Ted Husted wrote:
> 
> "Geir Magnusson Jr." wrote:
> > The best solution, I think, is the most minimal and open one, and as I
> > see it, the most open one is simply a Jakarta project that acts as a
> > container for independent Jakarta mini-projects.
> 
> That's what we're doing -- in exactly the same what that tagblibs is a
> container for custom tag mini-projects.

So then the above is *not* what we're doing - we are doing a tablibs
model, not a jakarta model.  They are different, and I would argue the
jakarta model is preferable, as you can implement the taglibs model
under jakarta model as a self-ruling entity (like Taglibs is under
Jakarta), but the converse isn't true - you can't have a 'Jakarta zone'
under the taglibs model.

> The only functional difference between a Commons package and a Jakarta
> product is whether you have to vote a committer into * other * packages.

It's not the only difference, and it is a big difference.  No one has
explained what is wrong with the conventional jakarta model, why it has
to be changed.  

I interpret that as a guidance / control difference - the people
dedicated to working on a component can have their actions vetoed by
*anyone* without the conventional requirement that the vetoer have a
demonstrated understanding of and investment in what is being  vetoed by
being a part of the component's committer set.

There is also a CVS difference, as the components aren't separated into
distinct CVS spaces.

There is a user-community difference.

Documentation might be a tad challenging.

etc

> One other small difference is that we're putting new packages up to a
> vote by all the Commons committers, rather than a PMC board.

That's cool.  I don't see a problem with that.  But then when you
approve a component and the initial committers, get out of the way - let
the component 'owners' do their thing.  The taglibs model does that
through a cultural understanding ( I assume ), rather than the
organizational guidelines.
 
> In both cases, we're probably just ahead of the curve, and Jakarta
> itself will probably be doing the same things eventually.

Hm.   Not sure about that, but I haven't been hanging around Jakarta
enough to know.

geir

-- 
Geir Magnusson Jr.                               [EMAIL PROTECTED]
Developing for the web?  See http://jakarta.apache.org/velocity/

Reply via email to