Re: [Springframework-developer] Current release plan - why delay package renaming?

[Rod J. <rod...@in...>]

> I'd also like to make up a concrete release plan for the follow-up, be it
0.9.2 or 1.0 beta2, with an approximate timeframe - probably beginning of
September. Obvious candidates are:
>
> - ResultSets from stored procedures;
> - Tiles integration;
> - advanced PropertyEditor support for JSPs;
> - compatibility testing;
> - polished sample apps;
> - very important: documentation and tutorials.

I'd call this 1.0 beta 1.

> I guess that stuff like metadata attributes driving AOP interceptors will
be beyond that timeframe. But we don't need to cram everything into this
follow-up: 1.0 RC1 in early October will be a fine milestone too. It would
generally be good to publish one release per month towards 1.0.

This is still my main priority; it's just that I've been distracted into
various other areas, largely to support actual projects. Anyone who'd like
to help on the metadata side, please let me know. I think this is very
important to the ease of use of our AOP APIs. We need a consistent metadata
solution in which source-level attributes are only one way (although often
the best) of providing metadata. .NET is of course a good example of
source-level metadata. JBoss 4 is also doing some pretty sophisticated
things with metadata hierarchies, although as I understand it it looks more
like text metadata than the object metadata that .NET offers and I think is
important.

I think Attrib4j, Mark Pollack's project, has a part to play here, and am
working with him to try to make Attrib4j easier to use.

> Finally, regarding the scope of 1.0 final: I still vote for delaying JMS
and Web Services support until 1.1, to be able to stabilize and polish the
current features as much as possible for 1.0. The same applies to stuff like
TopLink support or JSR-168 Portlets. We can still decide to focus on certain
features earlier if current needs arise, of course.

Agreed. No one could say that Spring doesn't have enough in it! I think the
timing of getting these things in will depend on how many of us have an
urgent need for them. For example, I may well become far more interested in
TopLink integration and JMS because of work commitments, in which case I'd
definitely be prioritising them, whether or not they'd be identified as a
Spring priority.

Unless for some such reason we end up with the support earlier, I think
Spring 1.1 should have Web Services and JMS as the major new features.
Because these are essentially "bolt-ons" that don't require extensive
changes, we could probably get to 1.1 reasonably quickly from 1.0.

The next major feature on the horizon after that (1.2?) might be dynamic
reconfiguration of bean factories, as we discussed in the list a while back,
although I'm sure our growing user community will generate a lot of our
direction.

TopLink integration is a nice-to-have given our support for Hibernate and
JDO, but I suspect that not that many users are likely to be crying out for
it.

Regards,
Rod