Re: Proposed Experiment: More Meeting Time on Friday for IETF 73

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

 




--On Thursday, 17 July, 2008 18:19 -0400 Tony Hansen
<tony@xxxxxxx> wrote:

> It would be be best if the Fri afternoon slot were filled in
> early rather than as the last slots to be filled in. That way
> people would have more notice that they're being included in
> the experiment and there'd be less of a chance of a rude
> surprise.

Russ,

I agree with Tony.  It is perhaps worth mentioning that we've
tried versions of this before (and done so without clear
evaluation criteria -- Brian is right about that).  The mistake
in the past was that the scheduling algorithms have sent a clear
message that there will be few IETF sessions on Friday and that
most of them will be unimportant.

There are also two rather different experiments that you could
carry out here.  I'd favor the second, but you should at least
be clear about what you intend.

Experiment 1: The Friday afternoon slots are just slots, like
any other slots, and the Secretariat tries to distribute WGs
across the entire week including those slots.

Experiment 2: You treat the Friday slots mostly as an overflow
area, using scheduling rules like:

	* If you ask for three slots (and the relevant signoff
	process occurs), you are almost guaranteed a Friday slot
	and might get two or three.
	
	* If you ask for two slots, the odds of at least one of
	them being allocated on Friday are very high.

That would make things fairly predictable, and would work from a
scheduling standpoint unless several of the "we need multiple
slots" WGs have overlapping membership.   It would also probably
work out without disrupting the various groups that may be
meeting on Friday now (an IRTF group has been mentioned and, if
I recall, ISOC sometimes uses Friday time) in that missing one
session of a two-slot (or more) WG is typically much less bad
than missing the single session of a one-slot WG.

But the two models are very different and any experiments would
find themselves testing different things.


One other observation:  To the extent that the reason for doing
this is, as indicated in your note, "...Several WGs are not able
to get as much meeting time as they need to progress their
work...", I would encourage the IESG to very carefully evaluate
what is actually going on above and beyond whether having more
of Friday available would help with scheduling.  For example,
have these WGs shifted from getting most of their work done on
mailing lists to doing almost everything in meetings and, if so,
is that a problem that needs fixing?  If mailing lists are being
used adequately, do the WGs have a task management problem
(e.g., trying to work on too many things)?  Are they conducting
meetings efficiently and using the time well?  And so on.
Basically, if we believe in our claims that most work gets done
on mailing lists, if a WG routinely requires most than one slot
and especially if they require more than two, I would wonder
more about what isn't working and how to fix it than I would
about allocating more meeting time.

best,
   john



_______________________________________________

Ietf@xxxxxxxx
https://www.ietf.org/mailman/listinfo/ietf

[Index of Archives]     [IETF Annoucements]     [IETF]     [IP Storage]     [Yosemite News]     [Linux SCTP]     [Linux Newbies]     [Fedora Users]