Re: A proposal for publishing nominee names

[Dave Crocker <dcrocker@brandenburg.com>]

At 04:51 PM 7/13/2002 -0400, John C Klensin wrote:
> <sarcasm>

Since the point of my previous note seems to have been missed, let's try 
backing into this more pedagogically:

1.      The IETF has had a long-standing policy of maintaining intense 
privacy about the names of candidates.  The current proposal is therefore a 
replacement of a well-established -- and reasonably successful -- paradigm, 
with one that is fundamentally opposite.  Folks in the IETF know something 
about the risks involved with paradigm changes.

        Public nomination lists are quite common in the world; however 
many things are common in the world that the IETF has explicitly chosen NOT 
to adopted.  In other words, we have our own perspectives and styles of 
doing process.  The track record for the success of our model is marked, so 
we should try to maintain the foundation for it, as much as possibly.

        Blithely replacing a key component to the paradigm we have used 
for choosing our managers is simply not prudent.  A claim that we can "try 
it once" and then see how it goes is like saying that we should try to use 
LDAP instead of DNS, for awhile, and see how it goes.  We need to adopt 
changes based on a reasonably careful consideration of benefits, risks, and 
adoption process.  None of that is really happening here.

        This community is extremely expert at making technical changes, 
and yet we are always very careful about making changes to the Internet's 
technical infrastructure.  This community is very much NOT exert at making 
changes to social processes, so we need to be even more careful about 
making changes to the IETF's management processes, not less careful.

        Let's try to make the smallest change to our selection paradigm we 
can and let's be very clear about the problem that change is intended to solve.


2.      The proposal for this new paradigm introduces the change at the 
point of maximum scaling impact.  Folks in the IETF know something about 
scaling risks.

        As I said, "we like to think that we are free from such 
stuff."  That's different than saying that we ARE free from it.  To the 
extent that we are willing to be pragmatic and acknowledge that we live in 
the real world of people who have a variety of motives and styles, then we 
should factor such concerns into the design of our social processes, 
especially a social process that chooses our managers.

        Frankly, the change being proposed invites a substantial amount of 
public discussion, debate, abuse, and the like, for the entire duration of 
the nomcom process.  That's not just my opinion.  There is a fair amount of 
research on this sort of gaming and anyone trained in the topic will make 
the same comment about it.  They will also note that early vs. late 
disclosure is likely to make a significant difference in the nature and 
degree of that gaming.  We would do well to worry about making a change 
which affects the scaling of such gaming opportunities.


3.      The proposed change is being made without a clear sense of the 
benefit being sought, or real consideration of the likelihood that benefit 
will accrue.

        (To repeat, I think that some disclosure of names will permit a 
increase in community feedback and that this is likely to provide better 
insight into serious candidates. I doubt it will have any positive effect 
for evaluating basic candidate qualifications and it will definitely 
increase the load on nomcom members.)

        There also has there been no consideration of possible negative 
effects of the proposed disclosure.  Folks in the IETF know that changes to 
systems almost always carry unintended and negative consequences, so we 
would be wise not to ignore the probability.  The effects of changes to 
social processes are far more difficult to predict than changes to 
technical systems; we should therefore approach such changes with more 
caution, not less.

        My previous note suggested the alternative of publishing the names 
only of those making the final evaluation cut, rather than of all those 
initially nominated.  I gave reasons for that alternative suggestion

        Assuming that anyone has interest in serious consideration of 
needs, effects and alternatives, I'd suggest considering the trade-offs of 
such choices, rather than indulging in simplistic, dismissive sarcasm.


d/

----------
Dave Crocker  <mailto:dcrocker@brandenburg.com>
Brandenburg InternetWorking  <http://www.brandenburg.com>
tel +1.408.246.8253;  fax +1.408.850.1850