Re: Proposal for Describing Web Services that Refer to Other Web Services: R085

On Sat, Apr 26, 2003 at 04:55:59AM -0400, Sanjiva Weerawarana wrote:
> 
> "Mark Baker" <distobj@acm.org> writes:
> > The current approach, as reified in many Web services specs
> > and this wsdl:binding proposal, is akin to;
> > 
> > <phone>+16132864390</phone)
> 
> WS-Addressing's approach is a bit different. Its roughly like:
>     <phone>
>         <number>+1613...</number>
>         <international-access-#>011</international-access-#>
>         <outside-access-#>9</outside-access-#>
>         ..
>     </phone>

Well, that other information (other than "number") isn't identifying
information for the terminal.  The internationalized phone number
remains, by itself, sufficient information for a call to be made in
the context of the open phone system (i.e. after you get pass your PBX),
and with an up-to-date client which supports them.

Another way of looking at what's going on here is that the current Web
services approach creates a competing identifier syntax to URIs;

"tel:+16132864390"

versus

"<phone><number>+16132864390</number><phone>"

The issue is, there's a spec which describes the syntax of the former
(RFC 2396), but not the latter; a spec which identifies one string as an
identifier, separate from other data).

> Basically an endpoint reference in WS-Addressing is spsed to have all
> the data you need to get to the other end. The actual "address", a URI,
> in indeed required, but in many cases additional info may be needed
> for the recipient to really understand a reference.

Putting this additional info in the URI would be a Good Thing, as it
would permit late binding.

MB
-- 
Mark Baker.   Ottawa, Ontario, CANADA.        http://www.markbaker.ca
Web architecture consulting, technical reports, evaluation & analysis

Received on Saturday, 26 April 2003 23:22:23 UTC