Bug#760849: transition: glew

[Emilio Pozuelo Monfort <pochu@debian.org>]

On 02/07/15 09:33, Matteo F. Vescovi wrote:
> Hi Emilio!
> 
> On Thu, Jul 2, 2015 at 7:17 AM, Emilio Pozuelo Monfort <pochu@debian.org> wrote:
>> On 01/07/15 17:10, Matteo F. Vescovi wrote:
>>> Hi Jonathan!
>>>
>>> On Thu, 21 May 2015 11:59:44 +0100 Jonathan Wiltshire <jmw@debian.org> wrote:
>>>> Could you summarise which packages now ftbfs, and the ones you think may
>>>> be GLEW-related?
>>>
>>> So, I tested the 11 packages that were failing on 1.11.0 and built them back
>>> against 1.12.0 and the results are:
>>>
>>> * FTBFS but not GLEW-related:
>>>  - avogadro
>>>  - ball
>>>  - cegui-mk2
>>>  - megaglest
>>>  - spring
>>
>> What are the bugs for this?
> 
> I have not filed bug reports for these; just tested their failure:
> 
>  - avogadro: http://debomatic-i386.debian.net/distribution#experimental/avogadro/1.0.3-10.1/buildlog
>  - ball: http://debomatic-i386.debian.net/distribution#experimental/ball/1.4.2+20140406-1/buildlog
>  - cegui-mk2: http://debomatic-i386.debian.net/distribution#experimental/cegui-mk2/0.7.6-3.3/buildlog
>  - megaglest: http://debomatic-i386.debian.net/distribution#experimental/megaglest/3.11.1-1/buildlog
>  - spring: http://debomatic-i386.debian.net/distribution#experimental/spring/98.0+dfsg-1/buildlog

Please file bugs for these issues if they haven't been reported yet.

>>> * FTBFS and possibly GLEW-related:
>>>  - mupen64plus-video-z64
>>>  - openscad
>>>  - warzone2100
>>
>> Possibly?
> 
> Yes, because from their build logs I can't figure out if the failure
> is due to a direct GLEW issue or not.
> For mupen64plus-video-z64 I got (snip from [1]):
> 
> - - - >8 - - -
> g++ -g -O2 -fstack-protector-strong -Wformat -Werror=format-security
> -fvisibility-inlines-hidden -g -O2 -fstack-protector-strong -Wformat
> -Werror=format-security -flto -DNDEBUG -Wall -ffast-math
> -fno-strict-aliasing -fvisibility=hidden -I../../src -fPIC
> -I/usr/include/libdrm  -I/usr/include/libdrm  -I/usr/include/SDL2
> -D_REENTRANT "-I/usr/include/mupen64plus/" -g -MD -MP
> -D_FORTIFY_SOURCE=2 -D_FILE_OFFSET_BITS=64  -c -o _obj/rgl.o
> ../../src/rgl.cpp
> ../../src/rgl.cpp:80:18: fatal error: glew.h: No such file or directory
>  #include <glew.h>
>                   ^
> compilation terminated.
> - - - >8 - - -
> 
> while it's clear that "glew.h" is there, since the -dev package has
> been installed in the chroot.

Might be that it's no longer detecting glew support and missing the right
cflags. You can check if this is related to glew 1.12 by trying a build against
1.10 and seeing if that fails.

> For openscad, it is FTBFS already and it has been reported at #790776.
> 
> For warzone2100, the issue has been fixed upstream, as Paul already reported.
> I'll file a bug report later today, pointing to the git commit fixing it.
> 
>> Please file bugs for these FTBFS. Ideally with patches (or link to patches from
>> upstream / other distros).
> 
> OK, I'll try to do my best to reduce the efforts to fix them.
> 
>>> * builds fine:
>>>  - linphone
>>>  - opencsg
>>>  - psychtoolbox-3
>>
>> I see lots of other packages in the tracker[1]. What about those?
> 
> All the other packages were already tested to build fine against 1.12.0 version.

OK good.

One more thing: make the FTBFS bugs (related and unrelated ones) block this one.

Cheers,
Emilio