Santiago:
At the first Expert Group face-to-face meeting, JSR-286 spec lead Stefan
Hepper told us:
1. IBM was contracting with the Univ of Jena to help produce the JSR-286
Reference Implementation.
2. He wants Apache Pluto to host the RI code
3. He wants to base the JSR-286 RI on Apache Pluto 1.1
There are seven Apache Portals committers, including Stefan, who are
members of the Expert Group, so many of us should have known that this
was coming. Another constraint is the Java Community Process (JCP) rules
that state that EG information should not be disclosed before the are
formally released in document form. The first early draft was released a
couple of weeks ago, which is why this discussion is happening now.
Stefan has been away on vacation for a couple of weeks, so hopefully he
will chime in when he gets back.
A couple of weeks ago, Torsten Dettborn of the Univ of Jena contacted
David and I about beginning the process of incorporating their work into
Pluto's SVN repository. Torsten's command of English is limited, which
may be the reason he has not chimed in yet on this thread. I have
invited Carsten Zeigler to participate in talks with Torsten to
facilitate communication between Apache and the Jena group. Although
Ulrich Kuster is a Pluto committer, Torsten told me that he has not been
involved with them since Jan-Feb.
I for one am very happy that we have a paid group that can dedicate
their time to building Pluto 2.0. David and I, and I understand, the
rest of the Pluto contributors, are volunteers. That does not diminish
our commitment to the project, but we have to fit it in when we can. We
all want to contribute to Pluto 2.0, but it is good that we have the
Univ of Jena group to help move the project forward.
Please excuse my presumption that the Pluto 2.0 branch creation process
was not any different from what we did with the Pluto 1.0.2 and Pluto
1.1 branch. I understand their are some historical reasons to proceed
cautiously, so please have the Portals PMC vote on this matter as soon
as possible. My only suggestion is that keeping track of the Jena
contributions is better done in Jira rather than on pluto-dev.
Nevertheless, I am willing to abide by what ever the PMC decides.
TIA
/Craig
Santiago Gala wrote:
No, my reason for the -1 is that "IBM has contracted with a group at
the University of Jena in Germany to help create the reference
implementation of JSR-286 to be hosted at Apache. ", as cdoremus put
it. This means that the code result is a work for hire, and thus would
need to go through corporate CLA, vetting, etc. in the incubator.
OTOH, given the secrecy which has involved the whole JSR process (a
committer of the portals project was the chair, but we actually got
the first notice through the JCP website, etc.) I'm afraid we will be
again in a world of obscure changes, which go straight against the
Apache Community spirit, and is the kind of things that have been
slowly killing this project in the last couple of years, from the
community point of view.
Now, if the components of such team have the whole communications on
pluto-dev, and they send patches (except for Ulrich, which is a
committer) that get integrated until they earn committership, the
process would not need to pass through the incubator.
This is not an opinion, I think this is hard policy in the ASF. Apache
is about communities, not about code.
I (this is an opinion) don't care a dime about Reference
Implementations unless they are used and tested. This is the case now
with pluto 1.0, but it was not the case when it got donated.
I am afraid that we are facing a code dump that will take a lot of
time to clean up. Those are my concerns.
Regards
Santiago
On 8/25/06, Ralph Goers <[EMAIL PROTECTED]> wrote:
Santiago Gala wrote:
> I'm -1 about the branch.
>
Is your reason for the -1 because you feel that the JSR 286 work should
happen on trunk (assuming it was Pluto committers doing the work)?
Wouldn't you need to create a branch for 1.1 before you could start
that?
Ralph
