+ to vote 2.0.1 release in 3-4 weeks from today, and limit release to
bug fixes only, no new features.
+1 to do release in every 8-10 weeks timeframe, it is very important to
maintain regular release as the community is growing.
As for those possible new stuff for 2.1 release, we have to be realistic
about timeframe (8-10 weeks) and set some high priority features to work
on. I deem JAX-WS 2.1 and JAXB 2.1, Better REST service support -
WADL/JSR 311, SCA integration, and build refactoring. I am sure there
will be a lot of bug fixes also.
For things beyond 2.1 release, I also see following possible agendas:
- better dynamic language support (refactoring JavaScript support, add
more dynamic language support such as JRuby, Groovy)
- JSON support
- Some more transport support such as SMTP?
Cheers,
Bo
Dan Diephouse wrote:
Yes, its that time - we get to start thinking about post 2.0 releases!
(and
2.0 isn't even fully out yet - but its close ;-))
First: 2.0.1. I think its important to follow up with bug fix releases
quite
often in general, but especially after the first release. We'll see a lot
more people start to use CXF once 2.0 is out I believe. And I'm sure
they'll
find all sorts of bugs. So I'd like to propose that we follow up with
2.0.1within 3-4 weeks of
2.0 (i.e. vote on it 3-4 weeks from today).
The question of 2.1 is a bit more complex. I see several things on the
possible agenda for 2.1:
- JAX-WS 2.1 & JAXB 2.1
- Refactoring the build better to help with the tool dependency issues
- WS-SX (Trust, SecureConversation, SecurityPolicy)
- WS RM 1.1?
- Better REST service support - WADL, JSR 311 early draft?
- WSDL 2? (I see this one as increasingly irrelevant as there has been
almost no take up of WSDL 2, and I think it'd be a distraction for us
right
now)
- SCA integration? (coincidentally, I noticed the Fabric3 team has
started
CXF integration)
- Others? I'm sure there are many other important issues out there...
I see two philosophies by which we can release 2.1
1. The fixed time philosophy - i.e. We do a release every 8-10 weeks
2. The fixed feature set philosophy - i.e. we do a release only once
these
features are ready
I'm a much bigger fan of philosophy #1 as I think it serves users much
better. But I am open to to either provided we make frequent bug fix
releases for 2.0. The one constraint that I see on 2.1 is that we will
need
to do JAX-WS 2.1 TCK testing which will take how ever long it takes.
So even
if we go by the "fixed time" philosophy, our window can't be smaller than
the time it takes to implement those APIs and do TCK testing.
Thoughts?
