Right. In this case, however, I think what we do is shift the "burden of
proof": If you want to register a header field in the "non-standard"
registry, that's all well and good, but the IETF may, by some consensus
process, put a note on your registration that says, "This is boneheaded;
don't use". What we want to do is get the bar lowered enough so that some
benign and widely deployed header field can be registered and documented
without having to wait for consensus on "the best way to do this".
Good idea. This matches what is done for non-standards track RFCs.
What happens is this: when someone wants to publish a non-standards
track RFC, the IESG is asked to look at it. The IESG can in principle
ask to put a note on the RFC that says "this is boneheaded; don't use".
In practice (at least while I was on IESG) the IESG spent a tremendous
time reviewing such documents and trying to craft language that explained
*why* a particular idea was boneheaded. because although the IESG
could also recommend "don't publish", the RFC Editor might publish it
anyway, and because many people don't bother looking at the status
of an RFC before implementing it. So the IESG occasionally spent time
crafting detailed explanations, but the RFC Editor wouldn't always
publish these intact, so there ensued a debate between IESG members
and the RFC Editor about it.
The time spent doing such reviews was time taken away from other,
far more useful activities. If the IESG had the option of saying
"don't publish this as an RFC" and having it stick, then it
wouldn't have been necessary to take up so much time explaining
why bad ideas are bad.
I don't see any reason at all that IESG should be spending much
time debating ideas that don't have a supporting constituency
larger than the set of authors of the document just in order to
keep it from looking like IETF is supporting their boneheaded ideas.
Similarly, I don't see why IETF should create a header registry
that makes it easier to legitimize boneheaded ideas than it is
to legitimize good ideas. That's exactly opposite of IETF's mission.
Header fields that are worthy of wide deployment will nearly always
take more time to develop than a boneheaded or naive approach - because
in order to make an idea worthy of wide deployment it is necessary
to consider a wider range of interests, and that takes time and effort.
Keith