incubator-general mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Tommaso Teofili <>
Subject Re: Clerezza status (Was: [Incubator Wiki] Update of "August2012" by BertrandDelacretaz)
Date Thu, 09 Aug 2012 07:58:56 GMT
Hi Jukka,

2012/8/9 Jukka Zitting <>

> Hi,
> On Mon, Aug 6, 2012 at 11:26 AM, Apache Wiki <> wrote:
> > +   As in our last report in May, we believe Clerezza should graduate
> soon, but
> > +   unfortunately that hasn't happened yet. Activity is currently fairly
> low,
> > +   and it looks like Clerezza might remain a small/low activity
> project, but
> > +   the PPMC is functional, has done releases and invited additional
> committers
> > +   so there's no need to stay in the Incubator any longer once a plan
> to attempt
> > +   to grow the community is in place.
> Do you have an idea what happened around a year ago when dev@ activity
> dropped from the hundreds it was for a long time to the dozens where
> it's mostly stayed since then? Alarmingly the low mark seems to have
> been last month when only a single non-automated post was sent to
> dev@.

As far as I can understand that mostly depends on some of the most active
salaried developers got a new job thus they weren't able to keep up the
same "contribution rate" to the project.
Also, if I remember correctly, some discussions about the Clerezza security
module were quite "hot" and that may have played a role in this, but
actually I don't think that's the main reason.

> I recall Clerezza having release trouble due to complex/unreleased
> dependencies for a long time. Could that have contributed to the loss
> of momentum? I think it would be useful to somehow capture experience
> like this, perhaps ultimately for use by ComDev in something like a
> "How to maintain community momentum?" guide.

While I think that guide would be very useful (even if it'd quite hard to
grasp the very deep reasons / solutions) I think the unreleased
dependencies problem was resolved in not so long time.

I think the main reason for the low activity lately is related to the
previous point about salaried devs and to the fact that the most commonly
used pieces of Clerezza (e.g. the RDF API in Apache Stanbol) look quite ok
and stable requiring a minimal development effort.
At the same time there are a number of improvements that the PPMC is aware
of and that could be done.
Another hurdle to newcomers is in my opinion the huge number of modules /
pieces Clerezza is made of which may look scaring at first :-)

> Anyway, it sounds like the community has a reasonably good idea on how
> to proceed, so I'm not too worried yet even though Clerezza is already
> getting pretty close to its three-year mark at the Incubator. Though
> I'd really love to see Clerezza showing notable improvement or even
> graduating before that milestone is reached.

sure, +1.

> If the efforts to grow or reactivate the community fail, would it be a
> good idea to seek to join forces with some related projects like
> Stanbol, Any23 or UIMA? Or do you feel that there are still enough
> active people to allow the project to function as a standalone TLP
> (able to reach 3 PMC votes for releases, etc.)?

I've not a strong opinion at the moment, I think Clerezza could be a nice
TLP but at the same time I wonder if merging efforts with Apache Stanbol
could be a better idea.
I see lots of discussions about the Clerezza API happening on
stanbol-dev@(since it's massively using it), and the two PPMCs share a
number of
members, however there could be good points against that merge (first of
all: different high level scopes).
Therefore my very early opinion is that Clerezza should go for TLP but
after having cleaned its structure based on the most commonly used modules,
tweak the website / docs in order to focus on providing users / devs with a
smaller set of outstanding features.

My 2 cents,

> BR,
> Jukka Zitting
> ---------------------------------------------------------------------
> To unsubscribe, e-mail:
> For additional commands, e-mail:

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