Hi Paul:

The principal issue I have concerning Phoenix escalation to beta concerns (a) the XML schema used to describe blocks, and (b) the specification of how blocks are declared. Currently the Phoenix platform uses <blockinfo/> and the existence of blocks is derived from the assembly/configuration info. In the past, block existence was declared in the manifest files containing the block implementation - however this does not seem to be the case anymore. I believe Pete is also in the process of adding support for the containerkit schema.

At the same time we have components using the Merlin model which includes additional meta-info dealing with (a) lifecycle phase dependencies, (b) lifecycle extension handler provision, and (c) a type specific default configuration. Merlin provides support for both the "containerkit" DTD and the Merlin "type" DTD so loading containerkit (Phoenix) based components into Merlin will not be a problem. However, the absence of manifest based declaration of the components within a jar file is problematic. Merlin uses this information to establish potential component service provider candidates whereas Phoenix assumes that the everything is declared in the assembly.xml file.

Assuming that the containerkit schema is included in Phoenix, then, the only issue concerning interoperability of phoenix style components in Merlin is the resolution of the Manifest question.

We could also go a step further and replace the containerkit meta-info model for the excalibur.meta meta-info model. This would ensure an even higher level of interoperability and its very simple to do - all it implies is that the (a) containerkit meta info uses excalibur.meta instead of the current containerkit.info package (which is really easy because excalibur.meta is already packaged as a separate extension jar file), and (b) Phoenix would need to check for things it cannot support (e.g. lifecycle extensions implied by phase, extension elements), and (c) manage or throw out types that declare default configurations.


Cheers, Steve.


Paul Hammant wrote:

Folks,

Can we debate a feature cap for a beta release please? Last time this was raised, we were talking of imminent JMX work being the blocker..... has that now been achieved?

- Paul


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



--

Stephen J. McConnell

OSM SARL
digital products for a global economy
mailto:[EMAIL PROTECTED]
http://www.osm.net




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



Reply via email to