I'm writing this because I think there may still be a lack of clarity on
why various people are insisting on various things, and we can't just do
something simple.   I don't think that the reasons for each of the things
that have been proposed is clear to everybody--I hadn't really thought
about the reason I'm giving for Step 1 until just now, but I suspect this
reason is clear in the minds of Terry and various other people in
leadership roles.   I would appreciate being corrected if I have failed to
express it correctly.

Step 1: officially get ".home" out of 7788 by updating it
Reason: right now if someone releases software with ".home" in it, it's the
IETF's fault, because the IETF is explicitly recommending this.   This is
why "do nothing" isn't an option.

Step 2: recommend something else and get it through 6761 process
Reason: people will still use ".home" and blame the IETF if they get in
trouble, because we haven't said to use anything new.

Step 3: explain how this stuff works.
Reason: right now our story for how naming works on homenets is sketchy,
and I would argue is not likely to provide a happy user experience.
Existing implementations by participants work and interoperate, but it is
not clear that this is a consequence of a good spec, or whether it's
because the implementors are talking to each other.

We cannot assume that this happy situation will continue as router vendors
who are not participating directly in the discussion begin to produce
homenet routers.   I think this isn't likely to happen soon anyway, so we
have time to address this problem, but we need to start working on that
now, not later.   It is not the case that what we have now is good enough.
_______________________________________________
homenet mailing list
[email protected]
https://www.ietf.org/mailman/listinfo/homenet

Reply via email to