[MPlayer-dev-eng] uau - svn account removal

[Uoti Urpala]

On Sat, 2007-02-24 at 20:51 +0100, Dominik 'Rathann' Mierzejewski wrote:
> When you decide to contribute code to a project, you are expected to follow
> that project's rules. That's basic politeness, not to mention other
> reasons. You may question the rules, but overtly ignoring them will only
> make people angry.

If you're contributing minor amounts of code to a project then you
should usually follow the requirements of the main developers (though
practice varies). MPlayer has no formal process to create "rules of the
project" that would have more authority than the degree to which the
opinions of the individual developers agree to them. Even if people
disagree with something they shouldn't talk of their opinion as "the
rules".

> > Of course some level of
> > cooperation is required, but the only people calling for account
> > suspension are ones who do not actually work on the same code.
> 
> Repeatedly violating written rules is not even close to "some level of
> cooperation". It's more of "I won't play by your rules and you must accept
> what I do anyway". Well, we don't.

Repeatedly violating which written "rules"? Some part of svn-howto.txt?
Which part exactly have I done differently "repeatedly"? And who's "we"
above? Are you speaking on behalf of every other developer?

Seems your idea of "cooperation" is telling people how to do their work
while you're not doing anything related to that work.

> > I do not claim that the commit was an ideal example or that you would
> > have to like it. But it does not justify your reaction.
> 
> Yes, it does. You were repeatedly warned not to break the rules. You were
> asked to fix your commit. You ignored us. Should I assume you don't intend
> to revert and recommit your changes to mplayer.c properly?

No, I don't intend to recommit changes to mplayer.c. I haven't seen any
concrete suggestions for what a "proper" way of doing them would be,
much less justification for doing so.

The only concrete suggestions related to the commit that I've seen at
all were about creating command.c with svn cp. I don't think that would
be particularly beneficial in practice, but it still remains as possible
as immediately after the commit (nothing has been "committed over the
changes" in that file despite what you might think from reading the
claims in this thread).

I get the impression that at least some of the demands for reverting are
made with the attitude "I don't agree with that, it must be reverted and
reapplied for show even if there is no practical benefit".

If some developers have problems figuring out exactly what the commit
does I could write a more detailed commit log entry for it. Reading such
structural changes from (necessarily large) diffs is not easy no matter
how such diffs are created.