Re: ISO/IEC 10646 as Document Character Set
Dan Connolly (connolly@w3.org)
Thu, 4 May 95 03:15:06 EDT
Glenn Adams writes:
> From: erik@netscape.com (Erik van der Poel)
> Date: Wed, 03 May 95 19:59:04 -0700
>
> So I'm asking if it's necessary to "negotiate" the absence/presence of the
> charset parameter, because I'm worried that older clients may not be able
> to handle content-type headers that include the charset parameter.
>
> There is already a means for such a negotiation; namely, the use of
> a Simple-Request vs. a Full-Request. A Full-Request request that its
> sender be capable of interpreting at least HTTP/1.0. Since the CHARSET
> parameter on the content media type is specified by HTTP/1.0, then
> the use of a Full-Request says that you must, at a minimum, be able to
> parse such this parameter. [Note that this doesn't mean you have to
> either interpret or make use of the parameter.]
>
> If a client is sending a Full-Request and cannot parse the parameter, then
> it is broken and should be replaced.
Hmmm... if all the WWW specs conflict with all the WWW implementions,
folks are going to continue to ignore the specs and read the Mosaic
2.4 code for "how it really works."
But this is an interesting idea. We could acknowledge that 1.0 clients
in fact do NOT support charset=, and only send charset= to 1.1 clients.
What else does a client have to support to be 1.1-happy? MD5 auth?
Anything else? What I'm after is: do we expect a lag until 1.1 clients
appear, or is the timeframe for 1.1 about the same as the timeframe
for internationalization in general?
Dan