Re: i93: Repeating Single-value headers

* Henrik Nordström wrote:
>> I would like us to formulate some guideline beyond the charter when
>> handling of specific errors is considered in scope of this group's
>> deliverables (so we would consider adding text pointing out the mis-
>> behavior, or making recommendations how to handle it). There are
>> some known cases where errors are common in practise and affect e.g.
>> efficiency (such as sending malformed If-Modified-Since headers),
>How to handle malformed If-Modified-Since is already specified. Such
>conditionals is simply ignored.
>The fact that this leads to degraded performance to clients sending
>malformed If-Modified-Since headers giving them the same functionality
>as not using conditionals at all is not something the spec needs to go
>into. Should be obvious to anyone who analyses the problem with their

The example does not really have much impact on my point, but even if
you are right and implementations are required to ignore the field if
it is malformed, we'd still have an issue, because existing implemen-
tations do not ignore it, and sending malformed headers is so common,
that they are unlikely to change. The specific case I had in mind is
appending garbage to a legal value, like in

  If-Modified-Since: Thu, 22 Jun 2006 18:33:10 GMT GARBAGE
  If-Modified-Since: Thu, 22 Jun 2006 18:33:10 GMT;length=12345
  If-Modified-Since: Thu, 22 Jun 2006 18:33:10 GMT

As an example, Apache 2.2 treats these as identical when deciding if
it should respond with 304 (as it ignores everything after the time).
Björn Höhrmann · ·
Weinh. Str. 22 · Telefon: +49(0)621/4309674 ·
68309 Mannheim · PGP Pub. KeyID: 0xA4357E78 · 

Received on Friday, 28 December 2007 19:43:16 UTC