db-general mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Geir Magnusson Jr." <ge...@adeptra.com>
Subject Status
Date Thu, 09 May 2002 12:33:44 GMT
Good morning.  Sorry about the delay - things got busy and was hoping more
people showed up.

I promised someone that I would put the a disclaimer at the top :

"The name 'db' is only a working name - we expect a much wider scope - so
don't get used to it."

I guess we have a bunch of people here now, so this starting status message
won't be wasted.

For those that don't know the history, we started this discussion group to
talk about the issues related to starting a new Apache subproject related to
data access, persistence, etc.  The original motivation was the proposal
that OJB move from SourceForge to Jakarta and how it would be a good 'seed'
to start a new community at Apache.

First, we are in *no* rush to get this done.  Thomas Mahler, the OJB lead,
is on vacation this week, and since this proposal is centered around OJB,
out of courtesy, I think we should wait for him to make sure that he's
comfortable with all that is going on.  He's new to our community, and he
might not want to object to anything if it looks like we 'already decided'.

Second, Peter had the great idea of waiting to announce this around the time
of a major trade events, as to piggy-back on the press and interest.  I had
the great idea of announcing this *between* major trade events, to avoid
getting lost in the press and interest in the trade event.  :)  We need to
sort this one out.

We currently have in mind 4 projects that will be designated as top-level
projects with which to start :

1) OJB - ObjectRelationalBridge, currently at sourceforge
2) Poolman - popular database connection pooling package, currently at
3) Torque - persistence layer from the Turbine project
4) commons - modeled after Jakarta commons w/ sandbox

Of course, Torque's developers and the Turbine community will need to make
the decision if they want to move.I think they already did that, but not

We also need to decide if Poolman should be there.  I think it should - it
has a huge user community.  Brief history : I am currently the 'keeper' of
poolman - it was turned over to me recently by Sean Neville, the author, as
professional and life commitments prevented him from adequately supporting
it.  It is currently LGPL, and has a developer base of 1, Sean.   He did
this to ensure freedom of action if he wanted to license the codebase in
another way - he didn't want to have to hunt down all the contributors.  I
have been in a holding pattern over the last month due to time constraints
and figuring out how to re-license the code.  I think bringing to 'db' would
be great - I have several other Jakarta committers ready and willing to
participate, so there will be the required developer base of 3, at least.  I
would offer this as a 'conditional' top-level project - if it doesn't really
get moving after 6 months, we drop it.

I think that it's really important that we keep the top level projects of
high quality and clear scope.  That's why the commons is there - to be both
the container for 'smaller' things (components and utilities) as well as an
incubator via its sandbox.

Additionally, there are several other project ideas floating around :

1) Xindice : from the xml.jakarta.org - I just extended an invitation for
the community to come join and discuss if they want to move over

2) SqlClass : http://www.quiotix.com/opensource/sqlclass/ - written by brian
goetz, this may not be top-level project status, but would be a great seed
for the commons.

3) Peter Donald has a laundry list of acquaintances and un-indicted
co-conspirators that he will contact.

I asked Dirk-Wilhelm van Gulik, an Apache board member, about the idea.  He
had some questions, which I answered.  Out of courtesy, I will get his
permission to repost to this list.  He feels very positively and has copied
the exchange to the board as an FYI - they will of course take no action
until we approach them, but this enables them to watch what we are doing
(hopefully), and provide advice or warnings.

Moving forward :

As first steps, while waiting to have more people join as they find out
about us (Xindice, for example and peter's contacts), as well as waiting for
Thomas to come back from holiday and engage in the discussion, I think we
should figure out what else we wish to invite to the project.  Doing this
will help us find the boundaries of our scope.  There are no language
restrictions (it doesn't have to be in Java, for example) nor are there
specific technology restrictions ("must be RDBMS").

I think we have a solid list of near-definite candidates (the 4 above), so
putting together  a list of possibilities will be great.

I am going to keep a document up-do-date with what we discuss and decide and
repost every few days, so lets start throwing the ideas out and discussing.


Geir Magnusson Jr.
Research & Development, Adeptra Inc.

View raw message