RE: case-sensitivity

This falls under general application robustness, particularly for server
implementers.  Whether it's written in the spec or not, inevitably someone
will build an application that works against server X but not yours
(client/server in-breeding) unless you are forgiving about minor details.

Here is my short list of deviations I have seen (including my collage work
on Fnord Web Server):
o Do not assume case on any value that is not explitly case sensitive for
functional reasons
o Do not assume a particular white space padding
o Do not assume anything about the formatting of paths (trailing slashes, /
or \, etc.)
o Do not assume real CR-LF pairs (accept CR or LF without the other)

Your millage may vary.

Brian Morin
Software Development Engineer
MSN Communities


-----Original Message-----
From: Greg Stein [mailto:gstein@lyra.org]
Sent: Saturday, May 27, 2000 1:58 PM
To: w3c-dist-auth@w3.org
Subject: case-sensitivity


I've been assuming that words such as "infinity" and "T" and "F" are
case-sensitive, using the case defined in RFC 2518. However, I just found
the following in RFC 2616:

  Section 2.1 (part of the Augmented BNF description):
    (for literals)
    ".... Unless stated otherwise, the text is case-insensitive."

In RFC 2518, Section 2, we state that WebDAV uses the same conventions as
HTTP/1.1.

Does that mean that the literals in RFC 2518 should be treated in a
case-insensitive manner by the server?

Cheers,
-g

-- 
Greg Stein, http://www.lyra.org/

Received on Saturday, 27 May 2000 18:38:54 UTC