I think we should explicitly say that SGs follow WG rules.
I would agree. There are several reasons why this is probably the best
way to go. Aside from the openness issues, the amount of extra work
created for the IESG/Secretariat to have a different set of rules would be
very substantial. A SG group chair has all the same obligations to keep
things on track as a WG chair does -- and if they can't, then that is
probably a good sign that they wouldn't be a very good WG chair. A SG can
also have a design team, much as a WG can.
- to what extent is a SG allowed to frame the problem to be solved in a
way that would constrain a later WG if one were chartered? it's clear
that they're not supposed to develop protocol specs, but what about
requirements? goals? models of interaction between communicating
parties? terminology?
I think that depends on what items are chartered, and what their status
is. It is probably not a good idea to have charter a SG to develop a
Standards Track document (or maybe even a BCP), but the document currently
doesn't prohibit that (it just prohibits milestones relating to protocol
specifications). Do we need some additional language on that?
- are SGs allowed to request meeting space at IETF plenary meetings?
(this is a resource utilization question - does IESG need to treat SGs
more-or-less as WGs for the purpose of resource allocation? could SGs
preempt creation of WGs due to a lack of resources?)
Yes, they're allowed to request meeting space, same as WGs. In terms of
impact on meeting slots (which has been tight lately), the question is
whether a given SG would consume a slot by meeting that would not
otherwise be consumed. If the SG is going well, it's more likely
that either an additional BOF or a WG slot would have been
allocated anyway. If it's going poorly and dragging out, then yes, it
might consume additional slots. I think this speaks to the desirability
of chartering SGs for the minimal period of time. An alternative to "six
to twelve months" might be "six months", or perhaps "one or two IETF
meetings."
_______________________________________________
Ietf mailing list
Ietf(_at_)ietf(_dot_)org
https://www1.ietf.org/mailman/listinfo/ietf