SV: A Merlin Top Level Project

Hi,
I think that is definatiely a good idea. in fact, I have been witing for the
proposal some time since Merlin is really not avalon as I understand it. It
is much more a project like Pico, Excalibur etc, focusing on implementation
of a component infrastructure instead of specifying it as avalon is known
for. Since there are other projects claiming to implement the avalon
contract (which is not being clear since there is no consensus on what
constitutes this contract), even Merlin as the "Avalon container" sounds a
bit like hijacking the framework, which I think is why the flames arise.

I love Merlin and would like to see it a TLP!

/peter

-----Ursprungligt meddelande-----
Från: Niclas Hedhman [mailto:[email protected]]
Skickat: den 22 juli 2004 03:43
Till: Avalon Users
Ämne: A Merlin Top Level Project 



Hi everyone.

This mail is directed to EVERY USER of Merlin, if you just starting out,
half 
way ahead and experienced, _please_ read and respond.

== Abstract ==
Avalon has yet again been under fire for making too much changes to the
Avalon 
Legacy (read Framework, site and documentation), what constitutes compatible

changes and how these changes are perceived by the Avalon-dependees, such as

Excalibur, Cocoon, James and Loom. The aftermath of this last storm, makes 
_me_ believe that Avalon should NOT be the home of Merlin, and I am lobbying

for the support of a Merlin Top Level Project.

== Purpose ==
Merlin is so much more than what the founders of Avalon anticipated, and the

Merlin committer crew (esp. myself, Steve, Alex, and Timmothy) feels that 
there are an unnecessary restrictions on how we can evolve Merlin, without 
creating fear at the other projects depending on Avalon.

We think it is better for Merlin to be a user of Avalon Framework, like 
Excalibur, Loom, James and Cocoon are, without any special privileges or 
obligations.

The Merlin committers wants to forge ahead with a vision of a serious 
alternative/compliment to J2EE, where component oriented programming finally

gets to its full potential as an efficiency enabler.

We want to send a clear message that Merlin is not a 'project' shuffled into
a 
backyard called Avalon, but that it is a serious _product_, on par with
other 
ambitious products in ASF like Geronimo, Cocoon and Eve. Avalon have lately 
had as many or more web site visitors[1] than high profile projects like 
Struts, Tomcat and Ant. This is a sign that there is a strong interest out 
there, and that we need to channel this interest into a stronger community.

I think that a Merlin Top Level Project is due, the technology and community

is mature, the increased visibility will spark even higher interests, and 
that we can accelerate the development of the platform better without the 
concerns of Avalon legacy.

The Merlin committer crew is also determined to increase the frequency of 
releases, especially on facilities and generic components. Less of our
energy 
will be spent on the 'infamous in-fightings of Avalon', and we can serve the

community better.


== Mitigation of Consequence to Users ==
This will not be a pain-free exercise.
If we are granted "The Apache Merlin project" (http://merlin.apache.org), we

will most probably be required to change the codebase package names.

However, I hope that we will be able to bridge the inconvenience for the 
"Facilities Users", who would have wrong classnames for the facility 
declaration, so that Merlin will explicitly recognize both namespaces for 
some time, and generate a deprecation warning.

Native normal components would work as without change.

Blocks (aggregated components, i.e. block.xml files) should also work
without 
change.

Applications that embedds Merlin is the only major area, where we can't 
provide a smooth transition. I.e. import statements will have to change, but

with modern IDEs (alt. some script files), the work would be fairly straight

forward and quick.


== YOUR action is required ==

I NEED you to respond to this mail. If you like the sound of this idea, just

send a mail to the list saying exactly that. If you think this is not a good

idea, please express that too, to this list. This is not a vote, just a call

for support.

The more support I can rally from the users, the more likely the proposal
will 
be viewed positively by the ASF Board, whio ultimately decides on this.


Cheers
Niclas

[1]
Daily stats (90 days):
http://reverycodes.com/apache/stats/stats.pl?type=projects&day=&month=&year=
&interval=day&periods=90&projects=ant&projects=avalon&projects=struts&projec
ts=tomcat

Monthly stats (6 months):
http://reverycodes.com/apache/stats/stats.pl?type=projects&day=&month=&year=
&interval=month&periods=6&projects=ant&projects=avalon&projects=struts&proje
cts=tomcat


-- 
   +------//-------------------+
  / http://www.bali.ac        /
 / http://niclas.hedhman.org / 
+------//-------------------+


---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]

---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]