[MPlayer-users] Compiling SVN; was: Jerky 1080p h264 playback

[Uoti Urpala]

On Tue, 2010-05-18 at 20:12 -0700, RC wrote:
> On Mon, 17 May 2010 17:57:26 +0300
> Uoti Urpala <uoti.urpala at pp1.inet.fi> wrote:
> 
> > And what determines that trying to move improvements to svn would not
> > help users is this: it would take much less effort to fix any problems
> > in git than to backport features to svn.
> 
> "Less effort" for you, doesn't help users in any way...

Not only "for me". It's just not possible to create a competitive
version based on svn without wasting a lot of effort.

Where did you invent that "doesn't help users" from? You think the users
are unhappy when the code "only" solves their problems but doesn't come
from your preferred svn? I doubt many of the users themselves agree.


> > svn is not being "actively developed by several people". Reimar is the
> > only developer who prefers the svn repo and can be said to be actively
> > working on code. Carl has done some smaller stuff.
> 
> Dozens of different people have made comits to the SVN repo this year. 
> Are you trying to say they ALL woud prefer to switch over and develop
> your fork?  If so, why don't they do so right now, and instead continue
> to work against SVN?  Or are you trying to say they aren't developers?

The following accounts have made at least 10 commits this year to the
svn repo (commits to translations and to libswscale excluded -
libswscale is considered to be part of FFmpeg; technically inside the
MPlayer svn repo but for historical reasons only):

cehoyos
compn
diego
komh
reimar
sesse
zuxy

So on that list are the 2 people I mentioned and 5 others. That is
already pretty far from your "dozens of people" claim (your claim is
strictly false even if you use the most liberal interpretation and count
people with a single commit). Of those 5 people, compn's commits are
fourcc additions and typo fixes. diego has a high number of commits but
those are mostly simple cleanups. komh does OS/2 porting. sesse has done
some nontrivial code development, but with the exception of a qt change
he only works on the binary loader module (which also has little to do
with git/svn differences). zuxy has mostly simple commits with a couple
of less trivial ones.

sesse's commits form the most significant development work out of those,
but they're also the least relevant to git/svn repo differences.


> Yes, I'm tried of this, but if you're just going to keep repeating the
> same lies...

You've been the one to make false statements (and demonstrably so). I
doubt whether you could objectively describe the development situation
even if you tried to be honest; you haven't shown much understanding of
it so far.