FESCo Meeting Summary for 20090424

Adam Williamson awilliam at redhat.com
Sat May 2 05:24:33 UTC 2009


On Sat, 2009-05-02 at 01:55 +0100, Matthew Garrett wrote:
> Put simply, any argument based on "Nobody's complained, so it's fine" is 
> fallacious. People expect Linux to be dreadful. If something works 
> they're happy - if it doesn't, it's because Linux isn't ready for the 
> desktop yet. And when we've still got a community that's more inclined 
> to tell people that they should read the documentation rather than ask 
> whether a specific UI is sufficiently understandable, it's hardly 
> unsurprisingly that people aren't going to spend a great deal of time 
> complaining about how unintuitive they found some aspect of UI.
> 
> If people raise issues with a suggestion and the counterargument is 
> "Users haven't raised this problem" then we can't draw conclusions about 
> why they haven't. Maybe it's because the average user is smart enough to 
> figure out that the multitude of volume controls we ship are all 
> intended for subtly different purposes. Maybe it's because they can't be 
> bothered filing bugs. Maybe it's because they're scared of raising a 
> contentious subject. Who knows? In the end we still come down to making 
> decisions based on the opinions of people we deem to be experts in the 
> field. And if you don't trust the desktop team to make the appropriate 
> decision in this case then it would be helpful for you to say so 
> plainly.

Let's see. I don't think you're really wrong. It's not a perfect basis
for argument. However, let me see if I can summarize the desktop team's
position:

"We're in the best position to know what's right for the desktop, and we
say that the best thing to do is just ship one mixer application. The
one that can't handle anything except analog stereo output when there's
no problem with ALSA. We consider surround output, digital output, input
switching, input monitoring and all the known and as-yet unknown but
undoubtedly existent cases where there *is* a problem with the ALSA
mixer and hence g-v-c doesn't work properly to be, as a whole, not
significant enough to make it worthwhile shipping a different mixer,
either alongside g-v-c or instead of it.

We're the best people to make the decision because we know what we're
doing when it comes to building a desktop. Yes, we've actually been
shipping two mixers by default for three releases now (and we didn't
notice this until someone else pointed it out), but that's just, er, a
mistake. We're still the *only* ones who can be trusted to know what
we're doing here!"

I'm really, honestly, not in a nasty way, not finding that to be 100%
convincing.

I'm not working on the basis of "do I trust the desktop team to make
this decision and get it right". That's not really how I'm looking at
it. I'm just evaluating this particular case, on its merits, and I don't
think the call that the desktop team made was the right one. I don't
really care about the reasons behind that or the politics or yadda yadda
yadda. I certainly agree that the process could have been managed better
- objections could have been raised much earlier - and that's something
it's entirely worth dissecting...but I don't think it should take
precedence over making sure Fedora 11 is the best product it can be. I
just don't think shipping a mixer that doesn't handle analog surround
output, digital output, input switching, input monitoring or the
multiple cases of buggy volume supplied by ALSA (there's six of these
bugs filed so far, btw, and counting) as our *only* graphical mixer is a
good decision.

I raised this concern through the proper channels, it was picked up by
FESco - not at my instigation, btw, I was called into the meeting after
the topic had already come up - and FESco came out in the same position
I've been espousing. At that point I thought the whole thing was settled
and I went on to actually work on the technical side of what needing
doing, whereupon the whole argument got dragged out of its grave on this
list and we had to go through everything again, three times over.

I'm sorry if I don't handle the political side of things that greatly,
and I'm sorry if I sometimes come off as too aggressive. That's my
fault. I'm not trying to rub anyone up the wrong way. It's just a bit
frustrating that no-one seems to want to engage in the actual blood and
guts of this issue; all I'm hearing is vague theoretical arguments about
why *this* group of people is better entitled to make the decision or
why *that* line of argument is theoretically flawed or yadda yadda. I'm
just not finding it very productive.

Please credit me with a bit of experience and knowledge in this field
too...yes, I've not been around Fedora very long, but I *have* been on
the very sharp and pointy end of distribution releases at Another Place
for three or four years, being The Guy for several tens of thousands of
people to yell at when they don't like something in a new release, also
being the guy who tracks release critical bugs and writes all the
release documentation, including the Errata. I think I have a pretty
decent track record of knowing a hot-button issue when I see one, and
the idea that we should ship the new gnome-volume-control as our sole
GUI volume control application on the desktop operating system images
that most people think of as 'Fedora' (the default install you get off
the DVD, and the GNOME live spin - the live spin you get if you don't
explicitly select a different one, and the one we mostly hand out at
events, covered in Fedora logos) just does not feel like a winning
proposition to me. All I've been trying to do is back that position up
with solid reasoning and logic.
-- 
Adam Williamson
Fedora QA Community Monkey
IRC: adamw | Fedora Talk: adamwill AT fedoraproject DOT org
http://www.happyassassin.net




More information about the fedora-devel-list mailing list