in /opt/csw/etc ? Even better would be /etc/opt/csw, please see for details http://www.opencsw.org/manual/for-administrators/setup.html#etc-opt-csw-vs-opt-csw-etc Best regards -- Dago