- From: Scott Lawrence <lawrence@agranat.com>
- Date: Tue, 28 Jul 1998 20:47:03 +0000
- To: Henrik Frystyk Nielsen <frystyk@w3.org>
- Cc: http-wg@cuckoo.hpl.hp.com
> At 19:20 7/28/98 +0000, Scott Lawrence wrote: > > >I believe that the MUST should stand; making it a SHOULD renders the > >Expect feature almost useless. Henrik Frystyk Nielsen replied: > I agree, but isn't this in fact the case in practice? Wouldn't it be > better to let Mandatory handle this as it has a stronger mechanism > for enforcing client based requirements based on the M- method name > prefix? ... but have you tested what happens with old servers (and CGI programs) if you send them new methods? At least in the case of CGIs they often ignore the method - in short, you have exactly the same situation faced by Expect, but more complex. Don't get me wrong - I like the Mandatory mechanism because it is so much more descriptive, but I don't think that it is any better from a backward compatibility point of view (and cannot be made any better). -- Scott Lawrence Consulting Engineer <lawrence@agranat.com> Agranat Systems, Inc. Embedded Web Technology http://www.agranat.com/
Received on Tuesday, 28 July 1998 13:49:24 UTC