[FFmpeg-cvslog] r9601 - trunk/libavcodec/mpegidct.c

[Diego Biurrun]

On Thu, Jul 12, 2007 at 03:38:11PM +0200, Michael Niedermayer wrote:
> 
> On Thu, Jul 12, 2007 at 03:05:17PM +0200, Diego Biurrun wrote:
> [...]
> > > > > > > This is clearly non-free.  What were you thinking ?!?
> > > > > >
> > > > > > My mistake. I thought it is public domain.
> > > > > 
> > > > > At least 2 other GPL project use same routine and have same header in it.
> > > > > http://cvs.xvid.org/cvs/viewvc.cgi/xvidcore/src/dct/idct.c?view=log
> > > > 
> > > > Which one is the second?
> > > 
> > > libmpeg2 has a file which seems derived from it, some comments are identical
> > > if you ignore whitespace
> > > it has walken as author and GPL header ...
> > > 
> > > mpeg4ip has it with MPL license header
> > > 
> > > iam sure you can find more if you try ...
> > 
> > > could someone explain me the reasons behind this overreaction?
> > 
> > I don't think this is an overreaction.  Committing non-free code that
> > renders FFmpeg non-distributable is about the worst abuse of commit
> > privileges I can imagine.
> 
> you dont react that strongly if someone commits a file without license
> header,

Maybe now you can understand why I dislike files without license headers
so much.

I do not react (so) strongly to files without license headers for two
reasons:

1) The policy to include license headers in every file is recent and
only recently did I fix up all files to include a license header.

2) There is a big difference between committing a file without license
header and committing one with a non-free license header.  If a dev like
you who has contributed under the LGPL.

> or that we have GPL code but no COPYING with the GPL

I intend to fix that we do not have the GPL.

> or the IJG code without their README, ...

I addressed this issue by pasting the README we were not shipping into
the relevant files some time ago.  If you think this is not sufficient
we can talk, but I surely did not ignore this.

> and then you go and change other peoples liceneses to clean them up
> remove terms you (who is no lawyer) think are not needed ...

I do not change other people's licenses unless it with explicit consent.
I never have and I never will.

There is a difference between changing a license and changing the terms
that specify the license.  I think Uoti explained it quite well in his
mail to the other thread.

> root really should threaten me with revoking my account the next time i
> commit a file with no proper license header :)

I don't like the fact that you insist on others following the policy
stricter than you follow it yourself, no.  This is not meant as a
personal attack.  I would just prefer you to set an example even for
things you do not consider terribly important and/or a nuisance.  Surely
this would help encouraging others doing the same for things that they
consider a nuisance but are important to you.

> i dont know what ivan said or did on IRC but from my POV this looks
> like you pick ivan out and attack him for something he unintentionally did
> and what he corrected quickly (timewise, as iam unaware of what was disscussed
> on IRC)
> while others who do things intentionally and refuse to correct their mess
> can get away without a scratch ...

Ivan told me repeatedly on IRC that I should clean up his mess since I
was the one complaining and refused to do it himself until Mans stepped
in.

His wmv2/whatever commit broke regression tests, he never bothered to
fix it.  In the end Benoit did it for him.

I do not see myself singling Ivan out for anything.  I rightly complained
about his commit, but I stayed intentionally silent about his commit
access.  I'm far too biased to make a statement in this matter.

I dislike the fact that he asked you for permission in private instead
of on ffmpeg-devel.  (Especially considering that he dislikes it when I
discuss things like commit access for new dev X in private - but let us
not delve further into this side matter.)  Peer review on ffmpeg-devel
would have caught this mistake, I'm sure at least I or Mans would have
caught it.  I noticed it within seconds on ffmpeg-cvslog.

> > > it doesnt seem other projects share your interpretation of the license
> > 
> > I would rather think that they don't give a damn.  In any case they are
> > wrong, wrong, wrong.
> 
> hmm, honestly i dont give a damn either, this file is used by half of
> all mpeg related software free sw as well as non free 

That's a pity.

Diego