[MPlayer-cvslog] r26111 - trunk/mp_msg.h

[Reimar Döffinger]

On Sat, Mar 01, 2008 at 11:35:08PM -0500, Rich Felker wrote:
> On Sat, Mar 01, 2008 at 11:38:23AM +0100, Diego Biurrun wrote:
> > On Thu, Feb 28, 2008 at 10:53:06PM -0500, Rich Felker wrote:
> > > On Thu, Feb 28, 2008 at 07:31:35PM +0100, diego wrote:
> > > > 
> > > > Log:
> > > > TARGET_OS2 is never set, use __OS2__ instead.
> > > > 
> > > > --- trunk/mp_msg.h	(original)
> > > > +++ trunk/mp_msg.h	Thu Feb 28 19:31:35 2008
> > > > @@ -111,7 +111,7 @@ int mp_msg_test(int mod, int lev);
> > > >  
> > > > -#ifdef TARGET_OS2
> > > > +#ifdef __OS2__
> > > 
> > > Again, I'm against this general class of change. I've explained the
> > > reasoning many times! Would you please honor that or at least discuss
> > > it rather than continuing these bad commits?
> > 
> > I've discussed it before, I get tired of repeating it: I have little
> > interest in discussing your theories.  There is no practical alternative
> > to what I committed.
> > 
> > If you disagree: Patches welcome.
> 
> I could write the code to detect OS2 and add #define TARGET_OS2 in
> configure, but I figure you're more qualified as the configure script
> maintainer and someone familiar with it... While you're fixing the
> problem described in this commit, you might as well fix it right..

Most of these are hacks anyway, they are not only specific to OS2 but
possibly as well to the compiler and libc.
I can not really see why Diego should waste his time writing a configure
check that likely isn't any more reliable than this.
I'm sure there are quite a few things that could need a real configure
check and would actually provide an advantage, like a way to disable asm
without unsetting ARCH completely and other things.

Greetings,
Reimar Döffinger