>Cool! >That's what I was fighting for (specially with trygvis ;)) for couple of >months. > > Enough of this. Maybe I'm reading this wrong, but this looks to me like you are gloating and trying to put Trygve down by singling him out. There is no place for that here.
On this discussion, there is nothing on the mailing lists. The only opinion you had were your comments on the confluence document. We worked through them. I regularly asked if there were any outstanding issues. Given that the discussion stopped, I assumed we were in agreement. We were in agreement to use the other technique, because we saw the ejb-client situation being handled differently. In other areas, we adopted your suggestions. Now, after more time and consideration we've reconsidered and are taking on another of your suggestions, and Trygve, John and I at least are in agreement on it (I haven't heard from anyone else). It's fine for you to pipe up and agree or disagree with reasons, but I really don't want to hear any more about how you've had this all figured out from the start, without actively being involved. It's not helping us actually get some code done and out the door. >Definitly this makes implementation simpler and more consistent. > > Yes, I agree. >Actually it is rather neccesery to change it to make implementation at all >possible. > > I don't agree with this. We can still discover the POM everywhere except the parent - and it is probably reasonable that we could require that you only inherit from projects that are packaging = pom, but given the other solution is easier, and flexible - we'll use that. >The last thing about repository layout, which I don't really like is the way >the groupId is specified. > > As you've said. On this, it's just a matter of preference and we've gone with what users originally asked for and the majority of active committers think is best. I don't see any reason to change it. - Brett
