W3C home > Mailing lists > Public > ietf-http-wg@w3.org > October to December 1996

Re: HTTP response version, again

From: Larry Masinter <masinter@parc.xerox.com>
Date: Fri, 20 Dec 1996 21:32:21 PST
To: gjw@wnetc.com
Cc: dmk@research.bell-labs.com, http-wg%cuckoo.hpl.hp.com@hplb.hpl.hp.com, http-wg%cuckoo.hpl.hp.com@hplb.hpl.hp.com
Message-Id: <96Dec20.213221pst."2694"@golden.parc.xerox.com>
X-Mailing-List: <http-wg@cuckoo.hpl.hp.com> archive/latest/2152
The version number in the response indicates the capabilities of the
server. However, a server should not respond with a chunked transfer
encoding to a client that doesn't indicate it supports it. The
version number in the response can be ANY HTTP/1.x, but it would be
incorrect to send back a response that assumed HTTP/1.1 if the request
said HTTP/1.0.

# I think that the real issue here is we are using a single value to
# accomplish two objectives:
#   a.  Label the level of the response
#   b.  Declare the capabilities of the server

There is nothing wrong with this. In addition, we declare that 'a <=
b', that is, a response should be less than or equal to the
capabilities of the server. A response should also be less than or
equal to the capabilities of the request.

Since HTTP/1.x versions are upward compatible, you can tell the level
of 'a' by looking at the headers themselves.

Larry
Received on Friday, 20 December 1996 21:34:19 UTC

This archive was generated by hypermail 2.4.0 : Thursday, 2 February 2023 18:43:00 UTC