[dns-operations] Fixing the last Internet protocol which mixes service and server, HTTP (Was: Adding CNAME for the root domain issue

[Andrew Sullivan]

On Thu, Apr 28, 2016 at 06:36:58PM +0200, Stephane Bortzmeyer wrote:
> On Thu, Apr 28, 2016 at 12:01:28PM -0400,
>  Andrew Sullivan <ajs at anvilwalrusden.com> wrote 
>  a message of 17 lines which said:
> 
> > Including SRV in HTTP/2 was in fact proposed and rejected.
> 
> Do you happen to have a pointer to a documentation for this rejection?

That does make it seem a little more official than I intended it to be
:)  I recall the discussion of the proposal, and there was a lot of,
"No way," and that is all I remember.  The WG had a pretty good use of
the tracker during the development of version 2, however, so I imagine
it's in there.  Here's a list message from Mark Nottingham noting the
decision, anyway:

http://lists.w3.org/Archives/Public/ietf-http-wg/2015AprJun/0674.html

> Documenting rejected proposals would help IETF avoid the syndrom of
> "idea proposed every three years and rejected again".

I think you are perhaps optimistic about the tendency of people to
read the documents :)

A

-- 
Andrew Sullivan
ajs at anvilwalrusden.com