W3C home > Mailing lists > Public > www-archive@w3.org > September 2003

Re: Few questions about REST

From: Sergey Beryozkin <sberyozkin@zandar.com>
Date: Thu, 18 Sep 2003 23:17:57 +0100
Message-ID: <003c01c37e32$c9c82eb0$6ba9a5c2@zandarpc>
To: "Mark Baker" <distobj@acm.org>
Cc: <www-archive@w3.org>

> > Almost, but not quite 8-).  What you're missing is that POST - even when
> > extended as you describe - cannot RESTfully be used to have any insight
> > into what happens after the data has been submitted.  "echo" semantics
> > are such that there exists an expectation that what is submitted, is
> > returned.  It's that expectation which is not RESTful.
I have to admit I'm somewhat confused. It seems to me that "echo" semantics
are only 'visible' to a client and a service
A generic intermediary would probably be confused because it's not aware of
the semantics.
Is it more RESTful to POST Person details first and then GET it back or ask
a service to POST it to a provided URI, instead of combining it into a
single POST ?

I think that those uses of doc-lit SOAP which actually do RPC can be
considered unRESTful (empty input bodies, or perhaps with some structures
clearly representing parameters, many fine-grained invocations, etc)

> > Here's one that I think is RESTful;
> >
> > http://www.w3.org/TR/soap12-part0/#Example

This is why I'm confused. This example seems to be very similar as the one
with echoPerson.
That example shows an input message, which is followed by a response asking
for some clarifications. How POST in this example can be used RESTfully to
see what happens after the data has been submitted ?

> > Hmm, are you clear what "uniform" means?  It means that it's meaningful
> > to all resources.
Hopefully yes :-)

> > But you snipped my comment about TCP.  Using SOAP over TCP, this would
> > be RESTful (assume it uses the SOAP encoding);
> > <e:envelope xmlns=...
> >   <e:body>
> >     <m:post>
> >       <foobar:data>lsdfiasdfasdf</foobar:data>
> >     </m:post>
> >   </e:body>
> > </e:envelope>
That's what I meant when saying that a uniform encoded operation will look
like an encoded document,
looks like a good-example of doc-encoded style (which is probably extinct by
now), even though 'm:post' is really a uniform operation

Sergey Beryozkin
Received on Thursday, 18 September 2003 18:18:38 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Wednesday, 7 November 2012 14:17:36 GMT