On 01/27/2013 11:19 AM, t.p. wrote:
----- Original Message -----
From: "Stephen Farrell" <stephen(_dot_)farrell(_at_)cs(_dot_)tcd(_dot_)ie>
To: <mrex(_at_)sap(_dot_)com>
Cc: "John C Klensin" <john-ietf(_at_)jck(_dot_)com>; "Thomas Narten"
<narten(_at_)us(_dot_)ibm(_dot_)com>; <ietf(_at_)ietf(_dot_)org>;
<adrian(_at_)olddog(_dot_)co(_dot_)uk>
Sent: Friday, January 25, 2013 10:47 PM
Hi Martin,
On 01/25/2013 09:36 PM, Martin Rex wrote:
I don't know about the last time it happened, but I know about
one time in Nov-2009 in the TLS WG (now rfc5746).
I recall that and agree with the sequence of events you
describe, but I'm not sure that that situation is
relevant when considering this draft, for two reasons:
Sounds a bit like 'don't confuse me with the facts, my mind is made
up:-)'
Absolutely not. Handling the TLS renegotiation bug was just
not the kind of thing that's relevant for this experiment.
It'd have been entirely dumb, stupid and crazy to try use any
kind of process experiment to handle that situation.
*After* this experiment is run, and if it turns up something
that we might want to make part of the normal formal process
then perhaps we might want to think about whether security
incident handling could be done that way or not. (I'd think
maybe not.) That'd happen in a year or so at the end of the
experiment if the experiment seems to have gone well. But,
that's not relevant now.
This IETF LC is about whether or not to run a process experiment
for a limited time that's optional to use at the discretion
of WG chairs.
This IETF LC is not about whether this or any proposal ought
be part of our formal, long-term process.
The point that Thomas made and John endorsed is that when we want to
speed things up, our current procedures allow us to do just that. We do
not need a formal process (more complications, more work). And as John
pointed out, having two independent Last Call discussions, on two
different lists, on communities that may have little overlap, is not the
way, IMO, to establish a clear consensus.
John's point about the two discussions in parallel was well
made and already acknowledged as such.
S.
Tom Petch
- First, that was the IETF in security-incident-handling
mode, and that's just different from normal process for
us, whether fast-tracked or not. Its in the nature of things
that the vast majority of security incidents don't
directly affect IETF protocols so as to require a
backwards incompatible change. So I think that was a
highly unusual case. (And let's hope things stay that
way.)
- Second, there was significant controversy within the
WG before the last calls, (with many hundreds of mails;-)
so a set of WG chairs that chose to try a fast-track
experiment in such circumstances would be crazy basically.
(Remember, we're only talking about an experiment here.)
As for Eliot's question, I don't recall any case when
a WG skipped WGLC. Even if its not part of 2026, right
now it's a de-facto but mandatory part of the process
as far as I can see. I'd be interested if there are cases
where WGLC was skipped, esp. if its been regularly done.
S.