[FFmpeg-cvslog] random thoughts about SoC (was: Re: random thoughts about refactoring)

[Michael Niedermayer]

On Mon, Jan 11, 2010 at 12:08:22AM +0100, Diego Biurrun wrote:
> On Sun, Jan 10, 2010 at 11:13:31PM +0100, Michael Niedermayer wrote:
> > On Sun, Jan 10, 2010 at 10:45:16PM +0100, Diego Biurrun wrote:
> > > On Sun, Jan 10, 2010 at 10:23:34PM +0100, Michael Niedermayer wrote:
> > > > On Sun, Jan 10, 2010 at 02:48:00PM -0500, compn wrote:
> > > > > 
> > > > > i'll paste some relevent irc chat:
> > > > 
> > > > thank you!
> > > > while not as usefull as i thought, it is usefull.
> > > > 
> > > > some comments below
> > > > 
> > > > > [17:49] <DonDiego> Dark_Shikari: we have no students who reach the patch review phase, so this is not a problem
> > > > 
> > > > ehm? there are plenty who did reach patch review, theres even a patch laying
> > > > around that i should review. You make it look quite a bit worse than it is
> > > 
> > > SoC is not going too well and what's worse, it's not improving year over
> > > year..
> > 
> > Thats not ffmpeg specific.
> 
> I was not talking about any other projects, I know no statistics for
> others.

Knowing about other projects in similar situations and how the try to
solve it and how and if that was successfull seems better to me than to
simply assume you are right and the soc mentors are all wrong


>  For FFmpeg it's less than stellar and not improving.  Clearly
> something we must work on.

We can also work on squaring the circle.
This probably is exagerating but at least i do not think the difficulty of
tasks is the part that needs significant adjustment.

Areas that do need adjustment is getting more highly qualified students to
apply to ffmpeg-soc in the first place and
Improving communication, as jason suggested IRC, this is an option.
It could help us better know the students and their abilities early
to better select qualified students and we could better help during
the actual soc task if we are aware of what is done and where problems
are.


> 
> > Also if u ask for little you get little, if you ask for alot you get alot
> > and some failures.
> > Itd be certainly possible to make qual tasks and soc projects so simple
> > that we had a maybe 70% success rate but the code that would come out of
> > this would be less than what we get now.
> 
> You are missing the point.  Students and tasks should be evenly matched
> for optimal results.

Students submit applications so they choose what they want to work on
to match students better, mentors would need to know their abilities
better and there also need to be matching applications submited by the
student.
Within these constraints we surely have tried our best in the past, you
dont know it but we did reject applications from students when we where
pretty certain that the student was not able to finish the work in any
reasonable timeframe.


> 
> > > > > [17:51] <Dark_Shikari> For example, if you were writing an h264 decoder, you would have to write a header + bitstream parser
> > > > > [17:51] <Dark_Shikari> that printed dct coefficients for cavlc streams
> > > > > [17:51] <DonDiego> yes, i notice that many people from there apply and i suspect that money could be a motivation, there's $4500 to earn, i can live off that kind of money for 1/2 year in germany..
> > > > > [17:51] <Dark_Shikari> that kind of thing is _useful for your actual project_
> > > > > [17:51] <kierank> the x264 qualifications are meatier than ffmpeg ones
> > > > > [17:51] <DonDiego> mru: jai has just taken money and done nothing in return :-/
> > > > > [17:52] <Dark_Shikari> 2) You must be on IRC often, and get to know the community
> > > > > [17:52] <Dark_Shikari> you must be involved
> > > > > [17:52] <Dark_Shikari> you must respond to pings
> > > > > [17:52] <Dark_Shikari> If I ping you and say "there's a bug in your code here", you should be able to respond
> > > > > [17:52] <Dark_Shikari> obviously, I can't control timezones/sleep schedules/etc
> > > > > [17:52] <mru> DonDiego: jai has done stuff though
> > > > > [17:52] <Dark_Shikari> but in general, you shouold be available.
> > > > > [17:52] <kierank> 22:42] <@Dark_Shikari> you must respond to pings --> like janihan ;)
> > > > > [17:52] <Dark_Shikari> lol
> > > > > [17:52] <DonDiego> mru: that was last year, this year it was money for nothing..
> > > > > [17:53] * mru takes the chicks for free...
> > > > 
> > > > > [17:53] <Dark_Shikari> I don't think the x264 requirements are harder than ffmpeg
> > > > > [17:53] <Dark_Shikari> the ffmpeg qual tasks are really hard
> > > > > [17:53] <Dark_Shikari> and don't even relate to the main project often
> > > > > [17:53] <Dark_Shikari> I think the main thing is that we create a community and force people to be involved
> > > > > [17:53] <Dark_Shikari> and if they can't stay online and involved, they don't get in
> > > > > [17:53] <mru> that's how communities work
> > > > > [17:53] <mru> _saying_ you're in doesn't make it so
> > > > > [17:54] <Dark_Shikari> yup
> > > > > [17:54] <Dark_Shikari> having a qual task that leads into the main project is the best idea of mine IMO
> > > > > [17:54] <Dark_Shikari> it gives people a feeling that they're actually part done with the project
> > > > > [17:54] <Dark_Shikari> before SOC even starts
> > > > > [17:54] <Dark_Shikari> and people are less likely to quit in the middle
> > > > > [17:55] <Dark_Shikari> as opposed to simply bullshitting and quitting after getting nothing done
> > > > > [17:55] <DonDiego> we need to work on our soc, it's in a sad state and not improving
> > > > 
> > > > > [17:55] <DonDiego> also, our tasks are *much* too hard
> > > > > [17:55] <Dark_Shikari> well, here's my advice for improving ffmpeg SOC
> > > > 
> > > > > [17:55] <DonDiego> basically the only student that ever finished a decoder is kostya
> > > > 
> > > > complete nonsense
> > > 
> > > Absolutely not.  There is only one non-Kostya-implemented decoder that
> > > made it into FFmpeg start to finish during SoC: MPEG-4 ALS.
> > 
> > Thats nonsense, just count how many we just got as a qualification task
> 
> This is nonsense, they're much easier than the ones that make up SoC
> tasks of course.

I thought someone said they where too hard as well


> 
> > also, ffmpeg is made of many parts, encoders, decoders, muxers, demuxers
> > and alot in between. If you split the students we have over these there arent
> > that many left in each category, now saying there are just 2 in one specific
> > category is a little weak.
> > 
> > just a quick grep for a random student name:
> > libavcodec/ac3dec.c: * Copyright (c) 2007-2008 Bartlomiej Wolowiec <bartek.wolowiec at gmail.com>
> > libavcodec/ac3dec_data.c: * Copyright (c) 2007 Bartlomiej Wolowiec <bartek.wolowiec at gmail.com>
> > libavcodec/eac3dec.c: * Copyright (c) 2007 Bartlomiej Wolowiec <bartek.wolowiec at gmail.com>
> > libavcodec/eac3dec_data.c: * Copyright (c) 2007 Bartlomiej Wolowiec <bartek.wolowiec at gmail.com>
> > libavcodec/lzwenc.c: * Copyright (c) 2007 Bartlomiej Wolowiec
> > libavcodec/lzwenc.c: * @author Bartlomiej Wolowiec
> > libavcodec/nellymoserenc.c: * Copyright (c) 2008 Bartlomiej Wolowiec
> > libavcodec/nellymoserenc.c: * by Bartlomiej Wolowiec
> > libavcodec/tiffenc.c: * Copyright (c) 2007 Bartlomiej Wolowiec
> > libavcodec/tiffenc.c: * @author Bartlomiej Wolowiec
> > libavformat/spdif.c: * Copyright (c) 2009 Bartlomiej Wolowiec
> > libavformat/spdif.c: * @author Bartlomiej Wolowiec
> 
> That's a straw man.  I mean what I say, not what you put into my mouth.
> 
> I never claimed we got nothing out of SoC nor am I unaware that demuxers
> are easier to write than decoders.  I was talking about decoders and my
> point still stands.

10 matching in libavcodec 2 in libavformat, we dont have demuxers in
libavcodec
and thats just one student ...


> 
> > > Our SoC tasks are too hard, it's a fact.
> > 
> > thats your oppinion
> > and as you consider to apply for this years soc ill surely not consider
> > your oppinon as it could be heavily biased due to personal interrest.
> 
> Everything is perfect.  You assess reality without flaw.  Nothing needs
> to be changed, ever.  FFmpeg must continue in all regards without
> adjusting course in the slightest.

we adjust course, there where many adjustments over the years with respect
to soc. For example when kamil failed to write a j2k encoder and decoder
i insisted that we will not accept such combined tasks in the future or
make it clear that the decoder is the primary goal and the encoder optional
kamil did work and still couldnt finish the code.
Also due to dissapearing students we made more certain that people who
did zero work would end up "disqualified" and not get more payment than
had already been sent out.
...


> 
> I'm sick and tired of these discussions.  First you complain loudly
> that you want to hear about all the problems that get talked about
> "behind your back", then you shoot at the messenger.

I do want to hear about the problems, several have been raised here in
this thread, that i will try to look into like spliting svq3 out of h264.
Also theres jasons ideas of using IRC more for SOC student communication
and making qualification tasks more related to the actual SOC project.
Also the issue of reviews being too picky on irrelevant details has been
raised by jason and i belive vitor too has mentioned something similar
recently.
But listening to what people consider problems does not mean i have to
agree on all points raised. As mentioned above i agree with many but
i dont think SOC tasks are too hard also even if so students submit
them not us. If a student claims he can do it its not always easy for
us to judge if he is capable to do it.
Besides this students are being paid 4500USD so accepting completely
trivial tasks that i could do in a few hours does not seems fair to
the others.
That said i dont remember straight of my head an application we rejected
due to being too trivial but there probably where some.

besides you didnt post the irclog so you arent the messenger

That said, what simpler tasks would you have in mind that we should
suggest students to pick for their applications?


> 
> Fuck it.

iam not sure how i should interpret that, could you elaborate?


>   I'm not making the mistake again.  You know what happened to
> Cassandra when she stayed in Troy.  

actually no i dont, which of homers works do i have to read?


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

Those who are too smart to engage in politics are punished by being
governed by those who are dumber. -- Plato 
-------------- 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-cvslog/attachments/20100111/20ce072f/attachment-0001.pgp>