fedora mission (was Re: systemd and changes)

[Jon Masters]

On Fri, 2010-08-27 at 15:23 -0700, Jesse Keating wrote:

> >Again, I feel it is necessary to have a survey of Fedora users.
> >Preferably annually. And listen to the feedback. If they say "yep, we
> >just love the churn, the number of updates" and so forth, then fine. If
> >they say "actually we'd like less than 800 updates after installing",
> >then also fine. I'm sorry to beat a dead horse, I just feel it is very
> >important that we finally, clearly articulate who our users are and what
> >they want by treating more like customers and gathering their input.

So I'm filing another FESCo ticket as I type this. They can decide to
reject my proposal again, in which case I will bring it up with the
Fedora Board by way of appeal. I will take "no" for an answer once it's
been escalated all the way, because then at least I will have tried.

> The cynic in me would expect that the people who want something
> different than the fire hose we have now are silently leaving,
> and those that are left are going to say they like the deluge of
> updates.

On some level, if they say that, I suppose I get to eat my hat and deal
with the deluge. At least we'll know. Maybe we'll be surprised :)

> Or I could just quote Henry Ford, or any other people who talk
> about design by committee.  

But I feel it is "design by committee" now, just the Open Source
version. A small group of people (f-d-l and similar) arbitrarily decide
what the userbase "wants" without asking them, and based on what they
would like to see in the distribution themselves. There are elections,
but only those with a very strong involvement are motivated to stand.
And it's hard to vote if you're not very informed on the issues, whereas
almost everyone and his/her dog has an opinion in a poll. Yes, that can
be bad too (people manipulating the survey, etc), but there are ways to
write polls and conduct them in a more scientific way that is /less/
(but not perfect) vulnerable to that. And it's better than /no/ data.

Look upon this as a democratic action, an opinion poll, a survey. The
kinds of things governments, corporations, non-profits, and other groups
do to find out how they're doing and what could improve. If over 50% say
they like the rate of updates and are willing to accept breakage, then
fine, we can quote that number and I will have to eat my words. It
doesn't mean the end to the discussion, but it provides data. If the
results are very skewed in favor of churn, I will then also owe Kevin a
beer or two for being in the minority opinion, and maybe eat a hat.

Jon.