[18039] in Athena Bugs
Re: gtk-1.2.6 installed on Linux Athena.
daemon@ATHENA.MIT.EDU (Richard Tibbetts)
Wed Jul 26 15:42:17 2000
Message-Id: <200007261942.PAA03433@hikari-no-ken.mit.edu>
To: tb@MIT.EDU (Thomas Bushnell, BSG)
cc: Richard Tibbetts <tibbetts@MIT.EDU>, Greg Hudson <ghudson@MIT.EDU>,
bugs@MIT.EDU, tibbetts@MIT.EDU
In-reply-to: Your message of "26 Jul 2000 15:26:28 EDT."
<u1hu2dc3d7f.fsf@alice-whacker.mit.edu>
Date: Wed, 26 Jul 2000 15:42:10 -0400
From: Richard Tibbetts <tibbetts@MIT.EDU>
On 7/26 tb@MIT.EDU (Thomas Bushnell, BSG) wrote:
> Richard Tibbetts <tibbetts@MIT.EDU> writes:
>
> > Nearly every app that uses both gtk and autoconf will have issues, I
> > believe. This is because of the two stage process by which autoconf
> > checks for gtk libraries. First it runs gtk-config
> > (/usr/athena/bin/gtk-config, version 1.2.8 since that is the first one
> > in most people's path). Then configure builds a test program which
> > links with libgtk (which gets /usr/lib/gtk, version 1.2.6). Since the
> > version numbers are different, it blows out.
>
> It should use gtk-config to compile the test program! Why on earth
> doesn't it?
I spoke to Thomas, and it seems that the bug might be better described
as follows:
/usr/athena/bin/gtk-config returns information about libraries that
eat in /usr/athena/lib, but does not actually cause programs to use
these libraries.
A solution to this would have /usr/athena/bin/gtk-config --libs
including R-paths as well as -L and -l options.
Does this make sense? I could do a patch if it was desired.
tibbetts
PS: This might make contained builds more interesting.
-*- http://www.mit.edu/~tibbetts -*- finger tibbetts@monk.mit.edu -*-
