[FFmpeg-cvslog] r13444 - in trunk: libavcodec/acelp_vectors.h libavcodec/cook.c libavcodec/dnxhdenc.c libavcodec/h264.c libavcodec/qdm2.c libavcodec/qtrle.c libavcodec/vc1.c libavcodec/vp3.c libavcodec/vqavideo.c libavfilter/avfilter.h libavformat/avidec.c libavformat/ipmovie.c libavformat/nsvdec.c libavformat/rl2.c libavformat/rmenc.c

[Diego Biurrun]

On Wed, May 28, 2008 at 03:25:23PM +0100, M?ns Rullg?rd wrote:
> Diego Biurrun <diego at biurrun.de> writes:
> 
> > On Tue, May 27, 2008 at 05:30:02PM +0200, Michael Niedermayer wrote:
> >> On Tue, May 27, 2008 at 03:06:35PM +0200, Diego Biurrun wrote:
> >> > On Tue, May 27, 2008 at 03:00:49PM +0200, Benoit Fouet wrote:
> >> > > Diego Biurrun wrote:
> >> > > > On Tue, May 27, 2008 at 01:39:48PM +0100, M?ns Rullg?rd wrote:
> >> > > >   
> >> > > >> Diego Biurrun <diego at biurrun.de> writes:
> >> > > >>     
> >> > > >>> but within FFmpeg indexes is much more widespread.
> >> > > >>>       
> >> > > >> Does it matter if both are used?  Pointless changes like this make svn
> >> > > >> blame less useful.
> >> > > >
> >> > > > Depends what you care about.  I do somewhat care about spelling
> >> > > > consistency, but of course this is not particularly high priority.  But
> >> > > > I also do not mind stepping through a few iterations in svn blame...
> >> > > 
> >> > > and do you care about people feeling the other way ?
> >> > 
> >> > That's kind of a loaded question.  
> >> 
> >> > Prior to other people voicing their
> >> > opinions, I cannot read their minds.
> >> 
> >> Iam glad to hear that you can do afterwards ;)
> >> 
> >> and
> >> Prior to the commit other people could not read your mind, thus "could not"
> >> voice their oppinion.
> >
> > It just did not occur to me prior to committing. Anyway, the issue is
> > very minor IMO, yet I sense a big bikeshed nearby...
> 
> I've lost count of the number of times a simple blame lookup has taken
> significant time because of useless fixes like this being committed,
> most of them by you.  Some were things like whitespace cleanup, which
> I agree should be done.  Pure renaming missions, like this one, I
> think are doing more harm than good, however.  If you feel strongly
> about consistency, make sure new patches are good, but please leave
> existing code alone.

I do blame lookups all the time and have resigned myself to the fact
that they can take significant time.  I'll have to note here that I find
viewvc significantly more comfortable than the command line client for
these tasks.

When I stumble over some particular sort of typo, I often just grep the
whole tree and fix them quickly.  I fix other mistakes I find on those
lines while I'm at it, but I don't go over complete files.  I know the
latter would be more desirable as it creates less commit clutter, but it
can take huge amounts of time.

Anyway, I can refrain from cleanups like this one in the future if
preferred.  I never claimed such pedantry to be important in any way,
but I did not assume that anybody minded either.

Diego