Hi
I try to summarize. If I'm wrong pleas correct me.
The plan (at least one of the possible scenarios) for the Karaf 3.x line
of SMX was to completely remove the code base of SMX and start again
with a new custom Karaf 3.x distribution which only assembles the
Camel, ActiveMQ, CXF features and features provided by Karaf Enterprise
Features subproject. Actual SMX 5 features like Activiti integration
could be moved into the Karaf Enterprise features too. It would give us
following advantages
* the features in Karaf Enterprise Features subproject could have a
separate lifecycle, could be easy installable on vanilla Karaf and
could by assembled by the SMX distribution
* SMX could be easily upgradeable to the newer Karaf version, because
we can create a new custom Karaf distribution and add the necessary
features.
There is yet another problem described some mails ago. We are not 100%
sure Camel, ActiveMX and CXF work on Karaf 3.x with no problems. Some
people have reported problems with usage of Camel or ActiveMQ on Karaf
3.x. Karaf 3.x has also still some problems which must be solved (or
will be solved by 3.0.1)
Because SMX 5 is almost stable, it would be wrong idea to remove its
code base and start SMX5 from the scratch. It would delay the SMX5
release. We can fix the SMX5 issues and release it based on Karaf 2.3.x.
We have time to start a new code base for SMX 6 and build it from the
beginning. We have time to solve the eventual problems with usage of the
esb components on Karaf 3.x.
We doesn't need to wait some next years to release SMX6 after SMX5 will
be released. When SMX6 is ready it could be released some months after
release of SMX5
Regards
Krzysztof
On 15.02.2014 02:16, Raúl Kripalani wrote:
Oops, didn't read JB's message before posting my previous one.
Why does SMX have to do a major release to upgrade from Karaf 2.2.x to Karaf
2.3.x, when within the Karaf project this is simply a minor version upgrade?
I would expect it to be a minor version on our side too.
Sure, I don't want to delay SMX5, but I don't want it to be a disappointing
release either. We've gone 5 years without a major release (SMX 4.0 was
released in 2009), and SMX5's gotta be big to avoid losing the confidence of
users in (1) the roadmap of the project and the (2) technical consistency of
the SMX community.
My 2 cent.
Regards
Raúl Kripalani.
--
View this message in context:
http://servicemix.396122.n5.nabble.com/Plan-for-ServiceMix-5-tp5719013p5719067.html
Sent from the ServiceMix - Dev mailing list archive at Nabble.com.
--
Krzysztof Sobkowiak
JEE & OSS Architect | Technical Architect @ Capgemini
Capgemini <http://www.pl.capgemini.com/> | Software Solutions Center
<http://www.pl.capgemini-sdm.com/> | Wroclaw
e-mail: [email protected] <mailto:[email protected]> |
Twitter: @KSobkowiak
