Re: CONTENT-ENCODING: FIXED revised proposed wording
Martin J. Duerst (mduerst@ifi.unizh.ch)
Wed, 23 Jul 1997 11:58:07 +0200 (MET DST)
On Tue, 22 Jul 1997, Jeffrey Mogul wrote:
> [OOPS. My previous message was missing one very important sentence,
> regarding when "identity" content-codings are implicitly allowed. -Jeff]
> (2) Replace section 14.3 (Accept-Encoding) entirely with
>
> 14.3 Accept-Encoding
> A server tests whether a content-coding is acceptable, according
> to an Accept-Encoding field, using these rules:
> (1) If the content-coding is one of the content-codings listed
> in the Accept-Encoding field, then it is acceptable. (Note that,
> as defined in section 3.9, a qvalue of 0 means "not acceptable".)
> (2) The special "*" symbol in an Accept-Encoding field matches
> any available content-coding.
> (3) If multiple content-codings are acceptable, then the
> acceptable content-coding with the highest non-zero qvalue is
> preferred.
> (4) The "identity" content-coding is always acceptable, unless
> specifically refused because the Accept-Encoding field includes
> "identity;q=0", or because the field includes "*;q=0" and does
> not explictly include the "identity" content-coding. If the
> Accept-Encoding field-value is empty, then only the "identity"
> encoding is acceptable.
A lot of what is written above turns up exactly the same in
Accept-Charset, Accept-Language, and so on. In many cases, there
is something special, but not too much. Is there a way to organize
the text different? This would hopefully a) save space, b) help
getting an overview, and c) make things more consistent.
For example, with respect to "*", the general description could
say that:
"*" in an accept field matches any value possible in that accept
field, including a default. A q-value given to "*" applies to all
values, including the default, but except those that are explicitly
present. An accept field may not allow "*", or may use a more
elaborate syntax for this functionality. Such cases are explicitly
mentionned in the description of the field itself.
Similar texts could explain the use of a default, the meaning of
q=0 (this seems already done to a certain extent), and so on.
For "Accept-Encoding", we would then just have to list the
possible values and identify the default.
Regards, Martin.