I know that Avalon has decided to focus on Merlin. I have no problems with that, and I wish you all the best in that endeavor. I thought that the exact direction of Fortress was still up in the air--beyond no longer supporting it.

If that is the case, I would like to host it over on D-Haven, or perhaps something like it that is compatible. While I know there should be no issue with the latter course, the first one does require a "buyout" so to speak with Avalon.


anyone talk to Shash yet? Keelframework being one of the big fortress users, his thoughts would be valuable :-D


Since when did users matter? :-) And, better yet, since when did the original developers matter?

Keel definitely derives a big benefit from Fortress, and to me it feels "just right". I will definitely do what it takes to keep Fortress alive. D-haven hosting works for me, particularly since Berin has never, ever shown any interest in keeping Fortress alive :-). I think the Keel community would welcome the CVS hosting of Fortress in Keel also, if that is what needs to be done. However, as Jaaron has suggested, keeping Fortress under the Apache umbrella (as long as Berin, Peter, Leo **2, and other original authors of the framework have authority over it) makes a lot of sense; let us not underestimate the value of the Apache name: the original authors deserved that label and support, and should continue to do so.

As I said in my earlier post, in response to Niclas' note about the IRC log, I am concerned that one implementation (even though it is Avalon's own) should drive the interface (framework) that should drive *all* implementations. I think that is purely asinine! If the interfaces don't change, and it is only about meta-data...then I'd be less concerned. We and a lot of other folks have developed a lot of code based on those interfaces. We shouldn't have to unnecessarily change a single line of code to keep using the INTERFACES, whether or not Fortress is forked; Apache and Avalon owes us that much.

Overall, I think what is happening in Avalon is completely bass-ackwards! Here we have an Apache Avalon interface that is stable, implemented by three mature containers (ECM, Phoenix, Fortress) and one getting mature but evolving (Merlin), in use by two superbly successful Apache projects (Cocoon and James). We have most of the original developers behind that. That to me is the legitimate legacy worthy of the Avalon name. Yet, all of that proven technology is being frozen or forced to fork, most of the original developers are quitting in disgust, while completely unproven stuff (as of yet) is being given the mantel and the Apache/Avalon name. Go figure. I thought only Leo's initials had LSD in it!

Leo, Berin, or even the Avalon PMC....anybody that is listening. What do we need to do to a) get ECM, Fortress, Phoenix, Loom, Jcontainer, Plexus and other containers using a common interface banded together? b) Can we start a mailing list with the known interested folks? c) I don't understand the beaureaucracy, what are our alternatives in taking our complaint and concerns outside Avalon PMC and to a higher body within Apache? In other words, what can users do to get their voices heard?

Shash


--------------------------------------------------------------------- To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]



Reply via email to