On 12/20/2013 4:46 AM, Jari Arkko wrote:
However, the preference is not to be an author. In particular, if you
are an author you should not be the chair deciding or document
shepherding about the topic. This is one reason why we have multiple
chairs.
In general, it is easy to get into a situation where it is difficult
to progress a document because all the parties who should do
something about the document are recused due to being authors. Having
other people hold the pen and the chairs doing their chair job is a
far better model.
On 12/20/2013 7:36 AM, Carsten Bormann wrote:
> More eyeballs see more issues, adding to the
> gene pool etc.
>
> But there is a limit to how far this takes, and a policy that
> entirely forbids WG chairs from generating documents will just make
> sure that people competent in the subject can no longer be WG
> chairs.
The flexibility of the model for IETF working group management means
that it can be tailored to the situation. The problem with this is that
it requires careful thought and sometimes-painful candor. The
trade-offs are between management complexity versus group efficiency.
In one extreme, a small, cohesive group, which has deep understanding of
the work to be done, will probably function smoothly almost independent
of who is running the group. Having the chair be an author in such a
case usually goes well.
In the other extreme, a large, divisive group working on a difficult
topic needs extremely careful management. This means both the choice of
chairs and authors and they way they behave needs to be calculated with
attention to avoiding problems while also making forward progress.
Remember that the ultimate goal of an IETF effort is community /use/ of
what is produced. Merely getting a document published isn't the
important accomplishment. Having the community use it is. The ultimate
point behind rough consensus is to develop the support that makes that
use likely.
At the outset of a working group, these concerns require considering
what is known about the group that is likely to participate, what is
known about any current authors, what is known about potential chairs,
and what is known about the work to be done. Any indications of
technical complexity or political or personal challenges makes it
advisable to set up the working group with assignments that are more
pristine and possibly less "convenient".
In other words, greater working group risk requires narrower working
group management roles.
Changing authors, as a document is adopted, is merely one of the ways to
attend to these issues.
d/
--
Dave Crocker
Brandenburg InternetWorking
bbiw.net