incubator-general mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Guillaume Nodet <gno...@gmail.com>
Subject Re: [PROPOSAL] Apache Aries incubator for Enterprise OSGi
Date Fri, 04 Sep 2009 11:51:00 GMT
There are a few things I don't understand well.   I thought the ASF over the
past years was trying to discourage umbrella projects.   I also thought that
overlap between the projects was indeed accepted (we already have multiple
JAX-WS or JAX-RS implementations in various TLPs / podlings).
That was rather clear in my mind, so I must have missed something.

But let's try to be constructive ...

Given the state and objectives of Aries (which means there's not much code
right now), I don' think it's wise to start splitting now between pure OSGi
specs implementations and other extensions, especially because some of the
possible specs mentionned are not yet spec, and Aries would become an input
/ feedback source for creating those specifications.

That said, we need to find a home for this proposal.   It seems Felix might
be a good place, so this basically means incubating Aries inside a Felix as
a subproject (I still think this is the role of the incubator, but let's put
that aside).
It means creating a subproject in Felix in a sandbox opened to all ASF
committers, and having a very low entry barrier for people that are not ASF
committers yet but want to contribute to Aries (as I said, it's difficult to
contribute patches when the code is inexistant at this point).   Another
thing to help creating a community around Aries would be to have the
possibility to have dedicated mailing lists and web site as a lot of TLP do
for their subprojects.  Plus, there are some code donations, so we also need
to take care of the IP clearance.  Binding votes would be those of the Felix
PMC of course.  Then, when the pseudo-incubation is completed, Aries could
either become a real Felix subproject, its own TLP, or be split.

I really don't  see the value in that, as it seems I've just described what
the Incubator has been set up for afaik.  And as any new podling proposal,
any ASF committer is free to join at this point.  So there's really not much
difference between every people from felix joining aries podling.  But I
really don't want to fight for that and I'd personally be ok if the Felix
PMC wants to become an incubator for Aries.


On Fri, Sep 4, 2009 at 11:29, Bertrand Delacretaz <bdelacretaz@apache.org>wrote:

> On Fri, Sep 4, 2009 at 7:50 AM, Niclas Hedhman<niclas@hedhman.org> wrote:
> > On Fri, Sep 4, 2009 at 1:23 PM, Kevan Miller<kevan.miller@gmail.com>
> wrote:
> >
> >>... What would be the
> >> benefit for the Aries community of developing these spec implementations
> at
> >> Felix?
> >
> > Ideally, you have more people taking care of any issues. More
> > importantly, you need to think outside the project and ask "What would
> > be the benefit for ASF...?", and IMHO having a complete spec suite
> > from one place benefits ASF as a whole....
>
> Very much agree with that. Being able to say "OSGi spec
> implementations happen at Felix, and
> customizations/application-specific stuff/extensions etc. usually
> happen in other projects" leads to a clear picture for "customers" of
> Apache OSGi stuff (and BTW the incubation proposal forgot about Sling
> apparently - we might use Aries' stuff as well in there).
>
> I don't like the idea of having part of the OSGi specs scattered
> around several projects - as a general rule. Several modules are now
> at Felix that were initially developed by the communities of other
> projects (ServiceMix, Slng, ACE), and moved to Felix once recognized
> as being "of general interest to OSGi users" - I think that's a very
> good model both for ASF insiders and outsiders.
>
> Several projects in the history of the ASF have granted commit access
> to committers of other projects on parts of their codebase. That
> wouldn't be a new thing. Forrest committers, for example, have commit
> rights on Cocoon, but they're expected to ask before changing stuff in
> there, to make sure Cocoon folks know what's going on.
>
> That works well, IMHO, and maybe Felix could similarly open parts of
> their codebase to other OSGi-related projects, with the same
> convention of "just let us know before making changes in there" and
> "make sure you're following our dev list".
>
> I know such issues can be discussed during incubation, not necessarily
> before the podling is accepted, but the fact that (AFAIK) no contact
> was made with the Felix project before creating the Aries proposal is
> very disturbing - so IMHO we should rather clarify this before
> accepting the podling.
>
> Maybe simply saying "modules that implement OSGi specs, or that are of
> general interest to OSGi users, will be moved to the Felix project, as
> much as possible" in the Aries proposal would help. That should be
> true for any project doing OSGi at Apache anyway, and I think the
> other projects mentioned above are working like that, which is a Good
> Thing,
>
> I'm very happy to see Guillaume as a mentor for Aries, that will
> hopefully help in building bridges between Felix and Aries.
>
> -Bertrand
>
> ---------------------------------------------------------------------
> To unsubscribe, e-mail: general-unsubscribe@incubator.apache.org
> For additional commands, e-mail: general-help@incubator.apache.org
>
>


-- 
Cheers,
Guillaume Nodet
------------------------
Blog: http://gnodet.blogspot.com/
------------------------
Open Source SOA
http://fusesource.com

Mime
  • Unnamed multipart/alternative (inline, None, 0 bytes)
View raw message