Ted suggests:
> 1) That we officially choose a project manager for the project.
I vote for Ted as PM. He's been doing a marvelous job so far....
> I think that everyone would probably agree that we are now ready to start
> tightening up the project and moving on to the project's Development
> phase. Thankfully, the Brainstorming phase of the project has provided a
> rich base of ideas for us to work with.
Actually, Ted, I'm going to be a bit contrary. I don't think we're ready for
development yet. We haven't got an architecture yet....and then we need to
spec out the interfaces and contracts that will make up this architecture. That
will likely require a lot of brainstorming still. Then we'll be ready to "develop"
some code. \
If you were using the term "development" in a more generic sense (develop
the architecture/specs/etc) then I'm in agreement...but would caution you not
to use a word that can be mis-interpreted so easily by programmer types.
> I think it is important for us to keep in mind that we will be developing
> a full-blown Embedlets specification which should be very similar to the
> Servlet specification. The current Servlet specification is 257 pages
> long, includes a large number of sections, a significant amount of
> formatting and a very detailed table of contents.
I think that the Embedlets spec needs to be a lot simpler than than the servlet
spec, otherwise we're going to spend too long on it, and risk making things a
bit too complicated for developers of Embedlets. It should also be possible to
spec out certain areas and then do some prototyping of a reference
implementation while other parts of the spec are still being worked on.
Interative and modular approach....delivery early, deliver often.
Keep in mind that documentation usually lags badly in open source projects. It
would be nice if Embedlets didn't fall into this trap.
Andrzej Jan Taramina
Chaeron Corporation: Enterprise System Solutions
http://www.chaeron.com
|
