FESCo Meeting Summary for 20090424

David Woodhouse dwmw2 at infradead.org
Sun Apr 26 08:38:52 UTC 2009


On Sat, 2009-04-25 at 17:15 -0500, Callum Lerwick wrote:
> On Sat, 2009-04-25 at 14:58 -0700, Jesse Keating wrote:
> > On Sat, 2009-04-25 at 16:45 -0500, Callum Lerwick wrote:
> > > A compromise that does absolutely nothing, that is 0%, to address my
> > > concerns is productive how?
> > > 
> > > Have the concerns of the person that sparked this whole thing become
> > > irrelevant at some point along the way?

Actually I raised the issue with FESCo based on something else, although
your issue and others certainly contributed to my concerns.

My original thought was just that FESCo would gently prod the packagers
who were closing regression bugs as WONTFIX, and encourage them to
behave more responsibly. When it became apparent that gentle
remonstration with the people concerned wasn't going to be sufficient,
something else came out of the discussion instead.

> > > This thing has certainly spun far out of my control, and no longer even
> > > has much to do with what I wanted in the first place.
> > 
> > Unfortunately it wasn't just your needs we were trying to address in the
> > meeting.

People will always need access to mixer controls. One set of people will
need them because they want to do things that the PulseAudio folks call
"weird", like using that line-in socket on the side of their laptop, or
playing CDs without chewing CPU time doing all those strange unreliable
heuristics we do to knit audio back together when we rip it off a CD. Or
turning their speakers on or off. Or setting the relative levels of bass
and mid-range speakers. Or any number of other things.

Other people need it just to get optimal use out of their hardware in
the "non-weird" cases. Although we're building up a hardware database
which tells us which of those sliders in the alsa-mixer-of-doom actually
do useful things, it can never be completely accurate and up to date.
People will _always_ be able to do better by experimenting for
themselves.

For example, if I want to get the best audible volume out of my laptop,
I may need to tweak the various mixer levels _just_ low enough to avoid
clipping, and the levels required depend on a lot of things -- including
the analogue circuitry in _this_ particular machine, and the ears of the
people present in the room. That information isn't going to be in the
database.

We have to accept that there are _always_ going to be cases where users
want better control over the mixer.

> Just sayin'. I'm sure not getting anything out of this so far.

Callum, I'm sorry we didn't address the panel issue. I wanted to talk
about the panel, but it quickly became clear there wasn't sufficient
consensus on that.

I think it's a fairly poor compromise that we ended up with -- but
that's kind of the point in compromise; it's not really _anyone's_
perfect vision.

On the whole, I _like_ the new F-11 gnome-volume-control -- I just wish
we could just make it work for the cases it currently doesn't handle.
That would be the ideal solution.

  (Obviously not just by adding more sliders; it needs to be possible
  for users to enable what they _need_ without just showing them the
  alsa-mixer-of-doom by default. I'm not really qualified to give a
  detailed suggestion on _how_ it's done -- that's a UI decision which I
  wish the UI experts would address. I wish they _wouldn't_ try to
  address it by simply declaring these uses out of Fedora's scope.
  That's a policy decision which I think it's properly within FESCo's
  mandate to decide.)

But since there's such resistance to doing that, shipping an alternative
volume control app at least restores functionality to those users who
need it.

It's even more suboptimal that this app is hidden in the menus and not
accessible from the panel. As far as the F-10 user is concerned, this
functionality is just _gone_ from the place it used to be; I'm not sure
they'll think to go looking around in the application menus to find it
again.

But that's the compromise we agreed on. It's far from ideal, but at
least it's only a regression in _usability_ for those users, not a
complete loss of functionality.


To be honest, I think the best option at this stage would really have
been to revert the VolumeControl feature for F-11, and to say "come back
and try again in F-12, and please do a better job next time". But I
definitely don't think we have FESCo consensus on _that_, so I didn't
push it very hard.

-- 
David Woodhouse                            Open Source Technology Centre
David.Woodhouse at intel.com                              Intel Corporation





More information about the fedora-devel-list mailing list