> On Wed, 12 Nov 2003, Henning Schmiedehausen wrote: > > > I still ask myself why we can't put HiveMind as its own project under > > the Jakarta umbrella. > > Isn't that what this proposal is proposing? If it isn't, then > [EMAIL PROTECTED] is the wrong list to propose it to. > > > > > We have projects with a much smaller scope as "normal" jakarta projects > > and we have and had framework projects such as Cocoon, Avalon or Turbine > > outside of the commons. > > > > IMHO the scope of HiveMind is already to big for the commons which I see > > as "software snacks": Small, easy to swallow and a side order for larger > > projects. > > > > >
Yes, this proposal is that HiveMind be a top-level Jakarta project; a peer to (say) Log4J and Tapestry. The commons is largely supposed to be for toolboxes; libraries that contain individual utility classes (commons-lang) or frameworks with a very precise, very focused use (commons-logging). Although you could scavenge a lot of useful stuff out of HiveMind, it is intended for use as a cohesive unit. In terms of homes, I can see a number of possible options: 1) Reject HiveMind outright --- and make Howard very sad. In this case, I would reorganize the Tapestry build to accomidate having HiveMind as a sub-project (once the grants from WebCT come through). I feel that is less than ideal; despite the overlap, and evolution of ideas from Tapestry into HiveMind, I'd prefer to keep them seperate, with seperate teams. 2) Keep it in the commons I have no real objection to this personally, but HiveMind doesn't quite fit with the commons charter, for the reasons discussed above (and in the commons charter). 3) Chuck it over to Avalon I've looked to see how we could graft HiveMind into Avalon and vice-versa, but they are really quite different beasts. The type-1 vs. type-2/type-3 split is intrinsic and difficult to reconcile. HiveMind's concept of a module doesn't map so easily into the Avalon space, and HiveMind's free-for-all approach doesn't jive with Avalon's dogmatic security model (including its explicit application construction descriptor). 4) New TLD I think HiveMind is a totally useful swiss-army knife that opens up the doors for a lot of really terrific approaches to solving common development problems. However, it's still a bit light to be a TLD ... something like 6600 NCLOC if memory serves (don't bet on it!). Certainly if boil-the-ocean frameworks like Struts, Tapestry and Turbine aren't TLDs then HiveMind isn't either. 5) New Jakarta Project As stated in the proposal; this is my preference. -- [EMAIL PROTECTED] Creator, Tapestry: Java Web Components http://jakarta.apache.org/tapestry --------------------------------------------------------------------- To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]
