Hi Dave, thanks for your thoughtful response.. my comments below.
On Thu, Jan 6, 2011 at 7:32 AM, Dave Neary
<dneary gnome org> wrote:
Hi,
Allan Day wrote:
> Sriram Ramkrishna wrote:
>> I think in this case, we want to be a buffer between devs and
>> enthusiasts.
Ah... I'm not a big fan of buffers. I prefer trying to quieten people
down when they're distracting. That means channelling the noise
elsewhere, and perhaps taking a slightly drastic measure of moderating
all posts to d-d-l for a few months.
I think that is the action, but to be more succinct, I am saying that devs don't have to do this. For the most part they aren't but there are a number of us who are advocating for gnome-shell but we're doing it in a vacuum. I'm merely trying to get those people together to do exactly what we are doing now but in a more cohesive manner.
If you moderate, you'll have a lot more angry people since it seems like you're trying to shut them up. They'll complain somewhere else. Better to let them talk and if they break the Code of Conduct, then we have grounds to kick them off the mailing list or put them into a position where all their posts have to be approved. Olav did something similar for Lefty on the foundation-list. That is a good example of managing the mailing list. It's imperative we let them talk and just as imperative we respond with data. And continue to respond. This wears out devs but it won't wear us down we're emotionally geared for it.
>> In general, I want to address the frustration devs feel
>> with having to continually defend design decisions by people who fear
>> change.
Characterising any questions/queries/doubts/fears about GNOME 3 as
"people who fear change" is not going to win friends and influence
people. We need to start framing the way we talk about this issue
correctly, because the way we talk about it affects the way we think
about it, affects the way others perceive us.
OK, point taken. Let's modify that argument instead to "people who are worried as there isn't any documented evidence that this design is going to work". We do not have official Gnome usability studies to back up our design of shell. We need something that we can point to that says that we thought of all possibilities. When there is a lack of evidence then indeed people do fear change. I know Jon and others are using research material, books and what not to back up their design and they have done some internal testing. None of which is public.
>> But my motivation is really to get people off devs back so that they
>> can focus on getting the design done.
So - to summarise: the problems I've seen are:
Identifying the problems and figuring out how to resolve them the first step at all. So I'm glad we are having this conversation because nobody else. Managing the people side is important since we do need to depend on these people to spread word of mouth on Gnome 3.0.
* A lot of GNOME members (be it foundation members, translators, members
of teams not directly involved in the shell, even members of the release
team, but also those "shrill activists") do not have all of the
rationale & thought processes that have gone into GNOME Shell. Most
people have not been following the shell list or the release team
archives. So some major decisions are coming out piecemeal and are
presented as faits accomplis - "this is the way things are, live with
it". The shell team and release team have not been their own best friend
in this respect.
This. You've definitely identified one problem here. There is no action to plan so to speak. We look for documentation and there is nothing there. Right now, I feel new features are coming out in piecemeal. I think that's somewhat OK, provided that we have a quarterly update on what that is so that we can at least point to it somewhere in a clear concise manner.
* There appears to be cognitive dissonance between the resources that
some GNOME people believe are there for maintenance and the resources
that are actually there - esp. related to fallback mode - panel +
applets + metacity.
I'm not sure I understand this point regarding maintenance. Are you referring to the maintenance of gnome 2.0?
* A lot of people (myself included) are worried whether we're going to
need hardware that a substantial proportion of our user base just don't
have.
I'm less worried about this. I make the assumption that clutter is going to work on any hardware where compiz works. In which case, I would argue that Ubuntu is making the exact risk we are by moving to Unity. Their "fail back" is on the same metacity, or probably whatever other third party window manager is out there. I rather we work harder on making clutter work on as many hardware devices to get as many of our existing user base as we can.
And we can fix all these things by:
(To fill the knowledge gap:)
- Documenting any major design decisions made in the shell and their
rationale, and presenting those in a nicer way
- Ensure that the release team is also framing things in a nicer way,
and explaining the rationale behind things rather than just saying
"that's the why". Notably, I think the release team should say that
panel support is a question of manpower not policy, and it'd be nice if
Jon backed that up.
- We need to start talking about all that's good in shell, what it
brings to the table, and why it's better than what went before. Videos,
success stories, interviews with happy users, all that kind of thing
I think this is pretty much what community management is. :-) Talk to stakeholders, get them to document what they are doing, give a unified consistent view of Gnome 3.0 to the community that we've extracted from stakeholders.I've been bothered by the whole communication of Gnome 3.0 and its feature set for a couple of months now just from the perspective how we are going to market this.
I think you're list is a good start.
(To be clear on what we want to see done but can't commit to, versus
what we definitely don't want to see done)
- Help the release team and the shell team draft a "Here's stuff we're
against" list (with justifications) and "Here's stuff we'd like to see
in GNOME 3.0, but there's just no way we can commit to getting it done
with the resources we have" (things like Orca) - I'd also live to see
the GNOME project as a whole be clearer on who's involved in the actual
maintenance of core modules and what developer resources there actually
are - I'd estimate it at 40 to 50 full time developers, but I suspect
I'm on the high side there.
I think as I pointed out before, not everything is going to be feature complete at launch. It will take some cycles before we are on par with Gnome 2.x. That especially is important since a lot of people expect to just change over. A document on who should switch might be in order.
(To allay fears of hardware compat issues)
- Draw up a list of graphics cards used in desktop hardware in the past
3 to 5 years, ordered by market share/volume of sales, and match that to
current compatibility of free software drivers for the hardware with the
3D requirements of Clutter. Ideally, someone with a GNOME 2.x desktop
should be able to run a command, get a chipset, and be able to tell
before he installs how well GNOME 3 will work.
I wonder if Emmanuelle has some kind of perf tool for clutter. That would make things simpler with an auto submission.
- Start publishing screenshots of fallback mode on chipsets that don't
support it & making sure it's still a nice experience.
Agreed.
> Personally speaking, I'd add a few other ambitions:
>
> * Ensuring that those from outside the project have a positive
> experience when they come into contact with GNOME. A bit more
> micromanagement here would help our public relations, I think.
This is a vast task, and I'm not sure how we can make a big dent in it
with limited volunteer resources - even doing basic communication of the
project goals and allaying concerns people have is probably more than we
can manage, but it is pretty much a minimum.
Yes, and that's why I believe that talking to people who are doing this engagement already and working together will help.
> * Communicating and explaining the direction of the project. 3 dot oh
> involves some big changes and our community would benefit from some
> positive messages in that regard.
Absolutely agree.
> Do those sound useful?
>
> Also: GNOME 3 is going to be met with criticism. We don't know how much,
> but there's definitely going to be some. Sri made a really good point to
> me the other day - we need well-rehearsed and effective responses to
> those criticisms, and this kind of community management is an
> opportunity to identify and rehearse those arguments. Maybe we could
> work together to produce some kind of PR play sheet? Or would that be
> taking things too far?
A FAQ will help for the most predictable criticisms, and a list of
talking points (the "what's great about GNOME 3" list I mentioned above)
should help us frame the messaging around the release in March, and
everyone giving interviews or presentations should have these down pat.
Shell has a FAQ it's not really getting updated. In fact we have all kinds of documentation, but it's not presented in a single area (eg
gnome3.org) and it is not present. Secondly, no matter how much documentation you have out there people will not read it. That's why engagement has to be part of the equation. Also we need to constantly in the front page and others to point to gnome 3.0. A banner on w.g.o saying "are you ready for gnome 3?" pointing to the website would be another refinement.
That's a big list of stuff though... Who can give time, to do what, over
the next few weeks?
I'm willing to do some work. I have a bit of a break before things start picking up again.
sri
