CTG: clarifications needed / HTTP header fields, URI patterns

It seems that during editing something got mixed up in section 4.1.5 "Alteration of 
HTTP Header Field Values"


Section 4.1.5 states:
Aside from the usual procedures defined in [RFC 2616 HTTP] proxies should not modify
the values of header fields other than User-Agent, Accept, Accept-Charset and 
Accept-Encoding header fields and must not delete header fields. It must be possible
for the server to reconstruct the original User Agent originated header fields by
copying directly from the corresponding X-Device header field values (see
Original Header Fields).

The first sentence is formally equivalent to:
	proxies MAY modify header fields User-Agent, Accept, Accept-Charset and 
        Accept-Encoding and SHOULD NOT modify other header fields.

The second sentence states that one can reconstruct modified fields from equivalent
X-Device fields. However, such X-Device fields are only defined for Accept, 
Accept-Encoding, Accept-Charset, User-Agent. If other header fields are modified
because of the transformation operations of a proxy, it is therefore impossible to
reconstruct them. 

Besides, there is an unnecessary repetition of "header fields" in the sentence.

A clarification is needed.


The same section states:


The heuristics discussed in 4.2.9 Proxy Decision to Transform relating to URI 
patterns are not part of the decision to alter HTTP Header Field values.

What is the reasoning behind this prohibition? Deciding not to transform a request
because a URI indicates a mobile site is quite a valid approach -- actually more
efficient than first transforming the request and then figuring out whether the
result was mobile or not in the first place.

A clarification is requested.

E. Casais


Received on Friday, 19 June 2009 14:59:29 UTC