Individually-submitted proposed RFCs (was: Re: what is a problem)

[Thomas Narten]

> Perhaps writing it down will help.  But the description of the
> process as it actually works is as follows, based on a long
> discussion with the RFC Editor during the June-September period
> of this year.

> (i) People can, as you indicate, request publication by
> contacting the RFC Editor.  The IESG may have given new
> instructions in the last few months (the original ones were not
> documented, so why should new ones be? :-( ).  But, it you have
> not, the RFC Editor believes that it:

> 	* Has been instructed by the IESG to not process any
> 	such requests, or even review the documents, during an
> 	IETF meeting or the several weeks before and maybe after
> 	it.  That blackout implies around three months out of
> 	the year in which those documents are not examined, at
> 	IESG request.

This is news to me, and I suspect other IESG members as well. At
least, I don't know if any formal blackout. Sounds like some
clarification is needed as to whether the IESG and RFC Editor are on
the same page here with regards to expectations.

> 	* Has been instructed that any processing of
> 	IESG-originated documents is to be given priority over
> 	documents originating elsewhere (except, perhaps, from
> 	the IAB).  Lower-priority activities include review of
> 	submissions before they are sent to the IESG for
> 	"standards process coordination".

Do you mean s/IESG-originated/IETF WG-originated/ above? This may be
an important disctinction.

I think the above is the general point that WG documents should get
higher priority over non-WG docs, and I guess it follows that review
of individual submissions also gets lower priority. My recollection is
that this topic was originally prompted by something like a plenary
discussion where folks wondered why it took so long to get RFCs
published and why WG documents didn't get at least some priority over
non-WG documents. But I have no knowledge of the details of what
"prioritization" means in practice within the RFC Editor and how it
impacts reviews of individual submissions. Real data would be useful
here, to understand the degree of the problem.

> 	* IESG has strongly delivered the message that any
> 	document that is intended to be processed in a timely
> 	fashion will come through the IESG first.  I.e., any
> 	author who wants a document to be processed in a timely
> 	fashion will find a friendly AD and get the IESG to
> 	process the document and _not_ submit it to the RFC
> 	Editor.

I don't think the above was the result any formal IESG policy. I think
it goes without saying that if someone can find an AD who is
interested in a document and agrees to review it, the document
generally has smoother sailing through the IESG. Or, stated another
way, when the RFC editor sends the IESG a document, a first step is
often to identify an AD to look at the document (i.e., based on the
document's topic, which AD should needs to look at it?). If the author
does this step in advance, a step is eliminated.

I am not aware of any policy or desire to have stuff sent to the RFC
editor first, as a way giving it lower priority or any such thing. I
know that at least some ADs (hi scott) generally just tell folks to
submit directly to the RFC editor, but I suspect that has more to do
with making sure it gets entered into the process formally where it
can then be tracked. If you go back a year or more ago (long before ID
Tracker), if a submission went through the RFC editor, the RFC editor
would keep track of the document and remind the IESG periodically if
things had gotten lost. If an individual submission went directly to
an AD (before ID Tracker), there was no such record.

> Now, a summary of the above might be "Sure, you can submit a
> document directly to the RFC Editor.  But the IESG has now
> arranged it so the document then goes into a black hole _before_
> any timeouts are initiated and before the document enters the
> tracker".

AFAIK, there was never any intention that the above take place.

Thomas