It seems no one is interested in the TLP idea anymore. We better discuss this now, rather than thinking silently. At the same time I don't think these discussion are reasons for people to get frustrated. Its just perspectives and being a diverse community we should be able to digest other ideas.
Anyway, I think pretty much everyone agrees we should move to a new TLP. But the question is what we should move with us. So I think its good to start a discussion on this. So what do you all think? Shall we move WS-Commons with it or not? I think we should, especially because of the convenience of axis2 developer. And I don't think that will hinder anyone using anything within WS-Commons components. What do you all think? Let's have a friendly discussion. Thanks, Chinthaka On Fri, Nov 28, 2008 at 1:32 PM, Eran Chinthaka <[EMAIL PROTECTED]>wrote: > Hi David, > > Please see my comments in-line. > > > On Thu, Nov 27, 2008 at 3:06 AM, David Illsley <[EMAIL PROTECTED]>wrote: > >> OK, I haven't discussd my thinking yet, so here goes. >> >> IMO the WS project is a disparate collection of stuff. This leads to: >> a. A lack of a coherent community > > > I agree with this point. Thats why I agreed to the idea of TLP, at last > (thanks to Ajith and Ruchith ;) ) > > >> >> b. A tough job for users to get what they want from the WS project > > > I'm not sure what you meant here. But what is so difficult for users? May > be I'm wrong as I am wearing Axis2-dev hat and not the user hat. > > > > >> >> >> I think that Axis2 and related technologies would be better served as >> a TLP which gathers just related technologies together. This should >> encourage a coherent community within the TLP, and making a TLP level >> decision to, for example, ship an Axis2 RSP bundle with Rampart and >> Sandesha pre-configured - the lack of whih is something I'm sure if >> hurting Axis2 at the moment, > > > Totally agree. > > >> >> >> Given those thoughts, in my perfect world, I'd split Axis2/Java and >> Axis2/C into different TLPs as they do have pretty different >> communities (user and dev). >> >> As to which Java based projects I'd ship to Axis2/Java and which I'd >> leave in WS... >> >> Axiom... leave in WS and promote to a proper sub project. It has a lot >> to give, but is currently very effectively hidden from users. I think >> this would continue to be the case if it were moved under an Axis2 >> TLP. >> > > It has lot to give, agree. But by moving that in to Axis2, will that be > hidden more? With Google, it is hard to hide something. Search for "XML stax > object model", "XML object model", etc. You will see Axiom in first 5 links. > If you use some other keywords, if Axiom is not coming up, then we need to > improve our Axiom page with SEO techniques, which I think is a different > question. > > So I don't think there is a visibility problem. Its all about, how famous > and effective it is than the URL or the location. > > >> >> Neethi... last time I looked, it was only really useful when used with >> Axis2, so I'd move that to an Axis2 TLP > > > +1 > > >> >> >> WS-Commons Transports... These are genuinely used by multiple TLPs and >> there is unlikely to be much 'user' interest in this project because >> we'll presumably ship them as part of Axis2 and Synapse. Therfore I >> don't think it matters much, so staying in WS seems reasonable. > > > I'd rather push these in to Axis2 TLP. > > >> >> >> XmlSchema... leave in WS and promote to a proper sub project. Used >> widely, not Axis2 specific, and deserves a little more visibility. > > > I'm 0- on this. > > >> >> >> Sandesha2, Rampart, Kandula2, Savan... move to an Axis2 TLP. These are >> (reasonably) tightly coupled to Axis2, drive Axis2 in interesting ways >> and are in demand by a large number of Axis2 users. Moving them to an >> Axis2 TLP would (IMO) allow that PMC to ship a compelling bundle and >> build a coherent community. > > > +1 > > Thanks, > Chinthaka > >
