[MPlayer-cvslog] r24941 - trunk/mplayer.c

[Uoti Urpala]

On Sat, 2007-11-03 at 16:37 -0400, Rich Felker wrote:
> A better commit message would be "statements before declarations not
> allowed in MPlayer". The policy is using a minimal subset of C99
> that's been supported by relevant compilers for over 10 (maybe 15?!)
> years, not explicitly supporting a particular compiler version.

You already made, and I refuted, a similar claim about gcc-2.95 features
matching some "natural stable subset" earlier, that time about asm
features. See
http://lists.mplayerhq.hu/pipermail/mplayer-dev-eng/2007-February/049660.html

Your claim this time is similarly false. How many 10-15 year old
compilers are there that can compile MPlayer? And how many of those that
can compile it do not fully support variable declarations?

> And actually the policy seems to be that people who care fix problems
> like this when they come up, since certain developers refuse to play
> by any rules...

The standard in open source is that those people who have skill and
actually do things decide how to do them. How many current developers
can work on the core functionality and overall architecture? How many do
changes that are even at the level of the audio changes this commit was
part of? Reimar seems to be the only one who does anything at all
similar. If someone tells me (or Reimar for that matter) how to do C
development based on rules that some other people decided I'm not going
to take them very seriously. Especially when the most often cited
"rules" seem to come from old development practices that were already
seen not to work and turned the code into a mess.