incubator-general mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Jukka Zitting" <>
Subject Re: Graduating a part of an incubating project into an existing TLP
Date Tue, 12 Sep 2006 07:47:19 GMT

On 9/12/06, Niclas Hedhman <> wrote:
> On Tuesday 12 September 2006 05:52, Jukka Zitting wrote:
> > Would this work in terms of the Incubator policies? Do we need some
> > other steps along the way, or can we streamline the process somehow?
> No, this is not what incubation is about. You have two main goals to reach
> before talking graduation;
>  1. Clear out any IP issues.
>  2. Show that the commnuity is healthy.
> Your suggestion seems to indicate that you are looking at the 'stability'
> or 'health' of the codebase, which has nothing to do with the Incubator.

Not really, the point of the exercise is that the Graffito project
(whose goal is to build a portlet-based CMS framework) is perhaps not
the best community for building a generic object-content mapping tool
based on the JCR API. There are a number of Jackrabbit committers and
contributors who are interested in participating, but don't really
care about the full Graffito framework. In fact it even seems to me
that the recent interest from within the Jackrabbit community is
threatening to shadow the real goal of the Graffito project, see for
example the distribution of discussion topics on the Graffito mailing
list within the past two months.

My point about a "clear and working baseline" release was about
simplifying the transition rather than being the primary goal.

> I can also read in that you might be talking about "donating" parts of the
> codebase to Jackrabbit. Again, I think the best way to proceed with that is
> to aim high for graduation of Graffito and then worry about the "svn mv"
> later.
> Set the bar high, aim for graduation... ;o)

The problem I see with this approach is that even though it seems that
we could fairly soon demonstrate a healthy community for the JCR
mapping tool, especially given the recent influx of Jackrabbit
contributors, it might take quite a while to do that for the full
Graffito project. If you filter out the JCR mapping commits and
mailing list discussions from the past six months, you're left with
little ongoing substance on the Graffito project. In fact I feel that
it would be easier for the Graffito community to focus and get back on
track for graduation if the JCR mapping tool was excluded from the
scope of the project.

One option would be to split the Graffito project into two incubating
projects, one for the portlet CMS framework and one for the JCR
mapping tool, but the process and infrastructure overhead for that
seems pretty high.


Jukka Zitting

Yukatan - -
Software craftsmanship, JCR consulting, and Java development

To unsubscribe, e-mail:
For additional commands, e-mail:

View raw message