On 14 Oct 2004, John Levine wrote:
With respect to the requirements document issue, nobody is
suggesting a waterfall model. My usual practice is to develop both
the requirements and the specification simultaneously.
Documenting what the spec is supposed to accomplish is a fine idea.
My concern was with attempting to do so before we know what the spec is.
Which is why we call it a requirement (or maybe more correctly goal).
We want to have list of what we want to accomplish available. If we
just document what we'are doing this is what I-Ds are for. I don't see
easy way to work on requirements together with implementation unless
we can separate what we can find way to separate what we want into
parts that can be accomplished separately (i.e. work on requirement
for one part while working on specification on another).
The rest of your suggestions seem reasonable too, particularly if we
can find a sugar daddy to pay for the conference bridge.
I'm against telephone conferences, this is likely to alianate non-US
and non-english participants and is not the way IETF works as it does
not really use internet.
Instead of telephone conferences we can do regular jabber sessions which
is typical way its done at IETF WGs.