Re: I-D Action: draft-ietf-6man-ipv6only-flag-03.txt

[Philip Homburg <[e-mail]>]

>How many heard about this for the first time during that lightning talk 
>and had an informed idea about this? How many do even run IPv6 at all? On 
>wifi where this is most important probably?

This draft was also mentioned in the IPv6 WG. Typically there are quite a few
people going to IPv6 WG sessions, though of course it is hard to say
how many of them are actually running IPv6.

>What I would like to hear is from wifi experts, how much does the IPv4 
>broadcast traffic cause problems in large wifi deployments? Arena settings 
>for instance?

Maybe we can turn this around and ask the wifi experts to compare this draft
to the DHCPv4 option and ether type filtering to see if this draft actually
saves enough that mobile device vendors would actually implement it.

This draft contains the most vague hand waving: 
"It may result in an undesirable level [...]" What's the real number if
you employ ether type filtering?
"In particular, this may overload switches [...]" How is that ever going
to happen?

If this is a real operational problem, then maybe somebody can measure it,
write an article and a future draft can reference the article.

This draft is pointless if host operating systems ignore it. And right now,
ignoring this draft does seem to be the most sensible approach.