Hello all.
IMNSHO, the preferable solution for Avalon and related communities is to have both an Excalibur and Merlin TLP. My suggestions of incubation and partitioning were always "second bests." We're half way there, so let's finally finish what we started and get Merlin a home of its own.
That still leaves us with several questions:
1. What goes with Merlin? 2. What can be or should become more generic? (ie- avalon-repository) 3. What's left in Avalon? (ie- What's Avalon's new mission?)
Here's my takes:
1. What goes with Merlin?
Initially just about everything. Magic and avalon-repository could be made more generic of course, but for now, leave them with Merlin. It's just simplier that way. Then the Merlin community can deal with how (and if) to best spin off such subprojects.
2. What can be or should be more generic?
My thought is to handle one things at a time. Sure Magic could be a new build tool (maybe there needs to be a ASF Build Tool Federation for things like Ant, Maven, Gump and Magic...). Anyway, point is that let's get Merlin up as a TLP first. Then we can worry about other subprojects.
3. What's left in Avalon?
For me, this is the more interesting question. I appreciate Stefano's sentiments expressed in his "My personal view of Avalon" email. Avalon's goal *has* been a holy grail. I think we all knew that. However, there has been useful and practical code come out of Avalon and that code shouldn't be locked up and frozen somewhere.
There's a couple of options for Avalon post-Merlin. My current favorite leaves Avalon with the framework and perhaps components like Cornerstone. Or the components could go with Merlin and Avalon would just be the framework. I kinda like the idea of a component commons, but I'm not sure if it would work.
Mostly I'd like to see Avalon house some really good documentation that would help users of the Avalon framework like Merlin, Excalibur, Cocoon, Turbine, James, and so on. I've been wanting to write this documentation for a long time but the climate has always been so tenuous that I haven't felt it would be worth my effort.
I once compared Avalon to a sort of Xerox PARC project: lots of cool ideas started here but never took off until they *left* Avalon. In fact, I don't really mind that. Perhaps Avalon could remain a place of experimentation in IoC/SoC/COP development, spinning off new technologies when they begin to mature and show promise. There are lots of options and possibilities and there's nothing wrong with a heritage like that.
So those are my thoughts. Along with Niclas, I would like to see a solid proposal by the August ASF Board meeting. If everything went smoothly there could be a Merlin TLP by September.
WDYT?
jaaron
--------------------------------------------------------------------- To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]
