General comment on draft-ietf-problem-statement-00.txt

[Margaret Wasserman]

At 07:56 AM 3/3/2003 -0800, Eric Rescorla wrote:
>I think this goes to the point I was making earlier about what the
>purpose of WG chair is supposed to be.
>
>The responsibilities of the WG chair that you enumerate above are
>essentially procedural responsibilities, namely making sure that
>consensus is achieved, that people do the jobs they've committed
>to. By contrast, you assign the substantive responsibilities (namely,
>making sure that the right thing gets done) primarily to the AD.  I'm
>not surprised, therefore, that you don't think having an AD present
>undermines that authority.

I think that this distinction is fundamental to many of the
differences of opinion that people in the IETF have about
out current management structure.

Some people are comfortable with the idea that all substantive
decisions rest with the IESG (with the possibility of appeal to
the IAB).

Personally, I think that the IETF probably did work better
when all substantive decisions _could_ be handled by 13 people,
but I don't think that model scales to an organization of this
size.

Instead, I think that we need to find ways to push (or pull?)
authority (including substantive decision making) out to a
larger group.  I don't know what the best way is to do that,
but some possibilities would include:

         - Enlarging the IESG, which could be done by:
                 - Just adding more people (but this has
                         group dynamics problems discussed before).
                 - Splitting the responsibilities between more
                         than one group.
         - Giving more authority to WG chairs, and holding them
                 more accountable for results.
         - Forming an additional "layer" of management between
                 the IESG and WG chairs with some substantive
                 authority (maybe to charter WGs within an
                 area, and approve Info, Exp and PS documents,
                 or something like that?).
         - I'm sure that there are many other possibilities...

Right now, though, it seems to be true that all major,
substantive decisions have to go through the IESG, including
both process and technical decisions.

One of the trickiest things in constructing a process for
a group to produce technical output is figuring out where
the major decision points are, and who should have the
authority for making the decisions at each point.  So, it
isn't surprising, after 10 years of strong growth and no
major changes, that the IETF processes could use some
tuning in this area.

Margaret