Date: Sat, 28 Feb 1998 19:20:46 -0800 (PST) From: "Gregory J. Woodhouse" <email@example.com> To: Larry Masinter <firstname.lastname@example.org> cc: Yaron Goland <email@example.com>, Josh Cohen <firstname.lastname@example.org>, email@example.com, firstname.lastname@example.org, email@example.com, Alex Hopmann <IMCEAEX-_O=MICROSOFT_OU=NORTHAMERICA_CN=RECIPIENTS_CNfirstname.lastname@example.org> Message-ID: <Pine.BSF.3.96.980228191201.23058Bemail@example.com> Subject: Re: draft-cohen-http-ext-postal-00.txt Yaron's statement is not a tautology. His point is that requests can vary along three dimensions: the URI, the method and the request header. If you commit yourself to the use of POST then you are not free to vary the URI because the URI identifies the resource that processes the request. With PUT and DELETE the URI identifies the resource to be modified, not a generic processor such as a CGI program or something similar. Personally, I happen to agree that in the case of tunneling a protocol though HTTP it is at least more intuitive for the resource to be a target rather than a processor of requests --- Gregory Woodhouse firstname.lastname@example.org / http://www.wnetc.com/home.html May the dromedary be with you.