incubator-general mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Stian Soiland-Reyes <>
Subject Re: [DISCUSS] Commons RDF to join the Apache Incubator
Date Tue, 10 Feb 2015 23:14:23 GMT
On 10 February 2015 at 20:31, Marvin Humphrey <> wrote:

> I also think it would be OK for the project to decide it wants to become a
> TLP.  Whether the project joins Commons or becomes its own TLP won't impact
> the number of people qualified to work on it.  Some Apache TLPs are
> effectively in maintenance mode and have very low activity, but still have PMC
> members willing to answer user questions, make security releases and file
> "still here" quarterly reports.  That seems like a legitimate aspiration for
> this project.

Right - I think it would be good to leave as a decision to be made by
the PPMC when we get closer to graduation. One problem with TLP is
that we would likely need a different name ;-)

> A potential Jena destination also seems as though it would have certain
> advantages, though my naive speculation is that it might be sub-optimal in
> terms of providing neutral territory for negotiating a common API for Jena and
> Sesame.

Right, the idea is for this to get a common, neutral ground.

Although several of the existing implementations, including Jena and
Clerezza, already have the 'abstract' sense of this API, but Commons
RDF API aims to be common across those, and not to "pick sides".

So we have agreed that the new API has to live outside the bigger frameworks.

> In any case it seems likely that if the project achieves its design goal,
> there will be people willing to work on it as long as both Jena and Sesame
> remain viable.  That makes it different from other potential "maintenance
> mode" TLPs which are in danger of stagnation because they cannot renew their
> communities.

I imagine the API would not evolve much once it stabilizes - there's
only that much maintenance you can do on an Java interface :).

But there are additional things we are trying out, like a simple
reference implementation, compliance tests, and event notifications.

Some general utility functions could also evolve later (e.g. "Copy
complete graph from implementation A to implementation B") - but
during incubation we would want to focus on the core interfaces and
integration with the existing implementations.

As our community evolves, I guess documentation would also play a role
- the API aims to be common across the RDF implementations - therefore
users of the API could be considered a single community (compare with
say the JAXB API) - so some usage documentation and tutorials could be
appropriate. This would however point out to details in the (Apache
and non-Apache) implementations.

> Those sound like final mailing lists rather than Incubator ones.  I might have
> expected these instead:

My guess is that Sergio just suggested "classic" addresses out of
habit. :-) Personally I would agree with the above.

> Do you expect to keep separate mailing lists after graduation, or will traffic
> be shunted onto existing Commons mailing list like and

Right - part of the community building (if we continue the Commons
route) is to gradually move (the decreasing) traffic to the existing
dev@commons lists. Commons don't want to split the community with new
component-specific lists, which we can understand.

However we felt that to start out with growing the Commons RDF
community on dev@commons would be a bit of a challenge (300 to 1000
messages/month!) - so part of the motivation for incubation is to have
a more separate space within Apache while we flesh out API design and
implementation questions - and then whoever "survives" so to speak
should be able to move along as we join @commons. :-)

> Lots of Apache experience in this group.  Four are PMC members of at least one
> Apache project.  Andy and Reto are ASF Members.  Andy and Sergio are both IPMC
> members.  Stian is a core contributor of the Taverna podling.
> You probably haven't been getting much feedback because there's a lot going on
> in the Incubator right now and everybody figures that with a group like that
> you're in good shape. :)

Agree, Commons RDF is a slightly different proposal - in a way we need
the incubator mainly to mature the API (e.g. fight over method names)
and grow the community, rather than to be a "podling" to learn the
Apache Way and battle with NOTICE files.

As you see below, though - we still need volunteer mentors.

> The Champion's main work is to help formulate the proposal.  That work is
> essentially done -- so it doesn't matter too much who takes that role, now.
> Are Andy and Reto opting out out as a gesture of openness to Sesame?

Sergio has effectively been the Champion for this proposal, but I
guess he's not technically admissible as the Champion needs to be a
Member or Director.

>> === Nominated Mentors ===
>>  * Benedikt Ritter (britter at apache dot org)
>>  * TBD
> Benedikt is a member of the Commons PMC, but he's not a member of the IPMC nor
> an Apache Member -- so although Commons input is important, unfortunately it's
> not a valid nomination.

We pointed this out, we would however still like to keep Benedikt as
an informal mentor and our "Commons" representative - in a way we have
to learn also the "Commons Way".

Voting-wise we have 2 IPMC members already, but obviously it would be
good to have more IPMC folks mentoring so we don't have to 'fish' for
release votes.

> I'd nudge newly elected IPMC member Rob Vesse, but maybe the roster is already
> Jena-heavy?

Good idea. :) And not unknown territory for Rob.

So I agree to be careful of getting Jena-heavy (I'm not quite neutral
either), but it would be great to have Rob involved - I'm sure he
would try to keep his clean fresh mentor hat on  - just like I try to
keep my "API user" hat on in this project. :-)

We were hoping to also get some "RDF neutral" mentors.

Stian Soiland-Reyes

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

View raw message