W3C home > Mailing lists > Public > www-tag@w3.org > April 2002

Re: FW: draft findings on Unsafe Methods (whenToUseGet-7)

From: Paul Prescod <paul@prescod.net>
Date: Tue, 16 Apr 2002 02:11:45 -0700
Message-ID: <3CBBEAD1.A46036DC@prescod.net>
To: www-tag@w3.org
Mark Nottingham wrote:
> 
> Do most Web developers know the background behind safe/unsafe methods,
> idempotence, and understand the depths of REST, or do they just use POST
> when they want to hide the query arguments?

Most web developers know not to use GET for non-idempotent actions
because of the meaning of the browser reload button. They also know not
to use POST for things that they would like to have long-lived URIs
("bookmarkable"). That's the essense of the difference between the two
methods.

>...
> Of course there will be people who misunderstand SOAP, or limit their
> understanding to a subset of its capability. Why is that "fishy?"

What is fishy is that there is no interoperable core. Some people
dislike the RPC part of it (especially section 5 encoding) enough not to
implement it. Others implement *only* that and none of the "envelope"
and literal features. I'm talking SOAP implementors, not developers
using SOAP. 

Obviously this is fishy because of the interoperability implications,
but it is also fishy from a procedural point of view that a
specification designed to be a fish evolved into a mammal during its
standardization process. What does that say about the initial
requirements gathering? And if it is primarily intended to walk on land
now, should the thing still have gills?

 Paul Prescod
Received on Tuesday, 16 April 2002 05:15:43 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Thursday, 26 April 2012 12:47:06 GMT