[MPlayer-dev-eng] The new Makefiles

[Alban Bedel]

On Tue, 27 May 2008 19:19:25 +0200
Diego Biurrun <diego at biurrun.de> wrote:

> On Tue, May 27, 2008 at 07:39:26PM +0200, Alban Bedel wrote:
> > On Tue, 27 May 2008 14:14:45 +0200
> > Diego Biurrun <diego at biurrun.de> wrote:
> > 
> > > TOOLS/netstream and TOOLS/vivodump still do not link under any
> > > circumstances.
> > > 
> > > Albeu, you said you knew a way to fix this.  Let's hear it.
> > 
> > I sent you a patch showing you what is needed to link them, then
> > warned you from the start that the file list _MUST_ be kept
> > slitted. You just ignored everything.
> 
> No, I just disagreed that
> 
> a) splitting the lists is necessary;

OK, now I'm burning to find out how in hell you will manage to not
split the list while not duplicating half of it.

> b) that if a) were true, it would be worth holding back my work on the
>    non-recursive Makefile to fix the linkage of a few obscure tools.

I never meant that! The only thing I insisted upon was to have:

LIBA_SRC = foo.c bar.c
LIBB_SRC = some.c

MPLAYER_SRC = $(LIBA_SRC) $(LIBB_SRC) another.c

Instead of one gigantic list. How that could in any way stop you from
converting the Makefile to non-recursive, I just don't know.

> > Now the work needed is way more important as one must split your
> > monster file list in logical parts :(
> 
> We will see if this is really so hard or if it cannot be done in
> another way.

Please remember that we don't want to have to maintain several file
list.
 
> > However you have all the information needed in the patch I sent you.
> > Apply the hunks on netstream.c and vivodump.c and all you need to
> > do is fix the makefile.
> 
> I'll look at your patch if I find the time.

If that mean you never even looked at it I'm really disappointed.
What's the point of asking my opinion and help if you anyway don't
really look at my proposals and keep doing things strictly your way?
That's just frustrating and a pure waste of time.

And if you just had listened a bit, and instead of a single list
simply converted each .a to a variable, which functionally is exactly
the same as what you did. The problem would already be solved, and you
would have spared quiet some configure changes.

	Albeu