Re: GStreamer regression analysis [was: GNOME and GStreamer]

[Luis Villa <luis villa gmail com>]

On 1/16/06, Christian Fredrik Kalager Schaller <uraeus linuxrising org> wrote:
> Hi Luis,
> You make some valid arguments, but here is my take on it.
>
> GStreamer 0.10 do have feature equivalence to 0.8 in terms of the free
> formats today. There are a few fixes that is needed in Totem still as
> listed by Ronald, but all in all these are minor issues considering we
> have as a paid developer (Tim) working on this.

They don't sound like minor problems at all, honestly- they sound like
blockers from calling the product stable. But I'm glad to hear that
they are fixable and hopefully will be fixed by release time

> The question is about the non-free stuff. For common formats like
> Quicktime, AVI, ASF/WMV, MPEG2, MPEG4 and Real. We will be as good or
> better than 0.8 by the time of 2.14. The same goes for re-enabling
> subtitle support. This is on Tim's todo list too.
>
> For weird stuff like some video game formats we will have a regression
> when 2.14 comes out unless someone steps up to try to fix it.

I'm not terribly concerned about the weirdo formats, and I agree that
Ronald is being slightly obstructionist in bringing them up. But the
major non-free formats are, realistically, big and important issues
for our users. Abiword wouldn't ship without support for .doc and
Firefox wouldn't ship with a plugin API that broke Flash. I don't see
why GNOME should hold itself to a lesser QA standard.

> For DVD playback there will probably be a regression, but here I have to
> say the release team can blame nobody but themselves. Thomas warned you
> against proclaiming DVD support with the previous GNOME release and you
> did it anyway. Ronald who promised to work on making sure DVD support
> would continue to work hasn't done a rats ass of effort in regards to
> DVD in 0.10 even though he even at the time of the last GNOME release
> should have known that 0.10 was close on the horizon and that it was
> targeted towards 2.14.

Again, users don't care whose fault it is. They care that something
that did work (even partially) no longer works. They won't write blog
posts and reviews that say 'man, Ronald Bultje sucks'; they'll write
blog posts and reviews that say 'man, GNOME sucks.' My job, such as it
is, is to avoid that second situation, and as far as I can tell,
gstreamer 0.10 is not really going to help me there during this
release cycle.

Again, I understand that there are perfectly valid engineering reasons
why this is the case, and I agree we should push hard to make sure
0.10 is tested and developed so that it is ready at some point in the
future. But responses like 'that isn't free' and 'that isn't our
fault' do not do much to convince me that we even agree on the
problem, much less that we agree on the solution or course to take
from here.

Luis

> If the release team want to save some face on the issue of DVD playback
> maybe Martin Soto is willing to commit his Seamless DVD player (which
> already use GStreamer 0.10, but in a different way than needed for
> Totem) to the GNOME release cycle which probably would give a better DVD
> experience than Totem can anyway. I don't think we should though for
> both legal and moral reasons.
>
> Christian
>
> On Mon, 2006-01-16 at 14:00 -0500, Luis Villa wrote:
> > On 1/16/06, Tim Müller <t i m zen co uk> wrote:
> > > As a GNOME user as well as a developer I really hope that GNOME will
> > > decide to go for the GStreamer version that is actively being worked on
> > > and constantly being improved, even if that means temporary feature
> > > regression.
> >
> > I've bit my tongue so far in this thread, but this really sort of
> > irritated me. End users do not know or care that the version is being
> > improved/worked on. What they definitely do see is regressions- 'I
> > upgraded and now I have fewer features and fewer things work.' No
> > normal user will think 'I upgraded and now the software is shittier,
> > but hey, I'm sure that means that in six months it'll be better!' To
> > think that end users will understand why their software suddenly works
> > less well is insane and, frankly, I think it is indicative of a
> > long-standing problem in the gstreamer project where the focus is more
> > strongly on the core technology than the actual user experience.
> >
> > >From what I can see, the real problem here is that gstreamer 0.10 is
> > really an api-frozen development branch still, which may be suitable
> > for server use with some codecs but is not an actual
> > feature-comparable end-user stable release. I'd urge everyone to
> > re-read this thread substituting 'api-stable development branch' for
> > '0.10' and see if it changes their thinking.
> >
> > Now, it might not- there are plenty of legitimate reasons I can see to
> > switch to 0.10 anyway, even if it should still be called 0.9.90.
> > Primary of them, of course, is that if we ship with 0.8, no one will
> > fix any reported bugs, and I can't blame the gstreamer/fluendo team
> > for that, given their limited resources. So possibly calling it 0.9.9x
> > instead of pretending it is stable and featureful doesn't change
> > anything.
> >
> > >From what I've read so far (and I admit I have not come close to
> > reading the whole thread) the right thing to do here is to encourage
> > developers and testers to use 0.10, and aim strongly for using 0.10 in
> > gnome 2.16, but to assume that we'll officially depend on 0.8 and
> > recommend it to distributors, end-users, etc. Yes, it'll be
> > effectively unsupported, but it will be exactly the same in those
> > terms as 2.12, whereas 0.10 will be a big step backwards in many
> > respects from 2.12, which should be unacceptable for us and will be
> > damaging to our users and our hard-earned reputation for stability.
> >
> > Luis
>
>