Howdy,
I can't entirely agree with the argumentation, in part
because (and again this goes back to how the text first
appeared) metrics are useful to establish the state of
the system, whether to critique the state or simply
understand it. Appeal is only one form of critique.
That said -- I'm not planning to say any more on the subject.
It's not something I'm suggesting should stop the document,
either way!
Leslie.
John C Klensin wrote:
--On Monday, 31 January, 2005 14:00 -0500 Leslie Daigle
<leslie(_at_)thinkingcat(_dot_)com> wrote:
Howdy,
I'm a little concerned about hacking the appeals path on the
fly (i.e., dropping the IESG and going straight to IAB), but
I can live with that.
WRT this:
Harald Tveit Alvestrand wrote:
> - Removed the para about metrics. That's not part of this
section.
> Could go under "IAD responsibilities". But I think they're
not critical.
I think it should appear in the document. Either in the IAD
responsibilities (as you've suggested), or I can send more text
as to how it pertains to appeals. That is, it's easier to
avoid appeals when there is concrete data on the table, and/or
support them where appropriate.
I don't think people want more text at this point. So, perhaps
just putting it in the IAD section is the easiest path at this
time.
Leslie,
FWIW, in the spirit of Sam's note, I'd like to suggest that this
doesn't belong in the document at all, but in some informal list
of things we expect the IAD to do and will hold her or him (and
the IAOC) accountable if enough of it isn't done. That is the
"less text" rather than "no more" theme. I don't consider that
a showstopper one way or the other, however...
If it is to be put in as an IAD responsibility, it needs to be
clear that "better measurement" is _not_ more important than
getting the job done. My version of your comment about appeals
is that, if the job is being done, and done well, meaningful
appeals won't happen. If it is not being done, then the major
benefit of extensive data is either to help prove that it isn't
being done (which will probably be obvious) or to aid in telling
those launching the appeal to go away... the latter if
extensive, but irrelevant, data are collected. And I note that
the original text did not require relevant metrics,
interpretable metrics, or the like, just metrics. It shouldn't:
those terms are nearly meaningless without careful definitions,
and attempting to make those definitions would draw us into more
and more detail that doesn't belong in the BCP even if we could
agree on them. But "there will be metrics" don't necessarily
aid in getting the job done and may actually impede it.
john
_______________________________________________
Ietf mailing list
Ietf(_at_)ietf(_dot_)org
https://www1.ietf.org/mailman/listinfo/ietf