[30/05/2005 22:44:40] brett: packaging vs type
[30/05/2005 22:45:52] brett: To be able to depend on only some of those artifacts, <type> on a 
 dependency may not match packaging - if you declare <type>ejb</type> you
 only get the ejb, and if you declare <type>ejb-client</type>, you only
 get the ejb-client JAR.
[30/05/2005 22:46:11] brett: This also means that while type is not considered in whether an project 
 is unique, it is for a dependency - so you can download tld files with
 the same name as the JAR they belong to.
[30/05/2005 22:46:19] brett: (just pasting the important bits)
[30/05/2005 22:46:30] brett: problems to arise:
[30/05/2005 22:46:34] brett: - inconsistency could lead to confusion
[30/05/2005 22:46:39] brett: - guaranteeing snapshots match versions
[30/05/2005 22:48:00] brett: I'm ok with the inconsistency - the naming is different to be clear, and I think we resolved that earlier
[30/05/2005 22:48:10] jason: yah, that's fine
[30/05/2005 22:48:54] jdcasey: alright, I like the idea of deriving the type name from the packaging, though
[30/05/2005 22:48:54] brett: artifacts are solely identified by groupId:artifactId now
[30/05/2005 22:49:18] jdcasey: and I'd be +1 for having an <attachments/> element inside of <build/> or somesuch for non-derived artifacts
[30/05/2005 22:49:18] brett: jdcasey: where appropriate. jar-tld is a bit funny
[30/05/2005 22:49:19] trygvis: should we separate between project and artifact id?
[30/05/2005 22:49:28] trygvis: g:a is project id, g:a:type is artifact id
[30/05/2005 22:49:45] jdcasey: trygvis: +version ;)
[30/05/2005 22:49:53] brett: jdcasey: noooo
[30/05/2005 22:50:01] trygvis: nope
[30/05/2005 22:50:03] brett: its still the same project
[30/05/2005 22:50:06] jdcasey: right
[30/05/2005 22:50:11] brett: version is about to become somewhat fluid :)
[30/05/2005 22:50:32] *** evenisse has joined #maven.
[30/05/2005 22:51:19] jdcasey: anyway, as for the snapshot issue, I think the <attachments/> could fix that...you just force everything to be deployed every time...otherwise, it's not really a snapshot, is it? I suppose you could suppress attachments for quick-n-dirty deploys
[30/05/2005 22:52:16] brett: jdcasey: I don't think they need to be listed?
[30/05/2005 22:52:39] jdcasey: what if my project doesn't want to deploy source artifacts, or javadoc artifacts?
[30/05/2005 22:52:45] jdcasey: or should I have that choice at all?
[30/05/2005 22:52:48] trygvis: deployment should always contain "tied" artifacts (clients, source bundles etc)
[30/05/2005 22:52:58] trygvis: (by default)
[30/05/2005 22:53:11] brett: yes, you should have a choice
[30/05/2005 22:53:13] jdcasey: I agree on that for derived artifacts, but maybe not sources, etc.
[30/05/2005 22:53:27] brett: I want sources and javadocs on by default
[30/05/2005 22:53:44] trygvis: for deployments I would say that the sources should be deployed by default
[30/05/2005 22:53:50] trygvis: if not people aren't going to do it
[30/05/2005 22:53:59] brett: hmmm
[30/05/2005 22:54:05] brett: I might have another look at this
[30/05/2005 22:54:06] jdcasey: the default is not really an issue here.
[30/05/2005 22:54:09] brett: its not needed for snapshots
[30/05/2005 22:54:11] jdcasey: we can do anything by default
[30/05/2005 22:54:38] jdcasey: snapshots shouldn't take a year to produce, and turning on javadocs for every snapshot will gum up the works pretty damned quickly
[30/05/2005 22:57:25] *** adc has signed off IRC (Leaving).
[30/05/2005 22:59:12] brett: ok, let me step through what we need
[30/05/2005 22:59:34] brett: type = ejb-client needs to somehow figure out that is packaging = ejb; classifier = client
[30/05/2005 22:59:46] brett: extension = jar
[30/05/2005 23:00:11] brett: type = tld needs to be extension = tld
[30/05/2005 23:01:03] brett: I think these all come from the artifact handlers again
[30/05/2005 23:01:09] brett: though tld is interesting
[30/05/2005 23:01:19] brett: I guess the war plugin could provide that one
[30/05/2005 23:01:45] jdcasey: just a quick question: do we need to tie these derived handlers to the main one through something like ArtifactHandler.getDerived()?
[30/05/2005 23:02:14] brett: I need to play this one out some more
[30/05/2005 23:02:24] brett: I haven't considered the binding at all I realise
[30/05/2005 23:02:25] jdcasey: that's cool
[30/05/2005 23:02:31] brett: will do tomorrow
[30/05/2005 23:02:45] brett: (but something like that, yes)
[30/05/2005 23:02:58] brett: but only for some
[30/05/2005 23:03:05] brett: eg, JarHandler need not know of Tld
[30/05/2005 23:03:26] *** adc has joined #maven.
[30/05/2005 23:03:29] jdcasey: sure
[30/05/2005 23:03:41] jdcasey: that's actually a good point
[30/05/2005 23:04:00] brett: yeah, we can do that once I flesh that out. I just forgot I hadn't dealt with that area of it.