Re: Case (in)sensitivity in (Non-)Compliance headers

Ross Patterson writes:

    The examples imply that compliance-namespace is to be treated in a
    case-insensitive manner, although that isn't stated anywhere.  I
    hate to keep harping on this issue, but living in the mainframe
    world where upper case tends to be the default for such things, I
    keep finding examples of case-insensitive tokens being handled as
    if they were always lower case (e.g., media types in HTTP
    Content-Type headers by at least one of the Big Two browsers).

If you look at RFC, section 2.1 (Augmented BNF), you'll find this
definition:

    "literal"
	 Quotation marks surround literal text. Unless stated otherwise,
	      the text is case-insensitive.
    
I.e., anywhere in the HTTP specification, if the BNF includes a
quoted literal, that literal is supposed to be case-insensitive.

However, since I didn't give a formal BNF for the compliance-namespace
values, and the set could be extended in the future, I agree that it's
worth stating this (case-insensitivity) explicitly.

And I'll also include a BNF for the RFC and HDR namespaces.  (Per
Henrik's comments, I'm removing the "PEP" namespace from the next
draft.)

Thanks
-Jeff

Received on Wednesday, 30 July 1997 12:30:53 UTC