RE: Working documents repository

["Spencer Dawkins" <sdawkins at nortelnetworks.com>]

I'm probably looking foolishly for too much consistency, but a recent thread
on this mailing list has been the difficulty of bringing new IETFers up to
speed, and what might be done about it (for example, see Charlie Perkins and
following posts on
http://www.ietf.org/mail-archive/ietf/Current/msg05005.html, "Generating
understanding about working group activities").

Preserving every version of every I-D, and the mailing list history, and
saying "figure it out", seems to say that the practice we use for catching
up with current work is so good that we should expand its scope to include
archeology!

If we have the opportunity to review an I-D and say "it's interesting, it's
impractical at this time for these reasons, it should be preserved as a
basis for future work" - this sounds like an OK thing. Publishing such an
I-D as an Informational RFC would provide an opportunity for this review. Is
this wrong?

Preserving every BAD idea the IETF has ever produced, with no
meta-information that says (for instance) "this protocol won't scale to the
Internet because it uses broadcast to reach all participants", sounds like a
less-than-OK thing.

I note that the best suggestions seem to require clueful editors to write
clueful summaries. Some of this is happening. Automatically keeping expired
I-Ds around would be easier, but in most cases would be just adding trees to
the forest.

Spencer