ietf
[Top] [All Lists]

Re: Proposed IESG Statement on the Conclusion of Experiments

2012-04-21 10:41:48
Hi -

From: "Ronald Bonica" <rbonica(_at_)juniper(_dot_)net>
To: "Randy Presuhn" <randy_presuhn(_at_)mindspring(_dot_)com>; 
<ietf(_at_)ietf(_dot_)org>; <wgchairs(_at_)ietf(_dot_)org>; "Brian E 
Carpenter"
<brian(_dot_)e(_dot_)carpenter(_at_)gmail(_dot_)com>
Sent: Saturday, April 21, 2012 7:44 AM
Subject: RE: Proposed IESG Statement on the Conclusion of Experiments
...
The proposed IESG statement encourages community members to terminate
their own experiments when they have clearly ended. In many cases, the
experimenter moves on before the experiment is terminated. In that case,
the cleanup chore is left to others.

I think there are some  fundamental problems in this framing of the question.
The use of the phrase "their own" and "clearly" are the ones that concern me.
The problems with "clearly" are obvious, particularly for protocols that have
a long legacy of "just working", regardless of their theoretical shortcomings.
"Their own" assumes a level of "ownership" and, more importantly, *control*
which is rarely true, and in some cases the "owners" may actually have a
conflict of interest leading them to favor premature termination.

The proposed IESG statement *does not* seek to expedite the termination
of any experiment. Its only goal is to identify experiments that clearly have 
terminated.

I do not see any way in which this improves the criteria or process for 
determining
"clearly", or even the subtler problem of "terminated."  Sometimes the result of
the experiment may be "we don't want to standardize this", but also "in the
aftermath of the experiment, there are a lot of implementations of this, and
insufficient motivation for those who have deployed it to move to something 
else."

Randy


<Prev in Thread] Current Thread [Next in Thread>