Merlin & OpenIM

[Alexis Agahi <al...@users.sf.net>]

Folks,

I'm starting to study for real OpenIM/Merlin implementation. OpenIM is based 
on Phoenix and rely on Cornerstone for many stuff, but, for fun & new tech 
experimention (Merlin look funky) and also as Phoenix has some "heavyweight" 
approach, I would like to try something new :)

BTW, after a first look of many hours, Merlin is not that easy to mastering. 
Would it be possible to have a burnout-session/beer with one toplevel guru 
floating around, to push an openim release based on Merlin?

Is there a release roadmap for Merlin (in/outside avalon)? As always, as an 
integrator, I'm driven by stable release status to be able to tag the openim 
server 'stable'; this not mandatory, but it's better.

Regards

---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@avalon.apache.org
For additional commands, e-mail: dev-help@avalon.apache.org

Re: Merlin & OpenIM

[Leo Simons <le...@apache.org>]

Stephen McConnell wrote:
> Personally I think Merlin should have had a release here in 
> Avalon about a year ago. Maybe I should just post a vote and take it 
> from there.

you know the deal --

   1) propose a release roadmap
   2) propose a release manager
   3) release alphas (3 or so imo, but that's up to the RM)
   4) release betas (also 3)
   5) propose a final by uploading to personal space on the
      cvs box and calling a vote

no-one has done any of those things, so its not surprising there haven't 
been any releases :D

> One of the things that I see as slowing things down is assumption about 
> what a release means.

enough of that. I remember calling for a vote on loose rules for alpha 
and beta and RC releases. Everyone was in favor of the 
less-bureaucracy-more-just-go-for-it.

I think what has slowed releases down is lack of people with experience 
in generating releases and a willingness to it for merlin.

cheers!

- Leo



---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@avalon.apache.org
For additional commands, e-mail: dev-help@avalon.apache.org

Re: Merlin & OpenIM

[Alexis Agahi <al...@users.sf.net>]

On Tuesday 05 August 2003 02:42, Stephen McConnell wrote:

> >BTW, after a first look of many hours, Merlin is not that easy to
> > mastering.
>
> Lies, lies, damnable lies!

LOL


> Well, actually, it should be drop dead simple but you know how it is,
> users keep getting in the way, comming up with things thay want to do to
> which above and beyond the call of duty. I mean - what can you do?

Agree, it's not easy. Especially when you have to deal with many notions 
(distinct from Merlin) such as Maven, IoC, etc.

What about writting an Oreilly Book? AFAIR Oreilly is open on any technoid 
subject that could sale more than x books (having x not so high)
"Avalon Practice with Merlin" :)




> >Would it be possible to have a burnout-session/beer with one toplevel guru
> >floating around, to push an openim release based on Merlin?
>
> Absolutely. Let's do it!

Great!


> As to your question of "inside or outside of Avalon" - that's an
> interesting one. If Merlin stays within Avalon it seems that it will be
> disruptive. But I figure that this is better than a move outside of
> Avalon. Personally I think Merlin should have had a release here in
> Avalon about a year ago. Maybe I should just post a vote and take it
> from there.

My opinion is that we (the developpers outside the inner circle) need official 
frequent stable releases (generally speeking), and Merlin cannot stay in 
sandbox for years. CVS update is cool, but time consuming if you are not 
working on avalon mainstream stuff.
So maybe someone could push a 2.x release branch outside Avalon, and Merlin 
team could try to work with avalon community 'hand in hand' (lol) to converge 
with 3.0 (and if Merlin3 can maintain ascendant compatibity, it should be 
great).


> One of the things that I see as slowing things down is assumption about
> what a release means. Personally I want a Merlin release to be an active
> vibrant rolling beta process with a stable component contract combined
> with a volatile container contract until such time that we feel we have
> cracked the equation. I.e. moving away from:
>
> [a bunch of monks mumbling in monotone]

<snip mode="funny"/>



> Merlin 2.1 is fine as it is.
> That's what you end with using the current
> CVS and Merlin command line. Its ready to go, validated against all of
> the cornerstone components, James, a bunch of my own experiments, and a
> number of external projects (a couple of which have successfully pushed
> Merlin well beyond the things I’ve been playing with).


So let push him out somewhere. If big players here are ok to have an official 
Merlin 2 within avalon, this could be nice, otherwise, join the darkside for 
a while.

And could we keep dithyrambic rhetorics based on Merlin 3.0 and "community 
convergence" in this mailinglist.
No?

> Merlin 3.0 is a process that is starting to emerge - its shaping up to
> be about (a) refactoring of the assembly system, (b) separating out a
> management model we can easily handle things with JMX, (c) improving the
> APIs exposed to container extensions, (d) setting the groundwork for
> distribution and federation, while (d) making sure that we having a good
> time!

Sounds mirific, delirious, wonderful (MDW).

Regards,


---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@avalon.apache.org
For additional commands, e-mail: dev-help@avalon.apache.org

Re: Merlin & OpenIM

[Stephen McConnell <mc...@apache.org>]

Alexis Agahi wrote:

>Folks,
>
>I'm starting to study for real OpenIM/Merlin implementation. OpenIM is based 
>on Phoenix and rely on Cornerstone for many stuff, but, for fun & new tech 
>experimention (Merlin look funky) and also as Phoenix has some "heavyweight" 
>approach, I would like to try something new :)
>
>BTW, after a first look of many hours, Merlin is not that easy to mastering. 
>

Lies, lies, damnable lies!

Well, actually, it should be drop dead simple but you know how it is, 
users keep getting in the way, comming up with things thay want to do to 
which above and beyond the call of duty. I mean - what can you do?

>Would it be possible to have a burnout-session/beer with one toplevel guru 
>floating around, to push an openim release based on Merlin?
>

Absolutely. Let's do it!

>
>Is there a release roadmap for Merlin (in/outside avalon)? 
>

There have been no changes to Merlin 2 in the last couple of months 
other than the synchronization with the moving out of the meta info 
package into sandbox/meta (which I should probably move back under the 
Merlin namespace). In fact all of the stuff I've been doing recently is 
focussed on Merlin 3.0.

As to your question of "inside or outside of Avalon" - that's an 
interesting one. If Merlin stays within Avalon it seems that it will be 
disruptive. But I figure that this is better than a move outside of 
Avalon. Personally I think Merlin should have had a release here in 
Avalon about a year ago. Maybe I should just post a vote and take it 
from there.

One of the things that I see as slowing things down is assumption about 
what a release means. Personally I want a Merlin release to be an active 
vibrant rolling beta process with a stable component contract combined 
with a volatile container contract until such time that we feel we have 
cracked the equation. I.e. moving away from:

[a bunch of monks mumbling in monotone]

!
! "a release is something that we [pause] at avalon [pause]
! must uphold [long pause] for the benefit of [searching pause]
! our users, [excited pause] our past, our present, and our
! future [satisfied pause]"
!

as distinct from:

!
! "just do it"
!

>As always, as an 
>integrator, I'm driven by stable release status to be able to tag the openim 
>server 'stable'; this not mandatory, but it's better.
>

Totally agree.

+100

Merlin 2.1 is fine as it is. That's what you end with using the current 
CVS and Merlin command line. Its ready to go, validated against all of 
the cornerstone components, James, a bunch of my own experiments, and a 
number of external projects (a couple of which have successfully pushed 
Merlin well beyond the things I’ve been playing with).

Merlin 3.0 is a process that is starting to emerge - its shaping up to 
be about (a) refactoring of the assembly system, (b) separating out a 
management model we can easily handle things with JMX, (c) improving the 
APIs exposed to container extensions, (d) setting the groundwork for 
distribution and federation, while (d) making sure that we having a good 
time!

Stephen.

-- 

Stephen J. McConnell
mailto:mcconnell@apache.org
http://www.osm.net

Sent via James running under Merlin as an NT service.
http://avalon.apache.org/sandbox/merlin




---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@avalon.apache.org
For additional commands, e-mail: dev-help@avalon.apache.org