Re: [mpowr] WG Formation

[Pete Resnick <[e-mail]>]

On 2/14/04 at 2:58 PM -0800, Dave Crocker wrote:

>1. Impose a rigorous barrier to IETF entry
>
>It will filter out quite a bit of poor work, simply by requiring 
>that a group demonstrate that it can be productive, _before_ it is 
>made a working group. This is an entirely natural barrier.
>
>Do we have significant indication that working groups with truly 
>lousy initial behavior later turn out to do something valuable? (I'm 
>looking for a pattern here, not an exception.) If a working group 
>cannot have its act together initially, it is not ready for open, 
>standards-oriented engineering prime time. So let's not give them a 
>ticket to the game until they are ready.
>
>Please note that I did not say to ignore them until they are 
>chartered. The suggestion to list them gives them visibility. I've 
>no doubt some other cheap and useful actions will also help.

I'm trying really hard to figure out what exactly is gained by not 
chartering a group but still letting them start their work, put them 
on a list, and expect the IETF-writ-large to start monitoring and/or 
working with them. If the IETF-writ-large is going to have to 
participate anyway (i.e., the folks in the not-yet-working-group 
request and get early cross-area feedback from other folks in the 
IETF before they produce their first I-D), that sounds like using 
exactly the same resources that a working group would. If the 
IETF-writ-large is *not* going to participate, there is some great 
likelihood that the group will not get good cross-area review, they 
will float off into the weeds, and they will have invested a great 
deal of effort in what will turn out to be useless work (i.e., the 
"late surprise"/"not enough early feedback" problem), which I thought 
was exactly one of the problems we were trying to fix.

(Note that I have not made a distinction here between "outsiders" who 
want a working group and IETFers who want a working group. There is 
now a significant enough number of people inside the IETF who are 
perfectly capable of doing lots of bad work if left to their own 
devices.)

If we're going to ask folks to demonstrate "good work", that's going 
to include:

- getting architectural input, security input, transport input (e.g., 
if your working in applications), applications input (e.g., if you're 
working down in the lower layers), etc.
- having a well-run mailing list with good moderation (i.e., a chair)
- producing I-Ds.

Once we add putting them on a public list or other "visibility 
mechanisms" (maybe having a document stating what they intend to do, 
i.e., a charter), what do we have that distinguishes that from a 
working group? If the answer is, "We don't have to spend weeks having 
the IESG argue about the charter, and the default action if they do 
bad work after a while is that they go away", why don't we just tell 
the IESG that they should stop arguing about charters and that bad 
working groups should be killed early and often?

pr
-- 
Pete Resnick <http://www.qualcomm.com/~presnick/>
QUALCOMM Incorporated - Direct phone: [phone], Fax: [phone]

_______________________________________________
mpowr mailing list
[e-mail]
https://www1.ietf.org/mailman/listinfo/mpowr