[FFmpeg-devel] DVCPRO HD: request for review

Michael Niedermayer michaelni
Tue Sep 2 22:55:35 CEST 2008

On Tue, Sep 02, 2008 at 07:46:29PM +0100, M?ns Rullg?rd wrote:
> Roman Shaposhnik <rvs at sun.com> writes:
> > On Sun, 2008-08-31 at 23:16 +0300, Uoti Urpala wrote:
> >> On Sun, 2008-08-31 at 21:33 +0200, Michael Niedermayer wrote:
> >> > On Sun, Aug 31, 2008 at 06:35:34PM +0100, M?ns Rullg?rd wrote:
> >> > > We can easily change the commit emails to use diff -w, if that's what
> >> > > people want.
> >> > 
> >> > That way someone could (unintentionally) randomize indention and move us
> >> > a step closer to mplayer.c style without anyone noticing
> >> 
> >> You could include another patch changing the indentation in the same
> >> mail, or versions of the patch with and without -w (in case they differ
> >> on >= N lines). I think the former would be strictly better than having
> >> commits creating incorrect indentation and then mostly pointless "fix
> >> indentation after previous commit" mails on the list and as separate
> >> commits in VCS history.
> >
> > It is nice to see that I'm not alone in considering this particular
> > policy to be suboptimal. Although, now that I'm using Git, it doesn't
> > bug me all that much that I have to split "code transactions", yet
> > it still leads to polluted history if you ask me.
> I agree, but I won't pick a fight over it.

There is noone with whom you could pick a fight with i think. If you can
automate spliting commits in functional parts + indention changes of otherwise
unchanged code i suspect few would complain.
Of course this must work in practice, which includes svn di (-x -w) and
developers not messing up like commiting unrelated cosmetics together with
functional changes. After all this is just about spliting reindention
when adding/removing a if/else/for ...

moving code around, renaming variables, ... all still would have to be split
by hand.
in that light its also not clear to me if this automatic spliting is worth it
but i surely do not object if (and only if) it does work in practice.

Michael     GnuPG fingerprint: 9FF2128B147EF6730BADF133611EC787040B0FAB

The greatest way to live with honor in this world is to be what we pretend
to be. -- Socrates
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 189 bytes
Desc: Digital signature
URL: <http://lists.mplayerhq.hu/pipermail/ffmpeg-devel/attachments/20080902/6f1a6aae/attachment.pgp>

More information about the ffmpeg-devel mailing list