- From: Gavin Thomas Nicol <gtn@rbii.com>
- Date: Fri, 29 Mar 2002 23:02:20 -0500
- To: www-tag@w3.org
On Wednesday 27 March 2002 06:47 pm, Paul Prescod wrote: > hyperlinking. Hyperlinking is a critical piece of any non-trivial > web service because it is the Web's way of managing state. While it is easy to prove that any RPC can be embedded in a URI, or that any continuation can be referenced via a URI, one has to ask whether this is a good thing.... simply because most URI's, especially in the face of arbitrary intermediaries, are globally scoped. It's unpleasant, but true that POST is often used for information hiding. This is in leiu of any other mechanism to do so. > SOAP actually works against the Web by discouraging people from > making data GET-able based on their intuitions. I think the fact that SOAP exists shows a failing in the ability to easily scope the set of available resources to a given user "session".... or at least that common practise (as established by CGI and promoted ever since) doesn't provide a conceptual framework for mapping such things into URIs well. It's pretty trivial to think of GET+all the headers and POST+all the headers and body as something that can be represented as a function call.... for many people, both return stuff, and so are roughly equivalent. The information hiding and bookmarking features give people a means to decide which to use. SOAP in HTTP is an extension of the desire for a general, and extensible means to make calls to server-side processing. Both GET and POST have enough problems that something else is desired, and as someone else noted, people have, for a long time, been rolling their own thing. You can't ignore either the user community *or* the vendor community. Both have already voted!
Received on Friday, 29 March 2002 23:14:00 UTC