Re: QUIC ossification

[Ted Hardie <[e-mail]>]

On Tue, Feb 19, 2019 at 7:52 AM Martin Thomson <[e-mail]> wrote:

>
> I completely get the don't segregate thing.  But to - I think - Mikkel's
> point about experimentation without permission, I do think that squatting
> is a perfectly sound approach in a space this big.  TLS people do it in a
> much smaller space.  As long squatters pick random values, they do so
> infrequently, and they get IANA registrations for codepoints relatively
> soon afterwards, the risk of collision is low.


The difficulty, of course, is that squatters use whatever process they like
for picking code points and it is easier to be lazy*.  One way out of this
is to have IANA generate random values for "unregistered" codepoints and
block those from appearing in the public registry for a period of time
(say, 6 months). At the end of six months they must either be formally
registered or dropped.  A slightly different, but probably similar effect
can be gotten by IANA generating random codepoints for a shifting
experimental range; they put out N code points at time T and age them out
and replace them at some interval afterwards.  That would, as a side
effect, give you a changing random pool that middleboxes would likely want
to track, rather than a static small number of versions.

Ted

*I have anecdata on this from ULAs but it is depressing enough that you
probably don't want to see it.