Re: [newtrk] Reviewing state of PS and DS standards

[Eliot Lear <lear at cisco.com>]

Dear All,

As a reminder, the decruft process was as follows:

1. Select a group of Proposed Standards from RFCs numbered less than 2000;
2. Present that list to an open mailing list that has been widely announced;
3. To remove an entry from the list, someone merely needed to send a note saying that a particular specification was still necessary for some reason.  We did NO validation of the reasons;
4. Present the trimmed list to other interested parties (e.g., rinse & repeat);
5. Last call to the IETF (e.g., rinse and repeat); and finally
   
6. Approval

Possible next steps include any or all of the following:

1. Go on to PS for RFCs 2000 to, say, RFC 3830, which brings us to two years prior to now, and approximately the same number of documents as the last time;
2. Take a look at DS for RFCs < 2000;
3. Develop a BCP for the bunch
4. First figure out what to do with the standards process and then circle back to this.

I prefer (d).

My primary reason for doing this was to allow the IESG to make true use of PS, such that if something less than useful slips through the cracks we can catch it in a next pass.  Before we advance any of the three options I would want guidance from the IESG that they would find these approaches useful.  I did so the last time based on an interest from the IETF chair who co-authored the document and hosted the old-standards mailing list. 

But my agreement with Harald had a second part to it, which said that we had to make forward progress on fixing the standards track.  Obviously Harald is not in the position he was.  However, I once again urge the IESG to recognize that RFC 2026 needs fixing.  We need to evolve as an organization.  It does not bode well that our processes have ossified so much in the face of a huge amount of evidence that they need fixing.

Eliot