On Sun, 16 May 2010 07:45:53 -0600, Jeremiah Foster 
<[email protected]> wrote:
> On May 16, 2010, at 1:52 PM, David Greaves wrote:
> > Carsten Munk wrote:
> >> So, this is primarily an e-mail to ask some questions to the two
> >> members of the TSG, that I think would not be sufficiently covered in
> >> a TSG meeting and answers might be better suited for the e-mail form.
> >> 
> > I think there are others who can contribute to the conversation too.
> > 
> > There are systems around meego that are only open to members "for technical
> > reasons" so the discussions around them could be opened up. (Yes, there'll 
> > be
> > some frustration at inaccessible links ... but that may help prioritise 
> > which
> > systems to allocate "opening-up" resource to. For those involved : Why isn't
> > this happening? What do you need to start discussions on open mailing lists?
> 
> Open Source is not just open source code - it is an open way of
> working. This means that we can have an opinion on the default file
> system in MeeGo and if we feel that btrfs is not yet ready, we can
> push ext3 to the repos to give ourselves a choice. The open source way
> of working does not yet exist in MeeGo, it is not yet an open project
> just a collection of git repos on gitorious.

I think it's a bit more complicated than that. Open source projects have
maintainers and maintainers make decisions. We are trying to open these
decisions to public discussions, but at the end of the day the
maintainers make the decisions. We can always have custom spins that do
things differently, but for the default builds, someone needs to make
the call - and that's what's happening.

I think this will feel much more natural as we have more maintainers who
don't work at either Intel or Nokia - but trust me, the btrfs decision
is something that has been debated hotly at many levels before (and I
haven't seen a single new argument here on this list) and at the end
Arjan made the call.

And just to make this crystal clear: if one of you had brought up
something new, the magic showstopper that says "this won't work", then
absolutely we would change the decision.

> >> We're not working in the open like we're supposed to - even though as
> >> has been said - Intel, Nokia and we all know how to do it! But when
> >> there's a big reveal mentality active, the mode of the people
> >> participating switches to internal/private development, even if you
> >> are only tangentially related to the object/UX being revealed.
> >> 
> >> And I think that's the main source of all our so called 'openness'
> >> problems - not malice, conspiracy, laziness or whatever.
> 
> I don't think there is a conspiracy or anything malicious
> either. These are two very large companies who now see the advantages
> of Open Source. It is a lot to expect that they will suddenly 'get it'
> and OPK and Ottellini will be showing up at FOSDEM with a penguin
> T-shirt and a UNIX beard.

I'll make that proposal tomorrow - I can't wait to hear PSO's answer (I
think it'll be "no" on the beard).

> What is happening is these two large
> dinosaurs are learning to dance to an open source beat - that is
> pretty cool. I think we all need to have a bit of patience and be more
> proactive in helping them with the transition from proprietary and
> closed to open and shared. Your email was a positive, proactive step
> Carsten.

Frankly, I think many of us know how open soruce works. We've been doing
it for 20+ years, since long before it was called open source. It's not
lack of understanding - it's the intersection between what you want to
do and what you can do. That's where the "big reveal" for the first
version comes from - and especially around art work and look and feel
we'll see quite a bit of the more protective thinking of the companies
in the future (as we have seen in Maemo and Moblin in the past).

But once the first relaese with the full UX is out I think this will
also turn into a much more natural way of working. Yes, I bet there will
continue to be "big reveals" around new iterations of the look and
feel. But the project itself will run as a community project. That's our
goal and that's what we are measuring our success against.
Well that and the quality and success of the resulting OS.

> >> For the near
> >> future until UX'es are out, the big reveal mentality will have to stay
> >> - and I respect that. I want a blazing start on a good future in this
> >> project with big fanfare. It's the future I'm worried about.
> > 
> > It's the credibility of MeeGo as an open project that I'm worried about.
> 
> It has all the credibility it needs. It has credibility inside the
> walled garden that is corporate Linux. MeeGo is not meant for ordinary
> hackers - they are encouraged to work and commit upstream. 

That part is definitely true - whenever we have an upstream, we
absolutely encourage people to work there and contribute there.

> MeeGo is a
> 'curated computing environment' where the file system is the best of
> breed but the only one and MeeGo will not listen to the debate. 

Here I disagree. MeeGo as an open source project will of course listen
to the debate. On the filesystem simply nothing new came to light and
the maintainer decided to stay the course. On other topics I expect this
will be different - as you guys certainly have a ton of expertize that
will help us create a better OS. If we weren't plannign to listen then
why bother with an open source OS?

> Its a lot like Flash on the iPhone - ain't gonna happen.

Uhh. Now I'm beginning to take offense. 

> If MeeGo alienates
> a few smart hackers . . . well, they can always contribute to
> Debian. What MeeGo wants is the appearance of meritocratic openness
> while retaining control over the infrastructure. Nokia did this with
> Maemo, and it caused certain problems, and it looks like MeeGo will go
> down this path too.

I'll tell you honestly that my (and our) goals are much much higher than
that. The "appearance of meritocratic openness" is not what we are
aiming for. We are aiming for the best client OS out there. And the only
way to get there is to harness what companies are good at doing (e.g., a
consistent well designed look and feel) and what the community is good
at doing (finding the best solutions to a ton of technical problems,
dealing with the endless combinations of hardware and the weird bugs
that are triggered by that, innovating on top of what exists today,
finding interesting new use cases and basically making the OS better).

Yes, the maintainers will continue to have some control - that's
necessary to create a resulting OS that actually works. And I think the
sooner we have community maintainers the better. Until then, speak up,
argue with us, help us to live up to these expectations.

> >> Finally, my questions to the two members of the TSG are:
> > [snip]
> >> Thanks in advance for your answers.
> > I too look forward to the response - nb maybe you should cc/mail them to 
> > ensure
> > they don't miss such an important discussion.
> 
> They're powerless! Quim Gil is doing his level best to make sure MeeGo
> is as open as possible, but what leverage does he have? He can
> persuade people to contribute, but he can't force them. And once we
> have moved past deciding the forums and other bike shedding, we are
> back to actual work on the actual distro. (And it is a distro, no
> matter how close to upstream they claim it is. Debian only ships
> _pristine_ upstream sources, so you don't get closer to upstream than
> that.)

I don't quite understand who in your mind is "powerless". But returning
to work on the OS is a great idea - the "big reveal" will happen very
soon (don't have a hard date, but I think we're very close) and with
that I think the floodgates will open. And frankly, I can't wait for
that to happen. 

> MeeGo's real audience is industry. It is a standardized. GNU / Linux
> API with key components supported by large multinationals and a
> compliance path so that other large multinationals can be confident
> that their code will be supported and gratuitous API changes will not
> occur. It is a commercial Linux distro, not an open source project.

It's trying to be both. It's trying to be a great platform to
target. But it will do so by working in the open, with the community.

> Maybe that is good thing?

I think it is.

/D

-- 
Dirk Hohndel
Open Source Technology Center
Intel Corporation
_______________________________________________
MeeGo-dev mailing list
[email protected]
http://lists.meego.com/listinfo/meego-dev

Reply via email to