Re: [Mingw-w64-public] OpenGL headers

[LRN <lr...@gm...>]

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

On 05.08.2013 22:04, LRN wrote:
> On 26.07.2013 19:59, Kai Tietz wrote:
>> Am 26.07.2013 01:22 schrieb "LRN":
>>> On 26.07.2013 09:33, Ozkan Sezer wrote:
>>>> On Fri, Jul 26, 2013 at 3:18 AM, LRN wrote:
>>>>> On 24.07.2013 00:22, Kai Tietz wrote:
>>>>>> thank you for the heads up.  Sure, it would me vety appreachiated if you
>>>>>> would take care for this update.
>>>>>> You should have commit rights.  So please sent mail with attached patch,
>>>>>> get approval for it, and then apply.
>>>>>> JonY, Jacek, Ozkan, or dw can give you assistance, if you require.
>>>>>
>>>>> OK, one question though:
>>>>> When previous generation of these headers was committed, what changes
>>>>> did the committer do to the original headers from the GL registry?
>>>>> Because newer version are quite different from the ones that are
>>>>> currently committed (this may be OK, since headers are generated from
>>>>> XML definitions, so they probably are not expected to produce small
>>>>> diffs when switching revisions, but still).
>>>>>
>>>>
>>>> They are some older version before opengl.org switched to using xml
>>>> registry instead of old .spec files. The last version of the old
>>>> headers are these:
>>>>
>>>> http://www.opengl.org/registry/oldspecs/glext.h
>>>> http://www.opengl.org/registry/oldspecs/glxext.h
>>>> http://www.opengl.org/registry/oldspecs/wglext.h
>>>> http://www.opengl.org/registry/oldspecs/glcorearb.h
>>>>
>>>> Now, I dont know which versions were the ones added to our repo,
>>>> because the ones from opengl.org as linked above do produce a diff:
>>>> see attached gl.diff to see.
>>>>
>>>> The latest official headers though, will indeed produce a large diff
>>>> because they are generated the new way.
>>>
>>> OK, so, i've looked briefly at the trunk vs oldspecheaders diff, it
>>> seems that no mingw-w64-specific changes were made, headers were
>>> committed verbatim.
>>>
>>> Anyway, for new 4.4 headers the diff is huge, and there's no adjustments
>>> are (presumably) necessary, which begs the question: what's the point of
>>> attaching a patch and getting approval for it?
>>>
>>> P.S. "presumably" means that i haven't tested these new headers with the
>>> few GL-using packages that i do build. But i will.
>>>
>> Well, the point is documenting the change in ML, too.  Not only in SVN
>> repository.
>> Additiinally it shows what patch will change and nobody gets surprised by
>> it.
> 
> Oh-kay. I've tested the patch, and the few GL things that i do compile
> (that would be gst-plugins-gl first and foremost, as well as glew) are
> fine with it.
> 
> So i'll attach the compressed patch to my next message. Do note that the
> patch is _bigger_ than new files themselves (103kb compressed vs ~79kb
> compressed), so the patch might have to go through moderation first.
> 
> 

And here's the patch.

- -- 
O< ascii ribbon - stop html email! - www.asciiribbon.org
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.11 (MingW32)

iQEcBAEBAgAGBQJR/+llAAoJEOs4Jb6SI2CwBd8H/AsyE/Pb//s/MZ+G07XFZfvm
DeeefjGMacsWs4VJidfzyJ6KkGEyrOt7bTjoKdA7zZDu65/XpAbaa2gk+7mRZGlw
7mZaWkGsp0CPSKmW/yFIw0mVtKsBjJP0z7SCpMNaWti3gPeOCsIy+mX38KI/tkJH
ve50yLFeaJQnzBOutA24wMC06kuWQt+NvY1fwSCkW+FbfNnZIV9exzFlJ0r7qZji
7NObQKxmPIiUg73dcRulP5juc6uVQa+GjI1J5U1s8azuf/BpJCsEKF2y4YqcHeeJ
+U70qJ+24dYRfgAYPYi5ogtw8BmX/t0DLHzKeXp5G61yzPhnM8KhMQrKKAsQYHg=
=o/dy
-----END PGP SIGNATURE-----