OASIS Mailing List ArchivesView the OASIS mailing list archive below
or browse/search using MarkMail.

 


Help: OASIS Mailing Lists Help | MarkMail Help

 


 

   Re: [xml-dev] Web Services Framework Interoperability (Was Re: [xml-dev]

[ Lists Home | Date Index | Thread Index ]

On Thu, 31 Jul 2003 10:39:41 -0400, Chiusano Joseph 
<chiusano_joseph@bah.com> wrote:


>
> - Suppose a process based on GXA wants to interact with a process based
> on WS-CAF;
 ...

> Is this a concern for interoperability? It seems to me that some sort of
> middleware product that translates between the two frameworks might be
> required.

Sure, it's a concern for interoperability.  The only 
Recommendations/standard profiles in this space are SOAP 1.2 (which is not  
widely deployed) and the WS-I interop profile (which is quite minimal).  
Also, it's pretty clear that the major vendors are not "playing nicely" to 
promote interoperability but are using the "standards" 
process/organizations as venues in which to pursue the same competitions 
that are going on in the courts, the analyst-for-hire industry, the trade 
press, etc. (hmm, sounds like von Clausewitz -- "Standards are the 
continuation of industry competition by other means.")

The competing stacks and ad-hoc consortia specs are fine *if* we think of 
them as experimental explorations of the possibilities that are opened by a 
common  conception of what "service oriented architectures" are all about 
and HTTP/XML and SOAP/WSDL offer as foundations.  But it is extremely 
dangerous to think of these things as "standards" on which interoperability 
can be expected.  That's the path to premature standardization of unproven 
technologies, vendor lockin, and all sorts of other evils.   It's important 
for consumers to demand standards conformance, but equally important to 
accept the principle that no "standard" should be promulgated before its 
time.

The more sensible path IMHO is to treat GXA specs AND  alternatives from 
other vendors and consortia as quasi-proprietary technologies that may add 
value in specific situations, but end users who value their ability to 
evolve and choose different vendors will NOT deeply architect them into 
their applications.  If they prove their value to solve real business 
problems in a conceptually efficient manner, then it's time to think about 
standardization and interop.  In the meantime, it's probably BETTER for 
end-users to use a "kludgy" middleware / transformation pipeline solution 
to achieve interop than to demand conformance to proprietary "standards" or 
premature standardization by OASIS/W3C/whoever.

[Speaking only for myself, blah blah blah]




 

News | XML in Industry | Calendar | XML Registry
Marketplace | Resources | MyXML.org | Sponsors | Privacy Statement

Copyright 2001 XML.org. This site is hosted by OASIS