Re: [rt-pmc] Juno Retrospective

[Jesse McConnell <jesse.mcconnell@xxxxxxxxx>]

it is deeper then that with regards to Runtime in general

For almost as long as Jetty has been part of eclipse there has been a
bit fat '?' over RT on what it represents and why it even exists.
Jeff McAffer was attempting to address that somewhat with his attempts
at making a product out of the runtime, where there was some
compelling story for using the different diverse systems of the
runtime project together in some way.  Jetty participated in that with
the WebStarter kits that there was a flurry of activity behind for a
while there...seems like couple years back at this point.

Personally I see it as a systemic issue within the eclipse foundation
and its focus on a tricked out editor.  On one hand the eclipse
foundation is pushing to be a more general entity that accepts open
source projects that care about IP and want to have their whole
existence validated and viewed through that spectrum.  This is a nobel
goal and one of the reasons that jetty came to eclipse.  However on
the other hand it seems the whole focus of the actual organization is
to produce this tricked out editor as a core which strategic members
can build their products on.  Enter the release train...the core
mechanism that realizes this goal.  Now I am not saying this is a bad
thing nor that is is not needed....it is just not the primary end goal
for a project like jetty.

Now, there has been some lip service spent on eclipse (equinox) really
being a software stack to build applications on, which jetty supports
and would certainly like to be a part of, but this so bound up in the
eclipse editor as the 'shining' example that it is hard to separate
the two.  Jeff was working towards that goal for a while but it was
obviously a side goal and not a mission of the organization at large,
which can be evidenced by the effort stalling and basically
disappearing once he left.

Which brings us back to the RT group...which briefly toyed with the
idea of producing a software stack that could do much of what
something like Ruby on Rails could do, only in an osgi fashion and
leveraging the projects within RT and perhaps a handful of other
projects within eclipse.  That never really materialized, so the RT
group now basically motors along on auto-pilot.

<thomas's mail showed up>

I'll paste in and tweak my response to Wayne when he asked me
privately this question.

- there is no _need_ for jetty to do it, we have no UI components that
are remotely critical to or of interest to core jetty developers or
users, we had a wtp component that we worked on some but really never
got much interest in

- the process is bothersome, it takes time to maintain the files (and
go re-figure out what we needed to do when the time came), build
metadata, p2
repositories, release docuware, etc

- we don't seem to fit the traditional eclipse project structure, ie
we don't live and breathe around the release of the eclipse platform
and we get no real benefit from committing resources to making it
happen, its ultimately just a distraction, when eclipse releases there
is no evidence that we have participated or had an impact on it

- the end result of our dropping off the release train is only a net
positive for us, we get the occasional hour or three back while
equinox and anyone else updates some line in their build to point to a
p2 repository...problem solved


cheers,
jesse

p.s. this is probably a good time to get this conversation going
again, we have had it off and on in the RT calls but its been a while
(though I have missed the last couple by a few minutes :( )


--
jesse mcconnell
jesse.mcconnell@xxxxxxxxx


On Wed, Aug 22, 2012 at 9:09 AM, Mike Milinkovich
<mike.milinkovich@xxxxxxxxxxx> wrote:
> Just a suggestion - feel free to ignore.
>
>
>
> It seems to me that we have a general issue with EclipseRT and the release
> train. More specifically, I think there is a sense that the release train is
> of less value to the RT projects. Or that the process is more burdensome for
> them. Jetty's recent announcement that they're not planning to participate
> in Kepler is a concrete example of this.
>
>
>
> Perhaps a general conversation about what makes the release train more
> difficult or of less value to the RT projects would be helpful? What could
> be done to change that for the RT community?
>
>
>
> Of course, it's possible that I am completely out to lunch on this
> perception. It wouldn't be the first time :)
>
>
>
>
>
> From: rt-pmc-bounces@xxxxxxxxxxx [mailto:rt-pmc-bounces@xxxxxxxxxxx] On
> Behalf Of Ian Bull
> Sent: August-22-12 9:49 AM
> To: Runtime Project PMC mailing list
> Subject: [rt-pmc] Juno Retrospective
>
>
>
> Hi everyone,
>
>
>
> The planning council is interested in feedback on the Juno Release.  If you
> have anything (good or bad) that you would like to add to the Juno
> Retrospective, please follow up here and I'll pass the comments along to the
> planning council.
>
>
>
> Also, the Kepler schedule has been finalized. You can see it here [1].
>
>
>
> [1] http://wiki.eclipse.org/Kepler/Simultaneous_Release_Plan#Schedule
>
>
>
> Cheers,
>
> Ian
>
>
>
> --
> R. Ian Bull | EclipseSource Victoria | +1 250 477 7484
> http://eclipsesource.com | http://twitter.com/eclipsesource
>
>
> _______________________________________________
> rt-pmc mailing list
> rt-pmc@xxxxxxxxxxx
> https://dev.eclipse.org/mailman/listinfo/rt-pmc
>