Re: A guide to the IETF's process documents

[Frank Ellermann <nobody at xyzzy.claranet.de>]

todd glassey wrote:

 [Internet Drafts]
> the issue was that I-D's were used as references for a process
> page without a process for permanently archiving them.

That's a tricky issue because I-D authors are (obviously) free to
drop an I-D, and (not so obvious) they can also ask that their
I-Ds are removed from the tools server. 

As an example, if I'd wish to kill my "???-options" draft I can
post a dummy (void) -02, after that the -01 is eliminated from
the "official" I-D repository.

Then I could ask Henrik to remove -00, -01, and -02 from the
IETF tools servers.  Maybe Henrik could say that he'll keep the
void -02 as tombstone until it expires.

After that I could ask Google who else has a copy of this I-D,
and try to eliminate them (what fun, it's more a theoretical
option).  IIRC John said he really sometimes posts I-Ds in the 
hope that they're ignored, because he can then claim that it's
a proven bad idea.  Admittedly I can't tell where he really did
that, maybe in some very obscure corner of newtrk / pesci.

But it's clear that authors are in theory free to withdraw I-Ds,
for good, bad, or ugly reasons.

> what all good I-D's do after six months and that is expire -
> making the page's links broken.

Yes, because xml2rfc uses links to the official I-D repository,
and arguably it should do that.  The tools server repository,
or other I-D collections, might need some way to delete I-Ds on
demand of te authors.  The official repository is automagically
prepared for the worst case "withdrawal". 

Another issue, if Brian references verion 01 of the BOF I-D in
the procdoc ION this doesn't automatically imply that he's also
happy with version 02 or later.

For the two cases you found (DISCUSS and BOF):

- The expired DISCUSS I-D is actually a kind of "grandfathered"
  ION, it's IMO unnecessary to fix anything about it before the
  ION experiment is evaluated-
- The BOF reference is automatically updated to the latest I-D
  whenever the procdoc ION is updated (unless Brian decides to
  remove it)
- the procdoc I-D (before the ION) had more informative links to
  unfinished I-Ds, e.g. I-D.carpenter-ietf-disputes-00 and maybe
  also the horrorshow-rescind I-D before it unfortunately failed
  in two Last Calls.  For the ION Brian removed all references 
  to remotely "controversial" stuff, only the BOF I-D survived -
  AFAICT nobody ever said a single word against this I-D.

Frank



_______________________________________________
Ipr-wg mailing list
Ipr-wg at ietf.org
https://www1.ietf.org/mailman/listinfo/ipr-wg