- From: Larry Masinter <masinter@parc.xerox.com>
- Date: Thu, 27 Jun 1996 13:35:31 PDT
- To: erik@netscape.com
- Cc: fielding@liege.ICS.UCI.EDU, http-wg%cuckoo.hpl.hp.com@hplb.hpl.hp.com
> 2. Data in character sets other than ("ISO-8859-1" or its subsets)...
Now that I think about it, that's a confusing way to say "either
ISO-8859-1 or US-ASCII".
>> In HTTP/1.1 _response_ messages, it is possible,
>> and will be recommended implementation advice, that for graceful
>> deployment a server might respond differently to a HTTP/1.0 request
>> and a HTTP/1.1 request.
>It would be nice if something like this was explicitly mentioned in the
>spec. Is it? If so, where?
I'll write something and send it around by Monday night.
> So you're suggesting that the server actually parse the output of the
> CGI program to check for charset, and, if absent, add charset? Are
> server implementors prepared to take this performance hit? Or are they
> already parsing CGI output for other reasons?
Somehow we have to manage the upgrade of HTTP/1.0 to HTTP/1.1 for
existing CGI programs. HTTP/1.1 places a number of new requirements on
responses in order to be compliant. Either the CGI programs will have
to themselves be upgraded and vetted, or else the servers have to
parse the output and validate it if they're going to be labelled
HTTP/1.1. I think this is a general requirement for protocol upgrades
when a gateway is involved.
This is an issue for any kind of plugin/API/CGI web server, isn't it?
Larry
Received on Thursday, 27 June 1996 13:44:26 UTC