Michael Schwendt pisze:
On Mon, 15 Dec 2008 21:57:47 +0100, Julian wrote:
> Dominik 'Rathann' Mierzejewski pisze:
>> On Monday, 15 December 2008 at 21:21, Julian Sikorski wrote:
>>> Michael Schwendt pisze:
>> [...]
>>>> ======================================================================
>>>> Broken packages in rpmfusion-free-development-x86_64:
>>>>
>>>> ffmpeg-libs-0.4.9-0.52.20080908.fc10.i386 requires libfaac.so.0
>>>> gstreamer-plugins-bad-0.10.9-1.fc10.i386 requires libfaac.so.0
>>>> libquicktime-1.0.3-4.fc10.i386 requires libfaac.so.0
>>>> mencoder-1.0-0.103.20080903svn.fc10.i386 requires libfaac.so.0
>> [...]
>>> I am somewhat responsible for faac, I suggested to update it but did not
>>> notice the abi bump. Sorry. Please let me know if I can help to bring
>>> this back to shape.
>> Well, you could check if there haven't been any API changes, IOW if the
>> affected packages build against the new faac. I would prefer it if this
>> build were removed because we're in the middle of rebuilding most of our
>> multimedia stack against the new x264 and ffmpeg and I don't want any
>> unnecessary release bumps and rebuilds.
>>
>> Regards,
>> R.
>>
> Hmm, I just installed the faac-1.26 on my Fedora 10, and it seems the
> soname is the same:
> $ rpm -q --provides faac
> libfaac.so.0()(64bit)
> faac = 1.26-1.fc11
> faac(x86-64) = 1.26-1.fc11
> $ readelf -a /usr/lib64/libfaac.so.0.0.0 |grep SONAME
> 0x000000000000000e (SONAME) Library soname: [libfaac.so.0]
And what are you trying to prove?
* Notice the repository id.
* Notice the package %arch.
* Draw your conclusion.
* Confirm by listing the repo directory.
Got it, faac did not get picked to be multilib :)
Julian