Re: process for workprocess cte

[singer@almaden.ibm.com]

Jon Bosak wrote:

>| I think the following should be added: "The anticipated
>| lifetime of the committee is [X] months, and a maintenance
>| committee is/is not foreseent at this time". I think the
>| person submitting the proposal should specify in the
>| proposal the anticipated lifetime of the committee.

>This requirement has caused such an unbelievable amount of trouble in
>W3C work that we're starting to give up on it.  (I can't go into
>further detail because I don't know for a fact that all the people on
>this list are W3C members.)  Both Eduardo Gutentag and David Singer
>have registered opinions to the contrary, but I maintain that in
>designing XML standards, the idea that you can force out a usable
>result by setting a date for completion can easily yield an
>illustration of Brooks's classic dictum about expecting nine women to
>have a baby in one month.

While it's true that you may not be able to stick to a deadline in
standards work (or even in childbearing...my son was two weeks late!), it
is also the case that work which starts without a target schedule tends to
wander on and on and on and on.  Setting a date helps focus the work of the
committee -- and if the date turns out to be unreasonable, it's always
possible to choose a new date, prompted by experience.  I would find it
very difficult to sign up for an indefinite committment, and quite frankly,
I believe that the sorts of standards OASIS will be approving will require
timely completion to be of value anyway.  I don't expect many OASIS
standards to be exploratory, but rather I expect OASIS to be standardizing
domain-specific work, which should be pretty well-understood when the
effort begins.

-- David Singer, STSM and Program Director, Web Technology
   Member, IBM Academy of Technology
   IBM Almaden Research Center, San Jose, California
   +1 408 927 2509 (t/l 457-2509); fax: +1 408 927 4073 (no t/l)