Re: [RAI] RAI processes for handling work effectively

Harald,

What you say makes some sense to me.

But if we were to act on it then I think we need to replace some WGs 
with review panels. (We have an SDP directorate, but I don't think that 
is a sufficient replacement for mmusic.) And so we will need some clear 
processes for review panels. How would we process drafts that would 
otherwise have been handled through the standanding wg? Would they be AD 
sponsored? And we may need to schedule meeting sessions for review 
panels too.

This seems like it would need to be an ietf-wide change, not just RAI. 
(But maybe we could pilot it in RAI.)

	Thanks,
	Paul

On 7/18/13 7:57 AM, Harald Alvestrand wrote:
> Gonzalo,
>
> let me make a radical proposition....
>
> Standing working groups are useless. All of them.
>
> What we have been doing in the IETF for 10+ years is to create things
> that we call working groups, which sometimes are used as places to do
> work, but really are places where people go to get a review and a stamp
> of approval.
>
> This isn't the role of a working group. It's a review panel.
>
> Working groups need a task they can complete. They need chairs that are
> willing and able to get the hard questions asked, get a decision out,
> and PUBLISH.
> Review panels need to have competent people on them, and they need to be
> able to say NO at need.
>
> The drivers for these 2 kinds of activity are orthogonal-to-opposed,
> they are not aligned.
> And by using WG procedures for both kinds, we are doing a disservice to
> ourselves.
>
> I'd suggest that among the requirements for a working group to be formed
> - AND for it to continue to exist - there should be:
>
> - A definition of the problem that needs solving
> - A definition of the work that needs to be done
> - A sign that one can look at in order to tell that it's finished
>
> (Picking on MMUSIC not because it's the only one, but because I'm
> working in it at the moment)
>
> In contrast, the MMUSIC working group includes statements like "Various
> extensions to SDP will be pursued to remedy the most urgent of SDP's
> shortcomings". How do we tell that we'll achieve that?
>
> To my mind, SCTP-in-SDP could have been a working group. Once there are
> two working implementations that can successfully negotiate SCTP, its
> work is done and can shut down.
>
> Trickle ICE could have been another. The day we have 2 implementations
> interworking - done.
>
> Both need to have competent review and input from people who are aware
> of the architectural issues of SDP, and the people who provide that
> input need to talk to each other so that each reviewer is in sync with
> the others.
>
> But that isn't the role of a WG. It's a review panel.
>
>          Harald
>
>
>
>
>
>
> _______________________________________________
> RAI mailing list
> RAI@ietf.org
> https://www.ietf.org/mailman/listinfo/rai
>