[FFmpeg-devel] [RFC] New library for shared non-generic libav* utils

[Michael Niedermayer]

On Fri, Jul 09, 2010 at 09:52:41PM +0100, M?ns Rullg?rd wrote:
> Michael Niedermayer <michaelni at gmx.at> writes:
> 
> > On Fri, Jul 09, 2010 at 07:57:16PM +0100, M?ns Rullg?rd wrote:
> >> Baptiste Coudurier <baptiste.coudurier at gmail.com> writes:
> >> 
> >> > On 07/09/2010 11:26 AM, M?ns Rullg?rd wrote:
> >> >> Baptiste Coudurier<baptiste.coudurier at gmail.com>  writes:
> >> >>
> >> >>> On 07/09/2010 11:02 AM, M?ns Rullg?rd wrote:
> >> >>>> Baptiste Coudurier<baptiste.coudurier at gmail.com>   writes:
> >> >>>>
> >> >>>>> On 07/09/2010 09:48 AM, M?ns Rullg?rd wrote:
> >> >>>>>> Michael Niedermayer<michaelni at gmx.at>    writes:
> >> >>>>>>
> >> >>>>>>> On Fri, Jul 09, 2010 at 04:41:59PM +0100, M?ns Rullg?rd wrote:
> >> >>>>>>>> Michael Niedermayer<michaelni at gmx.at>    writes:
> >> >>>>>>>>
> >> >>>>>>>>> I spended alot of time on libavutil and its only goal was to become
> >> >>>>>>>>> a general utils lib
> >> >>>>>>>>
> >> >>>>>>>> Said who?  It wasn't even your idea to begin with.  It was suggested
> >> >>>>>>>> and implemented by Alexander Strasser.
> >> >>>>>>>
> >> >>>>>>> svn blame of *.c *.h says:
> >> >>
> >> >> [...]
> >> >>
> >> >>>>>>> so id say, yes iam still the primary maintainer and author, even if
> >> >>>>>>> we consider that blame is not the worlds most idiot proof way to
> >> >>>>>>> check this
> >> >>>>>>
> >> >>>>>> Yes, you wrote more lines than anyone else, but not by any large
> >> >>>>>> margin.  Of the total ~11k lines, you only contributed roughly 25%.
> >> >>>>>> If lines were votes, you'd be losing.  You seem to like votes...
> >> >>>>>
> >> >>>>> Nah, this is heavily biased. A lot of lines are defines and macros in
> >> >>>>> *.h, not talking about the recent controversial documentation commits.
> >> >>>>
> >> >>>> Much of libavutil functionality resides in header files, so counting
> >> >>>> them is anything but biased.Are you jealous because your name didn't
> >> >>>> show up at all?
> >> >>>
> >> >>> Are you on crack ? You'd better stop your childish game and get your
> >> >>> shit together Mans. You are going too far.
> >> >>
> >> >> You are the one who needs to calm down.
> >> >
> >> > I'm pretty calm at the moment. On the contrary, during the past few
> >> > days I feel that you have been very offensive, and any thread seems an
> >> > opportunity to continue and provoke people, myself included. I think
> >> > your last sentence illustrates this well.
> >> >
> >> >> Did I do something to offend you?
> >> >
> >> > Not only me, but other people as well.
> >> 
> >> Where exactly did I offend YOU?  I could understand, though not agree,
> >> with Michael being offended by something I said in the last few days,
> >> although offence has certainly not been my intent.
> >> 
> >> >> A while ago when you and Michael had a big fight, I mostly sided
> >> >> with you.  Is this the thanks I get?
> >> >
> >> > That was a while ago, and as you can see, things have changed,
> >> > hopefully I'd say. Michael is the one trying to be reasonable
> >> > currently, and I second this.
> >> 
> >> Please do tell me what made you turn all your hate on me.  In the last
> >> year or so, you have not missed one chance to jump into a discussion
> >> for no other apparent reason than to contradict me just for the sake
> >> of it.  Why this hostility?
> >
> > as you are asking this. i must admit that i have a similar feeling and
> > question in relation to you
> >
> > the intreadwrite stuff (why does it bother you so much if we export
> >  it through a clean and documented api, projects use it anyway and
> >  as is its certainly filled with more issues than if we tried to make
> >  it at least work in the common cases and documented it we can even
> >  officially say its not recommanded to be used ...
> 
> Fine, I'll see what I can do there.  I guess something is better than
> nothing after all.

/me gives mans a virtual box of beer/cola/whatever you like

:)


> 
> > the gnu linker issues (you jumped at gnus defence but reading irc logs
> >  a while ago gave me the feeling you did not understand the issues there
> >  fully, so why did you attack me?
> 
> I don't remember exactly what I said, but I never meant to attack you
> in person.  If I said something in anger, I apologise for that.
> 
> I may not fully understand all the intricacies of the linker, but I do
> know this: it works the way the authors intend and document it to
> work. 

i doubt this is documented and iam unsure if it is intended like this
wasnt it drepper i would be certain that it cannot be intended


> It would be foolish of us to insist on changing
> well-established linker behaviour due to a weird corner-case where it
> would be more convenient for us if it did differently.  Even if we
> could get the linker changed, it would take many years for the change
> to trickle down to the places where it matters: distributions.  People
> building their own ffmpeg are not affected by the problem at all.

i understand it wouldnt help us. Still i think its something that should
be changed in the linker


> 
> > my svn repo on mphq, i remember you where against me being able to
> >  put my own little foss projects on mphq.
> 
> It's not that I personally have anything against _you_ keeping your
> stuff there.  The hosting and bandwidth is sponsored by people doing a
> favour to ffmpeg and mplayer, not to random projects they don't know
> or care about.  Were every ffmpeg/mplayer developer to host his
> private projects there, the sponsors might feel their generous (yes
> really) donations were being abused.  We do not want that to happen.

why did you not say this back then?


[...]
> > the removial of the gnu linker bug explanation/rant
> 
> Rants have no place in API documentation, especially not if they take
> the place of an actual description of what something does.  You cannot
> deny that the describes the macro more accurately after my change.

are you ok if i try to describe the problem in the gnu linker in that
doxy? Ill try to stay neutral in tone and i dont mind the native english
people to improve the wording ...

[...]
-- 
Michael     GnuPG fingerprint: 9FF2128B147EF6730BADF133611EC787040B0FAB

I do not agree with what you have to say, but I'll defend to the death your
right to say it. -- Voltaire
-------------- 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/20100710/8685469e/attachment.pgp>