Alfred Nathaniel wrote:
On Wed, 2008-08-06 at 13:19 +0200, Reinhard Pötz wrote:
Following the result of our recent discussion about the future of
Corona, I propose Corona to become Cocoon 3.
This means that any reference on Corona in source files, package names,
artifact ids, group ids or anywhere else will be dropped and the
standard Cocoon namespace org.apache.cocoon will be used.
This majority vote stays open for 72 hours.
Please cast your votes.
Here is my +1
-1
I think it is much too early to proclaim a tiny blossom like Corona to
be the heir to the huge thicket called Cocoon. It gives the wrong
signal to potential new users and will make them shy away.
They will read it as: "Oh, they are now working on C3.0. So C2.2 will
be legacy by the time my project is finished. I may be forced to
migrate to 3.0 with lots of incompatibilities. Better I use some other
framework for now.
That doesn't make sense. Then this user would have to migrate from the
'other framework' sometime which is most probably more difficult.
I'll have another look when C3.1 is out."
At least that was my personal reaction when in 1999 I first came across
Cocoon. I never bothered with C1.7 because C2.0 was already announced
as being a complete rewrite. Luckily, I passed by a second time in 2002
when C2.1 was in beta state.
Evolution instead of revolution is the key to success here.
C2.2 almost killed us because it was very bold and then took very long
to get out due to the feature creep during the long time it took to get
out. Porting stuff forward and backward between C2.1 and C2.2 did and
does cost a lot of resources. I would not want to throw in there yet
another branch.
There is no need to port things between 2.x and Corona - there is only a
very minimal overlap.
Before considering C3.0 we should have finished the C2.1 to C2.2
transition period. And that is not achieved by simply declaring the
C2.1 branch to be closed. For that I would like to hear more success
stories where people actually migrated non-trivial apps from C2.1 to
C2.2.
sure, I'd like to hear them too.
I don't want to stand in the way of progress here. Please carry on with
Corona and stay within the Cocoon context but just don't call to
Cocoon-x.y.
After 25 days of discussion this was the best solution we found. People
were very unhappy with the use of any codename. And meanwhile I think we
are all tired of the name finding game.
Cocoon 3 will be announced as alpha software. We will add warning
messages to all release artifacts and on the homepage that the code is
experimental and contracts can change from patch releases. We will also
state clearly that the focus of Cocoon 3 is much smaller (small pipeline
API & RESTful webservices) and that, thanks to the servlet-service
framework, it can be run very easily in parallel with Cocoon 2.2
Wasn't the original motivation for Corona to have a
programmable pipeline container which can be used independently of
Cocoon?
The original motivation was that Cocoon 2.x code is one of the most
difficult pieces of software that I've ever seen. We tried to refactor
it (see 'Micro-Cocoon' in the whiteboard) but found out that this is
everything else than simple. While doing this I wondered wow many people
do really understand how the environment handling exactly works and can
do changes without a long trial and error period?
Maybe stupid question: Why can't it be a set of experimental blocks in
trunk which may lateron replace the current sitemap processor?
It's not only the sitemap processor. Corona also has different contracts
at pipeline and pipeline component level.
--
Reinhard Pötz Managing Director, {Indoqa} GmbH
http://www.indoqa.com/en/people/reinhard.poetz/
Member of the Apache Software Foundation
Apache Cocoon Committer, PMC member [EMAIL PROTECTED]
________________________________________________________________________
