Re: Last Call: <draft-ietf-httpbis-http2-16.txt> (Hypertext Transfer Protocol version 2) to Proposed Standard

[Patrik Fältström <[e-mail]>]

> On 3 jan 2015, at 17:53, Eliot Lear <[e-mail]> wrote:
> 
> Second, even if we did handle versioning in SRV, there are known to be
> residential gateways out there that can't handle very many parallel DNS
> queries, and so we run into a loss problem.

If the future of the Internet is defined by the state of todays CPE's, instead of defining how the future of the CPE's should work, I think IETF is on the wrong track. Completely the wrong track.

Yes, these arguments do indeed come up now and then, and I every time stand up saying "be careful, be very careful".

On top of that I do not agree with your statement. More and more CPEs are linux based and run the same resolver code as anything else that is unix based, and if the world is to run DNS according to how DNS was defined in 1995 instead of how it is defined now, we can as well go back to HTTP 1.0 from 1.1 instead of going from 1.1 to 2.0.

I.e. I am much more concerned over the changes in the HTTP protocol than whatever requirements you put on DNS.


Regarding your other arguments, my main point was that even if there was a discussion in the wg two years ago, there is no trace what so ever in the draft about how DNS is to be used, so the questions will come back. Until there is a text. A text we can either agree or disagree on. But ultimately find consensus about.

   Patrik