W3C home > Mailing lists > Public > ietf-http-wg@w3.org > January to March 2007

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

From: Travis Snoozy (Volt) <a-travis@microsoft.com>
Date: Thu, 4 Jan 2007 13:28:43 -0800
To: "William A. Rowe, Jr." <wrowe@rowe-clan.net>, Paul Leach <paulle@windows.microsoft.com>
CC: "ietf-http-wg@w3.org" <ietf-http-wg@w3.org>
Message-ID: <86EDC3963F04D546BED8996F77D290F6049D118133@NA-EXMSG-C138.redmond.corp.microsoft.com>

William A. Rowe, Jr. said:

> > A strict interpretation of this would mean that other non-user-agent
> > clients (e.g., proxies) "MUST NOT" give an Accept-Language header in
> > their requests, because they can not (easily) give users a "choice of
> > linguistic preference". That could lead to all sorts of silliness (e.g.,
> > proxies stripping Accept-Language before forwarding) that very likely
> > isn't intended.
> Well, first off, you presume the spec replaces common sense of the authors
> of said proxy, but worse...

I think it's more egregious to presume that -any- author has common sense --
to me, that's just common sense ;).

> > I propose the following fix:
> >
> >   If the choice is not made available, then the Accept-Language header
> >   field MUST NOT be given in the request.
> not made available by ??...??  I presume you mean the end user of the
> proxy.

The prior sentence was omitted. Please elide omissions; taking that sentence
out of context does screw up the meaning, but that is an artifact of how the
original text was phrased. My proposed changes are not even in the clause
you quoted; they were in the omitted portion. The full proposal, again, was:

   As intelligibility is highly dependent on the individual user, it is
   recommended that <del>client applications</del><ins>user agents</ins>
   make the choice of linguistic preference available to the user. If the
   choice is not made available, then the Accept-Language header field MUST
   NOT be given in the request.

> But if the proxy can discern a locale by, say, the regional IP assignment
> of the end client, then that would be it's choice to 'fill in the gap'
> here. Or a special purpose proxy could very well prefer a specific
> language family based on it's anticipated user base.

None of these make the "linguistic preference available to the user". They make a choice, but the user is cut out of the loop (what if I use TOR, and get a German IP? Should the proxy rewrite my Accept-Language to be German, even though I want English?). Thus, none of the options you suggest conform to the spec.

> I'd suggest we please leave the existing text alone and leave it to the
> implementor to determine if and how Accept-Language should be presented.

... It's totally up to the implementer how to present Accept-Language options right now, and in my modification; I agree with you that this should be a choice on the implementer's part. But implementers of clients (proxies, user agents, or otherwise) do have to be present a linguistic preference choice -somehow- (according to the spec, right now) in order to transmit an Accept-Language header. This is the case even for things (like proxies) that don't have a reasonable way to present a choice to the user.

(I think I might be being unclear somehow; please, ask questions and I'll try to explain myself better.)


-- Travis
Received on Thursday, 4 January 2007 21:28:53 UTC

This archive was generated by hypermail 2.3.1 : Tuesday, 1 March 2016 11:10:41 UTC