<p>Em 27/07/2011 22:27, "Dagobert Michelsen" <<a href="mailto:dam@opencsw.org">dam@opencsw.org</a>> escreveu:<br>
><br>
> Hi Maciej,<br>
><br>
> Am 27.07.2011 um 12:07 schrieb Maciej Bliziński:<br>
><br>
> > Em 26/07/2011 11:26, <<a href="mailto:dmichelsen@users.sourceforge.net">dmichelsen@users.sourceforge.net</a>> escreveu:<br>
> > ><br>
> > > Revision: 15223<br>
> > > <a href="http://gar.svn.sourceforge.net/gar/?rev=15223&view=rev">http://gar.svn.sourceforge.net/gar/?rev=15223&view=rev</a><br>
> > > Author: dmichelsen<br>
> > > Date: 2011-07-26 10:26:31 +0000 (Tue, 26 Jul 2011)<br>
> > ><br>
> > > Log Message:<br>
> > > -----------<br>
> > > apr: Remove ISALIST, relocate build-1 from share/ to libexec/<br>
> ><br>
> > What do you (Dago and others) think about a check that throws an error if there is a binary placed outside the typical locations? The typical locations would be bin, lib and libexec.<br>
><br>
> That would require safe detection of binaries. IIRC that was difficult,<br>
> the scripts here were shell stuff like a special libtool variant.</p>
<p>I consider it a solved problem, using libmagic and hachoir_parser. That's pretty reliable as far as I can tell.</p>
<p>> Additionally I filed a test for a defect string in RPATH from a double<br>
> expansion fo $ISALIST resulting in just SALIST which can never be good :-)</p>
<p>It is already covered by the general RPATH check, but it might make sense to throw an additional error on a double expansion.</p>
<p>Maciej<br>
</p>