Vision (was: Re: Is the default process correctly described?)

[Jon Bosak <bosak@boethius.eng.sun.com>]

[Terry Allen, replying to a point of Eduardo's that's going
to end up in a separate, later discussion about TC voting:]

| Yet another reason I want to get clear on what our goals are

The goal is to set up multiple parallel industry efforts to define
industry- or application-specific DTDs, schemas, and namespaces.

| and what the vision of the overall process is.  Thousands of
| committees?  Industry coordination congresses?  How does it
| all fit together?

Thousands of committees.  Which means among other things that

(a) They must be self-supporting

(b) They must make the smallest possible demands on the board and the
    membership in general

(c) They must be able to do something that's useful in advance of or
    even in the absence of approval by oasis as a whole

(d) But oasis as a whole must be able to confer a special status on
    certain specifications created by this process.

Notice that (a) means that this whole thing runs basically through
voluntary association, with oasis itself providing not much more than
just what's needed to ensure a fair process and facilitate the
formation of cooperating groups.  And this in turn means that what
coordination there is will also have to be provided through voluntary
association.

I assert that a self-supporting process of this magnitude can't be run
from the top down.  W3C has demonstrated the practical limits of what
can be accomplished with a top-down model run quickly, and ISO has
demonstrated the practical limits of what can be accomplished with a
top-down model run slowly.  Neither of those models is going to work
here.

We don't need to do anything to make XML specifications happen; XML
allows separate organizations to start developing independent
information standards all on their own.  The design problem here is
how to get the protozoa floating around in the primordial soup to
begin to cooperate far enough to start doing a little tissue
differentiation and agreeing on where the notochord is going to go.

In addition to technical committees I believe that we will need at
least the following:

1. Chartering committees (already described in the process
   outline)

2. Industry harmonization committees (suggested by the process
   outline)

3. Coordination committees (supported by resources contributed
   by the committees desiring coordination)

4. Shared subcommittees (which should be covered under
   whatever rules we specify for subcommittees)

5. Maintenance committees (for reasons enunciated by Terry and
   Eduardo)

6. Translation committees (more on these shortly)

I believe that we will eventually need to specify all of these and
maybe even more, but that we can start with 1 and 2 right after we've
figured out how to fix our current process, about which more in a
separate message.

Jon