W3C home > Mailing lists > Public > ietf-http-wg@w3.org > July to September 2012

Re: #241: clarify eval order/interaction of conditional headers

From: Roy T. Fielding <fielding@gbiv.com>
Date: Thu, 12 Jul 2012 11:38:26 -0700
Cc: Mark Nottingham <mnot@mnot.net>, HTTP Working Group <ietf-http-wg@w3.org>
Message-Id: <3C81CC72-F639-41CF-9405-2210A1549C0C@gbiv.com>
To: Julian Reschke <julian.reschke@gmx.de>
On Jul 12, 2012, at 12:08 AM, Julian Reschke wrote:

> On 2012-07-03 09:37, Mark Nottingham wrote:
>> <http://trac.tools.ietf.org/wg/httpbis/trac/ticket/241>
>> 
>> Roy documented the evaluation order for conditional that Apache uses (which, he pointed out, is the only logical way to do it, once you look at it):
>> 
>> <http://trac.tools.ietf.org/wg/httpbis/trac/ticket/241#comment:4>
>> ...
> 
> I'd like to avoid to include pseudo-code like that *unless* it's really needed.

So would I ... the text provided is just the comments from the code.
Once the result is agreed, we can work on text.

> Do we really need to define the evaluation order? In what real-world situations well there be more than one condition?
> 
> Wouldn't it be simpler to:
> 
> 1) Formally define the condition expressed by each header field,
> 
> and
> 
> 2) State that all conditions must be met, otherwise a 412 will be returned (with the 304 variant special-cased).

No, because it isn't true in practice (nor desired).  We actually
do want an etag to override last-modified even when IMS is false.

....Roy
Received on Thursday, 12 July 2012 18:38:53 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Thursday, 12 July 2012 18:38:59 GMT