Dave Crocker wrote:
It uses Ned's language concerning restriction of changes to
draft-allman-dkim-*
Is that the following ?
| The working group will make only the minimal changes deemed
| useful to improve the viability of services that are based
| on these specifications
I prefer s/minimal/necessary/ and an option to add other I-Ds
(Murray's, Phil's, or William's) as further input documents.
I'm strongly opposed to both changes. Necessary raises the bar much too high
and we really need to focus on the current set of documents. Just look at how
many issues we already have open.
it states that the working group output will be
standards-track.
I don't see why it can't also create other proposals, it's
up to the IESG to decide about the final status. No way
to bypass IETF consensus for any WG output => no problem.
The issue is focus. Producing other documents is fine in theory, but in
practice likely to distract.
Remember, if we decide that other documents need to be produced the charter can
always be amended to say as much. Not only is a charter that says "may produce
some other stuff" increase the liklihood of rathole exploration, I suspect it
would not pass muster with the IESG, which has a pretty consistent track record
of rejecting open-ended charters. (It certainly would not have been acceptable
in my day, and not because of any objection of mine.)
| Keys will be stored in the responsible identity's DNS
| hierarchy
Apparently there are other ideas, maybe add "most likely".
Perhaps a little more wiggle room would be OK, but I think it needs to be
stronger than "most likely". I will also point out that your desire to allow
only necessary changes would effectively slam the door on a switch from
DNS-based keying, absent someone making a case that the DNS cannot be used in
practice (unlikely in the extreme).
Ned