[gentoo-dev] Re: [gentoo-core] [RFC] New metastructure proposal

[Alon Bar-Lev]

On 4/10/07, Alexandre Buisse <nattfodd@g.o> wrote:
> Hi everyone,
>
> as everyone probably noticed, there is a current atmosphere of sinking ship,
> with quite a lot of people leaving and many agreeing that gentoo is no fun
> working on anymore. Before it's too late, I'd like to propose a big reformation
> that would help solve some of the issues we are currently having and,
> hopefully, bring back some of the fun we all had developing and using this
> distribution.

Thank you for bringing this up. I think the discussion is very important.

It is first time I response to a "political" issue here... So forgive
me if I am totally wrong.

I don't think this is the reason people are leaving.
I think people are leaving because a lack of direction.
I am not aware of any goal Gentoo distribution wish to acquire. For
example: Do we wish to use a mainstream distribution? Do we aim to a
specific user community? Or Do we develop distribution for our use?

If you wish to be (And I think we should be) mainstream distribution,
we should derive targets, such as QA level, response times and
content.

Being more modular is one technical feature to achieve better
stability. But we should discuss the basics first.

I hear a lot that open source project with unpaid developers cannot be
committed to deadlines or requirements from its developers, but I
disagree. There can be an open source project with high quality
products and dead lines, if these properly defined.

I am quite new here, but it seems like there are few groups here, from
a group that interested in anarchy to a group that interested in
formal hierarchy.

I must disclose that in my view whenever a large group of people are
doing something together, some kind of hierarchy must be in place. And
I am not talking about current council, it seems that current council
does not LEAD Gentoo anywhere.

I read that sometime in history there was an effort to impose
structural format on the community, but then Daniel Robbins left?

If we wish to be a major distribution, we must grow. If we to grow we
must organize our-self better, and work toward a common goal. Common
goal forces decision making. Decision making forces leadership.
Leadership forces vision.

Is there any vision?

Now, for your idea.
When I written something similar in the past, someone told me that it
was already suggested... I don't know why it wasn't accepted.

I think too that ebuilds should exists in several overlays. At least:
Stage3 (current)
Base (baselayout, networking, boot loaders etc)
<herd>-tear-<N>

Each tear should have commitments for:
1. Time from upstream release package until it in overlay.
2. Time from security issue until it in stable.
3. Time from stable request until make stable per each platform.
4. Time for addressing bug, time for solving bug.
5. Keep last N stable version (major, minor).
I guess you got the idea.

For example, for crypto there can be several overlays, tear-1 overlay
with core packages, tear-2 with misc packages, tear-3 with supported
at free-time bases.
Each tear has its own measures.

tear-1 desktop cannot be dependent on tear-2 crypto, the desktop herd
need to ask crypto herd to move the package into tear-1 before
dependency is added.

I totally agree that each user may ask for package of specific
overlay, but I think that this should not be specified in the build,
but by the user at /etc/portage.

For example, I decide that dev-libs/gtk+ should be from overlay X the
portage should take it from there.

But I believe we should first discuss the community goals, then derive
a technical solution.

Best Regards,
Alon Bar-Lev.
-- 
gentoo-dev@g.o mailing list