RE: NEW ISSUE: Use of "Client" in 14.4

Maybe I wasn't clear -- the reason no one would do anything to their
code if we made the change is not that I believe they don't care about
HTTP/1.1 conformance, its that they already are compliant with the
suggested new wording (because the old wording made the intent clear
enough).

It's all about priorities -- until the HTTP/1.1 WG focused on just
getting three things done well enough, we were making progress at a very
slow rate, as everyone strove for perfection.

(Those three things: caching, keep-alive and the Host: header. They were
the minimum deemed essential to "keep the internet from melting down"
under the pressure of Web traffic using earlier HTTP versions.)

It was just another instance of "the best being the enemy of the good".

-----Original Message-----
From: Travis Snoozy (Volt) 
Sent: Thursday, January 04, 2007 1:13 PM
To: Paul Leach; ietf-http-wg@w3.org
Subject: RE: NEW ISSUE: Use of "Client" in 14.4

Paul Leach said:
> I don't think its worth making that change.
> 
> We need to be considerate of the people reading the updated spec and
> comparing it with the old one to see if they need to do anything. The
> diffs should be minimized, and since this one won't actually cause
anyone
> to do anything to their code, it might just as well be omitted.

... but that is then inconsiderate of anyone who is reading the spec
from 
scratch, is it not? And since this *does* center around a MUST-level 
requirement, I for one think it's very important to fix. Even though the

likely intent is clear to anyone who thinks about it long enough, what
the
spec *says* will cause clients that do the sane thing to be
non-conformant.

On the other hand, I don't think that anyone really cares if their 
client/server is "HTTP/1.1 conformant" at this point. I can't recall
anyone 
ever claiming conformance, and literal conformance is probably
impossible 
anyway (due to internal requirement conflicts). From this perspective,
we've 
got only an ad-hoc standard at best, and the MUSTs aren't really worth 
anything -- all that matters is "does my [client/server] work with 
[Apache/IIS/IE/Firefox/libwww/.NET]?"


Thanks,

-- Travis

Received on Friday, 5 January 2007 01:34:12 UTC