RE: Harvesting REST

> -----Original Message-----
> From: www-ws-arch-request@w3.org [mailto:www-ws-arch-request@w3.org]On
> Behalf Of Paul Prescod
> Sent: Tuesday, July 16, 2002 9:38 PM
> To: David Orchard; www-ws-arch@w3.org
> Cc: 'Mark Baker'; 'Champion, Mike'
> Subject: Re: Harvesting REST
>
>
<snip/>

> >  Hence the
> > onload/onrequest actuation axis, etc.  Also why ranges (for
> drag'n drop in
> > gui apps) were added for xpointers.
> >
> > There was explicit acknowledgement that hypertext linking
> was different than
> > more general purpose xml linking.
>
> XLink does not make such a distinction: "A link, as the term
> is used [in
> XLink], is an explicit relationship between two or more data
> objects or
> portions of data objects." Nothing about user agents.
>

You dispute my recollection of history.  so beit.  I was a member of the
group when it did decide that hypertext linking is different than general
purpose xml linking.  But I don't want to spend my time trolling through the
w3c mailing lists to find the minutes of the meeting where the vote was
taken for justification, and I wouldn't be able to post any links to this
mailing list anyways.  If you insist, I'll dig it up.  And you ignored all
the aspects of xlink that relate to user agents.  If you take a look at the
references for the XLink spec, or it's requirements, you'll see that almost
every reference is for hypertext/media systems and not a single one from
programming language or database backgrounds.  Certainly shows the focus and
heritage.

> But is the terminology worth arguing over? Roy Fielding says that REST
> is appropriate to hypermedia. Mark Baker says that it is
> appropriate to
> web services. Even if you define hypermedia as belly dancing
> there is no
> contradiction between those two statements. In my experience, REST
> handles all of the sam web services scenarios that SOAP/WSDL does, but
> with a much higher degree of standardization and interoperability.
>

The terminology difference is crucial because MikeC, myself and others
assert that general application to application integration has different
requirements than hypertext applications.  And that much of the focus of the
web has been on hypertext, not on a2a integration.  Simply redefining
hypertext to include a2a does not solve the problems that face a2a
integration.  And I for one insist that we are able to talk about a2a
requirements and leveraging the technologies that are already within our
activity.

> It is not surprising that this is so: the Web itself must
> already handle
> situations where reliability is paramount, where security is vital,
> where asynchrony is required etc. etc. The only difference is that now
> computers are the user agents rather than people and computers are
> stupider so they need pre-parsed XML rather than HTML. We do
> not have to
> reinvent the architecture from scratch. We only need to slip
> in XML for
> HTML as we promised to do four years ago when we defined XML.
>

There's a big difference between humans using browsers to interact with an
application versus an application interacting with an application.
Hypertext is about user interactions.  And I've even pointed out that the
fundamental linking spec of XML of an acknowledgement of this.  The very
fact that neither the hypertext nor the data-oriented vocabularies use XLink
is important yet separate point.

This extended debate is EXACTLY why I would like to have the harvesting task
force focus on soap and wsdl.  The TAG has said that it is happy with SOAP
1.2 in the web architecture.  That is sufficient to indicate that SOAP 1.2
fits into the web architecture, so SOAP 1.2 and WSDL should be sufficient
for us to harvest to start a web services architecture.  Developers have
real problems with figuring out what the web services architecture is and
how to use and extend it, and we're not addressing their needs by endlessly
debating REST's utility.  We've probably blown our July deliverable schedule
that we promised in June.

Cheers,
Dave

Received on Wednesday, 17 July 2002 19:14:23 UTC