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

Re: Large content size value

From: William A. Rowe, Jr. <wrowe@rowe-clan.net>
Date: Fri, 05 Jan 2007 18:24:22 -0600
Message-ID: <459EEC36.8060201@rowe-clan.net>
To: "Travis Snoozy (Volt)" <a-travis@microsoft.com>
CC: Scott Lawrence <scott@skrb.org>, Larry Masinter <lmm@acm.org>, "ietf-http-wg@w3.org" <ietf-http-wg@w3.org>

Travis Snoozy (Volt) wrote:
> William A. Rowe, Jr. said:
> 
>> There is no need, and significant arguments against defining a range for
>> DIGIT*1 as anything less than infinity, and it is up to the implementor to
>> decide what it can parse, what it cannot parse yet still work around, or
>> what it simply cannot work around.  Disconnecting, in the third case, is
>> the only remaining alternative...
> 
> Yeah, but the second case can be pretty ugly, given what the spec allows.

The second case, misparsing DIGIT*1 in this case, is your fault as an
implementor, and not permitted by spec.  DIGIT*1 is well defined.

>> So (s. 8.1.4 Practical Considerations) "A client, server, or proxy MAY
>> close the transport connection at any time" is the final word.
> 
> That's always an *option*, yes (but it's not the *only* option).

The other option is to leave the connection open and swallow the results.
Irrespective of how you [mis]parsed the headers, is there any other option
besides disconnect now, disconnect when you surrender (exceed some arbitary
number of bytes or whatever) or finish reading the response from the wire?
All three options are allowed by spec - it's your choice as an implementor.
The spec has no need or reason to dictate to you which you choose.

>> Can we please move on?
> 
> Sure. I have my answer, there are no new points here, and there's enough
> discussion for interested parties to draw an informed conclusion.

Excellent.
Received on Saturday, 6 January 2007 00:24:47 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Friday, 27 April 2012 06:50:00 GMT