* Adam Cécile (Le_Vert) [Tue, 03 Jul 2007 21:57:26 +0200]:
Do you mean having the right audacious-plugins dependency on audacious
wouldn't be enough ?
No, it would be enough. Just to be clear, I'm talking here about the
dependency scheme proposed in my first message to this bug report, for
Hi Steve,
I really hate circular depency and I'm not sure it's the better way. In
fact, audacious is broken in testing for ages and forcing a build of mcs
on mipsel would fix all that crap...
Adeodato Simó a écrit :
* Russ Allbery [Mon, 02 Jul 2007 14:39:20 -0700]:
Steve Langasek
Le mardi 03 juillet 2007 à 00:40 +0200, Adeodato Simó a écrit :
windlord:~ audacious
Failed to load plugin (/usr/lib/audacious/Input/libaac.so):
/usr/lib/audacious/Input/libaac.so: undefined symbol:
vfs_buffered_file_new_from_uri
Files in audacious-plugins are dlopened by audacious,
* Adam Cécile (Le_Vert) [Tue, 03 Jul 2007 08:46:45 +0200]:
Hi Steve,
(My name is not Steve.)
I really hate circular depency and I'm not sure it's the better way.
Well, do as Joss said and tighten the audacious-plugins dependency in
audacious, *and* introduce tight dependencies against
* Josselin Mouette [Tue, 03 Jul 2007 09:34:19 +0200]:
And the sane way to fix it is to make the plugins link to the library.
This way audacious-plugins will depend on libaudacious5 and testing
won't be broken again.
Yah, and what happens when the application (linked against
libaudacious4)
Adeodato Simó a écrit :
* Adam Cécile (Le_Vert) [Tue, 03 Jul 2007 08:46:45 +0200]:
Hi Steve,
(My name is not Steve.)
Wasn't talking to new or mistake.
I really hate circular depency and I'm not sure it's the better way.
Well, do as Joss said and tighten the
* Steve Langasek [Tue, 03 Jul 2007 12:25:11 -0700]:
But at a minimum, yes, the audacious-plugins package should be depending on
libaudacious by way of shlibdeps.
There is no NEEDED entry in the plugins against libaudaciousX. Do you
mean hardcoding the dependency instead? (Which, true, solves
Do you mean having the right audacious-plugins dependency on audacious
wouldn't be enough ?
Adeodato Simó a écrit :
* Steve Langasek [Tue, 03 Jul 2007 12:25:11 -0700]:
But at a minimum, yes, the audacious-plugins package should be depending on
libaudacious by way of shlibdeps.
There is no
On Tue, Jul 03, 2007 at 09:31:13PM +0200, Adeodato Simó wrote:
* Steve Langasek [Tue, 03 Jul 2007 12:25:11 -0700]:
But at a minimum, yes, the audacious-plugins package should be depending on
libaudacious by way of shlibdeps.
There is no NEEDED entry in the plugins against libaudaciousX.
* Steve Langasek [Tue, 03 Jul 2007 13:02:50 -0700]:
On Tue, Jul 03, 2007 at 09:31:13PM +0200, Adeodato Simó wrote:
* Steve Langasek [Tue, 03 Jul 2007 12:25:11 -0700]:
But at a minimum, yes, the audacious-plugins package should be depending
on
libaudacious by way of shlibdeps.
* Russ Allbery [Mon, 02 Jul 2007 14:39:20 -0700]:
Steve Langasek [EMAIL PROTECTED] writes:
Um. Which version of audacious was this? libaudacious5 isn't in
testing at all, and the stable (=testing) version of audacious works
fine for me with libaudacious4 which it depends on.
(Also
11 matches
Mail list logo