To: new-httpd@hyperreal.com
Subject: Re: MSIE + byteranges 
In-reply-to: Your message of "Fri, 14 Feb 1997 04:42:31 PST."
             <Pine.HPP.3.95.970214043859.21949A-100000@ace.nueva.pvt.k12.ca.us>
Date: Sat, 15 Feb 1997 04:39:47 -0800
From: "Roy T. Fielding" <fielding@kiwi.ICS.UCI.EDU>
Message-ID: <9702150439.aa21480@paris.ics.uci.edu>
Sender: new-httpd-owner@apache.org
Precedence: bulk
Reply-To: new-httpd@hyperreal.com

>> Likewise, since you are now looking at the user agent anyway, it would
>> make more sense to look for "Mozilla/[23]" instead of Range-Request,
>
>No, it doesn't really, because of all the Mozilla clones. I don't mind
>sending Netscape 4 x-byteranges, even if it supports no-x. Doesn't
>*hurt* anything.

Sure it does -- it allows Netscape to say that we aren't speaking HTTP/1.1,
so they don't have to either.  It is far better to force them to fix
their bugs on each major release, and that is why UA-based hacks should
always include a version number.

The Mozilla clones can go suck an egg; they asked for bugwards-compatibility,
so that's what they get (including MSIE).

>> BTW, coding according to an example in the HTTP spec is no excuse.
>> If they haven't read the MIME spec (which was 1521 at that time),
>
>I haven't checked the dates, but 2047 is referenced in HTTP/1.1 -
>weren't 2045, 2045 and 2047 all finished at the same time?

Yeah, but the RFC editor changes the numbers to reflect that post-WG.
When we wrote the spec, the only non-draft reference was 1521.

....Roy