Bug#861957: closed by Joachim Breitner <nome...@debian.org> (Re: Bug#861976: libghc-glut-dev: attempts to load libglut.so instead of libglut.so.3)
Hi, On 07/05/17 19:38, Adrian Bunk wrote: > Control: reassign -1 libghc-glut-dev 2.7.0.10-3 > Control: fixed -1 2.7.0.10-4 > Control: affects -1 src:raincat > > On Sun, May 07, 2017 at 01:51:04PM +, Debian Bug Tracking System wrote: >> ... >> With haskell-glut 2.7.0.10-4 this should be fixed in unstable. > > I am reassigning it so that version tracking covers this. > > I'd guess raincat might additionally need a binNMU? It does - filed as #862061. Thanks, James signature.asc Description: OpenPGP digital signature
Bug#861957: closed by Joachim Breitner <nome...@debian.org> (Re: Bug#861976: libghc-glut-dev: attempts to load libglut.so instead of libglut.so.3)
Control: reassign -1 libghc-glut-dev 2.7.0.10-3 Control: fixed -1 2.7.0.10-4 Control: affects -1 src:raincat On Sun, May 07, 2017 at 01:51:04PM +, Debian Bug Tracking System wrote: >... > With haskell-glut 2.7.0.10-4 this should be fixed in unstable. I am reassigning it so that version tracking covers this. I'd guess raincat might additionally need a binNMU? > https://packages.qa.debian.org/h/haskell-glut.html > does not tell me yet whether it will migrate to testing automatically > or not. Due to the freeze, there is no automatic testing migration since February 5th. Please file an unblock request. > Greetings, > Joachim cu Adrian -- "Is there not promise of rain?" Ling Tan asked suddenly out of the darkness. There had been need of rain for many days. "Only a promise," Lao Er said. Pearl S. Buck - Dragon Seed
Bug#861976: libghc-glut-dev: attempts to load libglut.so instead of libglut.so.3
Hi,
Am Samstag, den 06.05.2017, 21:09 +0100 schrieb James Cowgill:
> In haskell-glut 2.7.0.10 the behavior of loading libglut changed
> significantly. Before, it was linked in directly (using
> "extra-libraries") and dlopen(NULL) was used to access functions in
> it.
> In 2.7.0.10 the direct linking was removed and the loading is now
> done
> using dlopen("libglut.so").
>
> This is bad because:
> - libglut.so is the wrong library (should be libglut.so.3)
> - Debian packages have lost the automatic dependency on freeglut3
>
> This has caused raincat to stop working after it was binNMUed (see
> #861957). While it could be worked around by adding a "Depends:
> freeglut3-dev", this is a bit of a hack so it would be better of
> haskell-glut was fixed instead.
upstream has since then changed to code to try both libglut.so and
libglut.so.3:
https://github.com/haskell-opengl/GLUT/commit/a386914280dbc7e027e8ebbfe07774cde697e38d
Normal linking would indeed be preferable. This is the commit that removed it:
https://github.com/haskell-opengl/GLUT/commit/1b47ba6c8074dc163886baf5f3d5b1809501f19c
I inquired about the reason here:
https://github.com/haskell-opengl/GLUT/commit/1b47ba6c8074dc163886baf5f3d5b1809501f19c#commitcomment-22044194
Is this bug relevant for the release?
Greetings,
Joachim
--
Joachim “nomeata” Breitner
Debian Developer
nome...@debian.org • https://people.debian.org/~nomeata
XMPP: nome...@joachim-breitner.de • GPG-Key: 0xF0FBF51F
https://www.joachim-breitner.de/
signature.asc
Description: This is a digitally signed message part
Bug#861976: libghc-glut-dev: attempts to load libglut.so instead of libglut.so.3
Hi,
On 06/05/17 21:33, Joachim Breitner wrote:
> Am Samstag, den 06.05.2017, 21:09 +0100 schrieb James Cowgill:
>> In haskell-glut 2.7.0.10 the behavior of loading libglut changed
>> significantly. Before, it was linked in directly (using
>> "extra-libraries") and dlopen(NULL) was used to access functions in
>> it.
>> In 2.7.0.10 the direct linking was removed and the loading is now
>> done
>> using dlopen("libglut.so").
>>
>> This is bad because:
>> - libglut.so is the wrong library (should be libglut.so.3)
>> - Debian packages have lost the automatic dependency on freeglut3
>>
>> This has caused raincat to stop working after it was binNMUed (see
>> #861957). While it could be worked around by adding a "Depends:
>> freeglut3-dev", this is a bit of a hack so it would be better of
>> haskell-glut was fixed instead.
>
> upstream has since then changed to code to try both libglut.so and
> libglut.so.3:
> https://github.com/haskell-opengl/GLUT/commit/a386914280dbc7e027e8ebbfe07774cde697e38d
>
> Normal linking would indeed be preferable. This is the commit that removed it:
> https://github.com/haskell-opengl/GLUT/commit/1b47ba6c8074dc163886baf5f3d5b1809501f19c
> I inquired about the reason here:
> https://github.com/haskell-opengl/GLUT/commit/1b47ba6c8074dc163886baf5f3d5b1809501f19c#commitcomment-22044194
>
> Is this bug relevant for the release?
It broke raincat (#861957). Either this bug needs to be fixed or some
workaround needs to be applied to raincat for stretch.
Thanks,
James
signature.asc
Description: OpenPGP digital signature
Bug#861976: libghc-glut-dev: attempts to load libglut.so instead of libglut.so.3
Package: libghc-glut-dev
Version: 2.7.0.10-1
Severity: important
Control: block 861957 by -1
X-Debbugs-CC: pkg-games-de...@lists.alioth.debian.org
Hi,
In haskell-glut 2.7.0.10 the behavior of loading libglut changed
significantly. Before, it was linked in directly (using
"extra-libraries") and dlopen(NULL) was used to access functions in it.
In 2.7.0.10 the direct linking was removed and the loading is now done
using dlopen("libglut.so").
This is bad because:
- libglut.so is the wrong library (should be libglut.so.3)
- Debian packages have lost the automatic dependency on freeglut3
This has caused raincat to stop working after it was binNMUed (see
#861957). While it could be worked around by adding a "Depends:
freeglut3-dev", this is a bit of a hack so it would be better of
haskell-glut was fixed instead.
Thanks,
James
signature.asc
Description: OpenPGP digital signature
