You can read the following article by Mark at:
http://www.infoq.com/news/2009/03/burke-rest
Gervas
<<After the release of RESTeasy 1.0
<http://java.dzone.com/announcements/resteasy-10-ga-released>, Bill
Burke spoke with Jack Vaughan on how he sees REST comparing to WS-*
<http://itknowledgeexchange.techtarget.com/soa-talk/>. In the
introduction to the article, Jack starts by giving his views on the
actual name: 'Web Services':
The 'Web' part of Web services was different, somewhat exploitative.
The Web was a popular success, and you have to imagine someone
thinking that if they named the latest software architecture after
the Web, good things might happen. It wasn't a big reach; Web
services did tend to use the Web's bread-and-butter protocol, HTTP.
In Jack's opinion (shared by many
<http://www.infoq.com/news/2009/03/www.infoq.com/articles/sanjiva-rest-myths>
if previous InfoQ articles
<http://www.infoq.com/news/2009/03/www.infoq.com/news/2007/12/rest-ws-payback>
are anything to go by <http://www.infoq.com/news/2008/06/whoa-woa>) some
feel that REST is "truer to the spirit of the web than classic Web
services employing XML and SOAP". According to Bill (and again, many
others <http://www.infoq.com/articles/rest-introduction>) "The value of
REST architecture is that it takes better advantage of Web architecture"
and that the reason behind this is rediscovering HTTP and "trying to
understand how the web becomes so prevalent."
Bill goes on to target the WS-* standards, saying that they're too much
of a moving target:
Getting vendors to cooperate is hard -- ask Apache.
(Let's ignore the fact that all standardization is hard and as a result
takes time and effort
<http://qconlondon.com/london-2009/speaker/Paul+Downey>, including Java
<http://jcp.org/en/home/index> and the work behind
<http://www.w3.org/TR/WD-http-pep-960820> HTTP
<http://www.w3.org/Protocols/HTTP/Issues/http-wg.html>, or that efforts
such as WSTF and Stonehenge <http://www.infoq.com/news/2008/12/wstf>
show that Web Services vendors are willing to cooperate inside and
outside of the successful standards bodies such as W3C and OASIS
<http://www.infoq.com/articles/ws-standards-wcf-bustamante>.)
But back to Jack, who says that irrespective of the struggles in
standards suffered by WS-* (ignoring the fact that it hasn't exactly
been plain sailing for the WWW <http://www.w3.org/>) HTTP has "forged
on" so that now everyone has some kind of HTTP support within their
infrastructure, meaning that we don't need to worry about that aspect of
interoperability. But ...
REST forgoes certain levels of interoperability, but that may have
its advantages.
And Bill agrees:
What is cool about REST is you are focused on straight http. So
instead of worrying about interoperability between vendors...you
worry about interoperability between applications. You let http do
the heavy lifting.
To conclude, Bill says that REST isn't anti-SOA (again something that
others
<http://www.infoq.com/interviews/robinson-rest-ws-soa-implementation>
would definitely agree with
<http://www.infoq.com/presentations/mark-little-soa-rest>), although it
is "anti-WS-Star ... and SOAP".>
