Mike, I like this. Many thanks for your efforts in helping to get the
group moving along. It's really important that folks read Mike's email and
let us all know what they think of the suggested goals and milestones.
--On Thursday, December 13, 2001 17:55 -0600 Mike Dahlin
<dahlin@cs.utexas.edu> wrote:
> Goal: Near-final draft of first research update protocol complete for
> Dec 2002 IETF meeting
> (Rationale: if it isn't, then either the protocol being considered
> is too complex or the level of interest in the group is too low to
> make fast progress.)
Agree. For the most part I agree with the goals and dates here. If
possible we should strive to keep the final revision away from early March
if possible since that will likely interact badly with the I-D cutoff date
for Minneapolis. Also, in the (hopefully unlikely) event that I have to
return to the UK, I'd have to disappear around that time also.
>
> To meet this goal, we need to meet the following sub-goals:
>
> Sub-goal: "Last call" RUP requirements document complete by March
> IETF meeting
>
> Note: this will not happen if we wait for a "flurry of
> activity" in early March, so...
>
> sub-sub-goal: revised draft (fixing all "internal inconsistency
> details" such as definition of terms) posted to group by
> mid-January
>
> sub-sub-goal: assemble list of "significant open questions"
> from mailing list by late January/early February
Yep. But it's probably worth also pointing out that these two goals (and
the following one) are not necessarily to be considered in sequence. If
there are already significant open questions then it would be worth having
them sooner rather than later, even if we don't get around to considering
them for the mid-January milestone (it might be possible to resolve *some*
of them).
> sub-sub-goal: resolve these open questions through mailing list
> discussions and update document appropriately by end of Feb
>
> sub-sub-goal: authors/editors fix any new "internal consistency
> details" that have crept into document during revisions by
> early March
With respect to the author/editor part above, I think we're now at a point
where the document needs to move under editorial control. I'd like to
thank Dan for her great work in helping us get the document to its current
state, but suggest that we remove Dan from the day-to-day editing loop so
that her skills can be used elsewhere in the group.
I've been a busy little bee while in SLC this week, and one of the really
useful suggestions (thanks to Ned Freed and Joe Touch and others) is that
we rename the "requirements" document to be a "guidelines" document. We
have collected a lot of good issues in the text but I heed the warnings
I've received of the potential for future problems if we end up with
requirements that have not been met. Group consensus should be sufficient
to resolve any issues of that nature in the future.
> Sub-goal: rough draft of one or more proposed protocols + list of
> "hard issues" that need discussion by summer IETF meeting
> sub-sub-goal: one or more potential approaches identified at
> high level by March IETF meeting
While Mike's right in that the meetings should be used to hack out issues
that we can't resolve on the mailing list, I'd caution the group about
putting too much focus on those meetings. As Mike has also said, we don't
want flurries of activity just because there's a meeting due. The primary
work of the group must be done outside physical IETF meetings.
> Sub-goal: near-final draft by December...
This also seems appropriate.
--On Friday, December 14, 2001 09:28 -0800 Dan Li <lidan@cisco.com> wrote:
> At 10:53 AM 12/14/01 -0500, Patrick McManus wrote:
>
>> I think if webi begins useful protocol work with an expectation that
>> the work would be accepted you would seen engineering input in droves.
>
> I cannot agree more. Given that we've done much RUP requirements work and
> the main building block "invalidation" is well understood, I think webi
> should move on to protocol design, while closing the RUP req ID. Chairs,
> how about solicit interests and assemble a design/edit team?
I've been doing a lot of thinking on this issue this week.
My concerns:
1) Folks may be so involved in the fun protocol stuff that it's difficult
to get resolution on issues in the requirements/guidelines document
2) Addressing point #2 in David Martin's email regarding discussion at the
lunch BoF, it's possible that there may be some minor additions that we
might want to capture and understand before getting too far involved in
protocol work
3) The only protocol proposal on the table at the moment is WCIP. Other
than IBM I've heard at least a couple of other folks mention that they
might be in a position to document "hacks" (or more refined protocols) they
use. This will give the group a wider range of things to consider. I'd
suggest we provide folks some (short!) period in which to get drafty drafts
published before we start working on protocol specs. The chairs are open
to suggestions on an appropriate deadline for those specs to be posted
[though obviously any work submitted after that date must also be
considered by the group]
4) I'm still not 100% sure we've resolved the relationship between CDI and
WEBI on the RUP side of things.
5) While I think there's some additional process we can use to facilitate
protocol design work, I believe that team will simply be the set of folks
that participate on the list anyway *grin*. But seriously, I agree with
Dan that folks that are particularly interested in really doing work should
get in touch so that we can see about setting up separate mailing lists etc.
OK, so those are my concerns, here's my proposal:
A) Someone from IBM please tell us what "soon" means
B) If "soon" for a protocol draft is after mid-January 2002 then I think
I'd err to agree with Dan and recommend the group start to move forward,
ensuring someone that knows about IBM's stuff is definitely involved in the
design group
C) We get some real focus on the requirements/guidelines document between
now and mid-January. This work item is our #1 consideration until that
time.
C.1) Folks interested in protocol work continue to work on personal I-Ds
C.2) We try to make sure that WEBI and CDI know where the boundary is and
try to make sure we do develop one thing, but not two similar things
(unless that's really the appropriate thing to do)
D) Around mid-January we move protocol work to be our #1 consideration,
while striving to get the requirements/guidelines document completed ASAP.
Thoughts? (I tend not to make assumptions, so I'll take silence as total
lack of interest and an indication we should close the group.)
With respect to item #7 in David Martin's email, and the dates posted
above, I'll take an action to speak with the ADs on the appropriateness of
fixing some of the dates in the charter (I don't think any of us want to
change text since that would involve more work for all of us). In order to
do this I want to make sure that the dates Mike has proposed seem
reasonable to the group. Folks might want to consider whether we want to
re-date the IDD stuff that's in our milestones (see another message for
more on that).
This archive was generated by hypermail 2b29 : Thu Nov 18 2004 - 11:23:00 MST