W3C home > Mailing lists > Public > www-ws-arch@w3.org > September 2002

RE: WSDL, app protocols, URI schemes

From: Heather Kreger <kreger@us.ibm.com>
Date: Sat, 28 Sep 2002 12:31:24 -0400
To: "Champion, Mike" <Mike.Champion@SoftwareAG-USA.com>
Cc: www-ws-arch@w3.org
Message-ID: <OFE6552700.DEBE6FF6-ON85256C42.0050046D@us.ibm.com>





+1

How eloquently stated!

I think we all understand the rest-ers are unhappy....
We're not sure what to do about it or exactly what would have to change to
make
you happier (I don't promise complete happiness).

As I said at the face to face - criticism doesn't help, criticism and at
least one
suggestion on how to fix it does.

I think I've grokked one or two things... (correct me if I'm wrong)
-A way to indicate that a message in wsdl is cacheable
-A way to specify a service instance reference in wsdl
-A binding for service reference to HTTP GET
      - this would make it cacheable perhaps?
      - this would make it hyperlink-able  perhaps?

I'm sure I'm much too simple minded to get this right as netting out to 2
or 3 things that are already
recognized as work to be done....

Is it as simple as working with the Description group to
add a tag 'cacheable' to messages (which would be mapped into correct
headers in bindings)?

Should we work with the Description group and ask for a service reference?
This is a topic that
they've talked about before. It is a requirement from the Grid folks. Have
you looks at how OGSA
is doing this? Is a solution like this sufficient for your RESTfull needs?
Maybe we can raise its
priority with them...

If we did these things, does it make it then so that Web services does not
preclude RESTfull
designs and implementations?  (if no, what else is required?)

I'm not sure that we can get to a place where all things Web services are
also always RESTfull...
but maybe we can get to a place where all things Web services could be
RESTfull. But even
that will only occur with constructive assistance.

I know you think it was all possible before Web services, but apparently A
LOT of companies
have decided that Web services makes it A LOT EASIER than before. Web
services is NOT
going away. You can help adjust it to suit both our needs, or you can
complain continuously
that its broken, and then live with (or ignore) whatever broken thing we
come up with. I am not
claiming that you are not engaged... I'm just saying that so far the
interactions have been
critical and complaining, and not very constructive. Apparently we are but
simple children in
'the one true way' and need to be spoon fed answers.

I'm sure I'm way out on a limb here.

Heather Kreger
Web Services Lead Architect
STSM, SWG Emerging Technology
kreger@us.ibm.com
919-543-3211 (t/l 441)  cell:919-496-9572


"Champion, Mike" <Mike.Champion@SoftwareAG-USA.com>@w3.org on 09/28/2002
10:19:49 AM

Sent by:    www-ws-arch-request@w3.org


To:    Mark Baker <distobj@acm.org>, David Orchard <dorchard@bea.com>
cc:    www-ws-arch@w3.org
Subject:    RE: WSDL, app protocols, URI schemes






> -----Original Message-----
> From: Mark Baker [mailto:distobj@acm.org]
> Sent: Friday, September 27, 2002 11:53 PM
> To: David Orchard
> Cc: www-ws-arch@w3.org
> Subject: Re: WSDL, app protocols, URI schemes
>
>

>
> Anyhow, talk is cheap.

Indeed!

>  Show me a problem that Web services claim to
> solve that the Web doesn't have a solution for.

The same can be said (proven?) about a Turing machine.  I can imagine how
the REST operations can be mapped onto Turings operations, or onto Codd's
relational algebra, thus I will believe from my little thought experiment
that you can solve any problem with the Web.  Your point is well taken.

But uhh, so what?  As you say, talk is cheap, thought experiments are
cheap.
But how many real software projects really build on the formal model of a
Turing machine? (Well, I guess those that base their designs on FSMs do so
indirectly ...)  How about "pure" relational applications?  To hear people
like C. J. Date rant, the database industry has strayed a long way from the
purity of Codd's vision.  [aside: There's a relational model zealot named
Fabian Pascal who started ranting about the evils of XML a year or so ago;
I
made the mistake of sending him a long (attempting to be conciliatory)
reply, and my stupidity, poor education, and lack of knowledge about the
computing industry was featured in his column for the next couple of months
:-) ]

My point is that I hope we can get beyond this "you really don't need all
this other junk, you can simply Do It The One True Way" stuff.  Dave
Orchard, and I, and surely  a bunch of others here have noted how much we
have learned from Dr. Fielding, directly or indirectly.  I think it's clear
from the TAG - XMLP discussions 6 months ago that people in the Web
services
community are listening.


> Deal?

No, you'd be taking my money!  I want something in return: some
constructive ideas about how to do things like ...

-- draw the Web services architecture diagrams in a way that is
REST-friendly
-- use the Web to manage the state of long-running multipart service
invocations
-- use RESTful ideas to coordinate asynchronous services
-- use HTTP to help manage web servies (you might have something to day
about that,
   Mark Baker, if you choose to share the expertise that you sell ...)
-- Apply the ideas that have proven themselves on the Web to Web services
so
as to
   help ensure their scalability, etc.
-- Apply REST ideas at a much more palatable level of abstraction so that
we
   don't ask applications developers to work at the level of
   (metaphorically
   speaking) "assembly language" or even "Turing machines."
Received on Saturday, 28 September 2002 12:45:59 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Tuesday, 3 July 2007 12:25:06 GMT