[PATCH 3/4] Generate swrast GLX extension list, rather than using a fixed list

[Jeremy Huddleston Sequoia]

On Jul 25, 2012, at 9:14 AM, Ian Romanick <idr at freedesktop.org> wrote:

> On 07/25/2012 08:36 AM, Jeremy Huddleston Sequoia wrote:
>> 
>> On Jul 25, 2012, at 6:45 AM, Jon TURNEY <jon.turney at dronecode.org.uk> wrote:
>>>>> +#ifndef __APPLE__
>>>>> +    __glXEnableExtension(screen->glx_enable_bits, "GLX_SGIS_multisample");
>>>>> +#endif
>>> 
>>> As far as I can tell, this conditional is entirely pointless (now, I'm
>>> guessing that there is some historical reason for it), as this GL provider is
>>> never used by Xquartz.  Perhaps you can confirm that?
>> 
>> Correct, it's not used by XQuartz, but it is used by other servers (like Xorg) which can run on darwin.
>> 
>>  __APPLE__ just means that it's being built for OS X, not XQuartz.
>> 
>> It originally entered xserver with e56e24af252bd3b8e58076adf0f8eabf1103f187, and I changed it from __DARWIN__ to __APPLE__ in 54654815fa5e59b25cfd1fa72610120b72c10175.  But I don't know *why* it is there, and it doesn't really make sense to me.  The e56e24af commit was http://lists.x.org/archives/xorg-commit/2004-June/001142.html:
>> 
>> So that doesn't give much clue.
>> 
>> Giving it a punt is the right thing to do.  If it has fallout, I'm sure the 1 person who cares about it will say something.
> 
> I can't see how it would make any difference.  Like I said in a previous post, for GLX, the SGIS and ARB extensions are identical.

Yeah, but someone at some point had a reason for doing this, and I just wish I knew what it was.  I just like to unwrap puzzling mysteries ;)

--Jeremy