Re: GNOME and GStreamer

[Thomas Vander Stichele <thomas apestaart org>]

I'm replying to ronald below, but I'd like some advice from the release
team on the situation at hand.

Personally I'd like to do a gnome-media release this Monday, together
with the releases of GStreamer, so this version gets testing and
feedback.

I have the feeling Ronald doesn't want this to happen.

I have no interest in dragging this into an escalating I-can-so-I-will
situation, but I'd hate for GNOME to pick a GStreamer backend for the
next GNOME release two months from now, picked up by distros a few weeks
later, and needing support for the next year, when realistically
speaking no one is doing any actual hacking or maintenance on GStreamer
0.8, just because we let this decision slide by.

Both the GStreamer community and Fluendo have spent a lot of work on
GStreamer 0.10 - it'd be a shame not to reap the rewards.

Release team, please let me know what to do.

On to Ronald's mail ...

> First of all, because of your objections (and only because of yours)
the
> > committing was done one a branch.  AFAIK everyone is free to make
> > branches.
> 
> You could've asked to be sure that I would say no.

I don't understand what you're saying.

>  I've been the only
> active maintainer in the past >1 yr.

You've definately been a lot more active, I haven't hacked on
gnome-media for a year, primarily because the pieces I had been working
on were satisfactory to me, because GStreamer itself needed work, and
because gnome-cd has been neglected generally because everyone seems to
want it to die (I don't) and get replaced by totem and sound-juicer
(their maintainers don't agree).

OTOH, you've not done much on it lately either, and no new release was
ever made for this development series.  Now that something has come up
that I care about, I'm putting in time and effort again to integrate
GStreamer 0.10

> 
> > Why are you being this contrary ? I realize you've invested a lot of
> > time and effort in GStreamer 0.8 (as have we), but I don't understand
> > why you're actively resisting a move forward. 

> I want GNOME to have as good media support as possible for end users.
> Those are people that buy mp4 songs and watch quicktime trailers from
> Apple, or download mp3s/movies online, or watch DVDs/VCDs, or view WMV
> movies on msnbc or the likes, or put their whole song collection on
> their ipod or mp3 player (regardless of the source or target format) or
> buy music CDs and import that to their computer or other similar things
> or combinations of the above; people who are not native in English will
> use foreign subtitles, and people not only want media players, they want
> a webcam capture application, they want a tv application with capture
> support, audio/video editors, audio recording, use media in
> presentations, and so on and so on. GStreamer, in all those, is a
> toolkit, not an application. We don't make decisions, we provide the
> platform. The fact that you believe that doing anything besides Ogg is
> illegal is irrelevant for both developers and most end users. They just
> want to do it.

You make us out to be a lot more different than we are, please stop
polarizing.  I do care about all those other formats, and I've never
said or believed that anything but Ogg is illegal.

All I've ever said is that GNOME as a project that ships an LGPL/GPL
platform from the US cannot afford to ship a GPL/LGPL stack tangled up
in patents.  Everyone agrees with this - see spring-loaded folders.

Very little of what you just said applies to gnome-media at all by the
way.  The only thing in gnome-media that could possibly deal with
patented formats is gnome-sound-recorder.

>  I've put in a lot of effort in I-don't-care-what-version
> to make sure that users would be able to enjoy the until then
> theoretically marvelous but practically useless stable version. That's
> been a pain and costed me many 24/7 weeks (I've practically thrown away
> one study year by doing that), most of which is unpaid personal free
> time. Many people were thankful for it, but all that work has been
> thrown away, we're back at square one.

Many people are still thankful for it.  Your work has not been thrown
away.  Neither has anyone else's.  I tested jhbuild/totem with
gst-ffmpeg for 0.10 on my home machine today, it played my divx's just
fine.

What you are suggesting is that all the work to move forward on 0.10 -
and there's been a lot ! - should be thrown away.

>  However, my work to get support for those use cases
> back in the new 0.10 tree was actively obstructed by formal
> manager-driven (?!?) design rules and other political bullshit,

That's not true and you know it.  For those not involved - the
discussion revolved around a solution for sparse media streams, in the
form of filler events/buffers.  There's been a lot of discussion on this
topic and in the end it was decided to remove the concept from core, for
the *express purpose* of adding back a good solution (which can still be
this filler concept) *when someone works on it*.

What was done there was good - it means that the correct solution can be
added when someone puts it together, instead of ending up in a situation
where we have to support the API, even if the API is wrong.

As long as it's not actually implemented in 0.10, it's impossible to say
if the API is correct.

Note that nobody was stopping you from actually doing this - but you
didn't have time, which we respect - , and we'd still be happy if you
did this in 0.10. But you'll have to *do* something about it, not
complain about it because you're not getting your way automatically.


>  which
> caused me to resign as developer, since I have no interest in being told
> what to do in my free time by other, less capable people.

I didn't know the fit you threw on irc when you were last there was to
be interpreted as a resignation.   I don't even know what that means -
#gstreamer isn't very formal when it comes to its maintainer's
structure.  I also don't know exactly who you're trying to insult with
your statement, feel free to share :)

>  That's why I
> think 0.10 is no more than a theoretical blib right now. When it's
> reached end user usefulness, I'll be interested.

We'd appreciate you'd actually test 0.10 at some point.  We know
releasing it on schedule was a shock, but you should have recovered by
know.

Thomas

Dave/Dina : future TV today ! - http://www.davedina.org/
<-*- thomas (dot) apestaart (dot) org -*->
If you can call this luck
if you can call this love
if you can miss this much
<-*- thomas (at) apestaart (dot) org -*->
URGent, best radio on the net - 24/7 ! - http://urgent.fm/