W3C home > Mailing lists > Public > ietf-http-wg-old@w3.org > January to April 1996

Re: Charsets revisited

From: Gavin Nicol <gtn@ebt.com>
Date: Tue, 23 Jan 1996 20:57:22 -0500
Message-Id: <199601240157.UAA10720@ebt-inc.ebt.com>
To: nms@nns.ru
Cc: http-wg%cuckoo.hpl.hp.com@hplb.hpl.hp.com
> 
>National characters are allowed in query, but nothind said
>about charset. Per section 5, full-request can be just
>request-line, thus charset information must be in URI
>itself (how?).
 
Well, arbitrary character sets should also be allowed in URI's, but
little has been said about this either, other than "it's a problem". I
have suggested a few mechanisms for supporting this.

As for queries, I think that Larry's multipart/forms-data is the best
way to go. Ideally, all forms data, including queries should be sent
to the server as a message body, which *does* have a way of labelling
the content. This also applies to the GET method.

>2. Accept-Charset (section 10.2)
> 
>Any rationale behind having no quality factor for charset?
>Is the order in the list significant? Is there any explicit
>charset?

I never noticed this, and agree that it seems strange.
Received on Tuesday, 23 January 1996 18:01:11 EST

This archive was generated by hypermail pre-2.1.9 : Wednesday, 24 September 2003 06:31:43 EDT