From: Russ Allbery (rra@stanford.edu)
Date: Wed Jan 17 2001 - 13:59:17 CST
Brad Templeton <brad@templetons.com> writes:
> No need to get snotty.
I was being restrained compared to what I wanted to say. I find your
choice of presentation of your personal ethical preferences to be
offensive.
> Decoding tokens can be done any number of ways. The simplest is just a
> log file on the server with the token, message-id and real identity
> information.
If you don't see that in practice, in my experience, being able to tell
someone the origin of the problem and asking them to check a log file to
determine the origin of the problem is the difference between having the
problem be solved and having it continue, I guess I don't have anything
more to say on this specific point.
I will say this, however: I'm interested in standardizing and documenting
Usenet. Usenet is not a theoretical construction that we all get to make
up; it's already out there, already functioning, and is long overdue for
some standardization updates to reflect how it actually works.
Out there on that existing Usenet, there were 57,341 articles posted to
the misc.* hierarchy in the past 14 days or so. Of those 57,341 articles,
42,523 (74%) had NNTP-Posting-Host headers.
If you want to write up a standard for a new protocol called SuperUsenet
that has nice privacy features as part of the original design, has a PKI,
has all these other things that sound great, I'll gladly read that
protocol standard and see what stuff I can borrow from it and deploy on
Usenet.
If this working group ends up producing a standard for SuperUsenet, this
brand new communications protocol that isn't actually running anywhere and
hasn't actually ever been implemented... well then, I think we've all been
wasting a bunch of our time.
Similarly, if the dozen or so people who are active on this working group
decide that bits and pieces of Usenet that we don't like for whatever
reason were bad ideas and can easily be replaced, I think we have a
problem. If we write something that we think is better on paper and
publish it without any implementation to point to, without practical
experience with those ideas in real server installations, and without
regard to the reasons *why* people might be doing things we don't agree
with, then I don't think we're producing a standard. I think we're
producing a design document. Design documents are useful and interesting
things that don't get published as standards-track RFCs and that aren't
what I thought we were working on here.
But then, I've said things like this before.
-- Russ Allbery (rra@stanford.edu) <http://www.eyrie.org/~eagle/>