[MPlayer-dev-eng] [PATCH] Remove redundant XShmGetEventBase declaration

[Attila Kinali]

On Tue, 12 Aug 2008 05:23:35 +0000
"Bryan Henderson" <giraffedata at gmail.com> wrote:

> On 8/11/08, Attila Kinali <attila at kinali.ch> wrote:
> > I just checked this. XF86 4.1.0 has a "broken" headerfile that
> > does not contain XShmGetEventBase, it has been fixed in December 2001
> > and released with XF86 4.2.0 in June 2002.
> >
> > I think, we can safely assume that today noone is using XF86 4.1.0 anymore.
> 
> If by "safely assume" you mean without being wrong, then no you can't.

With "safely" i mean, that in 99% of all cases it will be true.
Of course, there are still people who use stone age systems for
some reason (heck, i used to have a 2.0.38 kernel running until
well into the 2.6.x days). But we can assume that these people
know exactly what they are doing and know how to handle problems
that arise from using such an old system.

>  There are people who use old systems.  I just checked the X I used
> the most: XFree86 4.0.2, and it too has XShmGetEventBase but lacks the
> declaration of it.

Hmm.. I remember XF86 4.0.2 being a quite good release (actually the first
4.x that was stable enough to replace 3.3.6)... but why do you still use it?

> But of course you can still assume, as a way of limiting your
> maintenance effort.

That's the main point. If we try to keep hacks for every bug
in any library we use ever had, then we'll be just a big collection
of workarounds that will soon become unmaintainable.

> BTW does it really matter?  Missing this declaration probably just
> adds a drop to the ocean of compile warnings.  That would explain why
> I with my 4.0.2 haven't noticed a problem with the other
> declaration-less references to which Diego refers.

IMHO yes: The correct fix is to change the header file and not MPlayer.

			Attila Kinali

-- 
A strange game.
The only winning move is not to play.
			-- Joshua, WarGames