My original note (please see below) was intended to address Ms
Foerster's argument, that one of XML's features would likely confound
the difficulties inherent in information exchange rather than alleviate
them. I attempted to suggest that one feature to which Ms Foerster drew
attention (independent tag selection), when taken together with certain
techniques established as an aspect of HyTime (that is, ISO 10744:1997
-- Hypermedia/Time-based Structuring Language)[1], as described in the
chapter "A.3 Architectural Form Definition Requirements (AFDR)" under
the name of "enabling architectures"[2], would permit the formulation of
automated techniques to perform the requisite translations,
transliterations, transcodings, and/or transformations. The techniques
were developed in relations to SGML, but have been demonstrated to be
just as applicable to XML. XSL, as alluded to by Mr Kammerer, can be
used to effect equivalent transformations.
These techniques, and similar ones not yet developed, which take
advantage of the meta-information intrinsic to XML-related standards,
permit interchange mechanisms which alleviate the noted problems rather
than exacerbating them.
I made, and make, no attempt to address the larger process issues.
Neither transactional and workflow aspects of the automated processes,
nor the political and social issues concerning reaching semantic
agreements. My note neither implied nor stated that "it should just all
work." It stated just that original message implied a conclusion which
was incorrect.
I've spent my time with COBOL. Don't go there. That one can is the
limiting case for that one should.
[1] HyTime: http://www.oasis-open.org/cover/hytime.html
[2] Enabling Architectures:
http://www.ornl.gov/sgml/wg8/docs/n1920/html/clause-A.3.html
--------------------------------------------------------------------------------
"William J. Kammerer" wrote:
>
> David RR Webber chided Rachel Foerster: "I believe what James [Anderson]
> was trying to tell you was that the W3C Schema has this all solved
> already with their datatyping and structure methods based on an OO
> approach. All we have to do is get everyone using this consistently and
> the parsers upgraded and it should just all work."
>
> Rachel retorted "...you know as well I as that the W3C has not yet
> approved an XML Schema Recommendation."
>
> Dear David:
>
> Nobody is arguing that XML isn't an "enabling architecture," to use
> James' words. I assume, as Rachel may have, that with whatever XML
> schema is eventually birthed by the W3C we'll pretty much be able to
> describe and constrain the syntax of any XML document. And with XSLT, we
> can pretty much transform anything that looks like XML into something
> else that looks like XML. I guess that's what James meant by "enabling
> architecture." But, so what? COBOL and flat files can do all that
> stuff already; so can EDI and translators (or COBOL programs).
>
> In any case, "it should just all work" simply ignores the effort that
> goes into getting two or more organizations interested in a message
> domain to agree on the formats, tags, structure, etc. to convey
> agreed-upon business semantics. Rachel pointed out ebXML, but the same
> give and take (and politics) goes on in the vertical groups like OTA,
> RosettaNet, etc. etc. Once you have two or more organizations that need
> to talk, some necessary sclerosis sets in and you can no longer move at
> "Internet speed."
--------------------------------------------------------------------------------
my original note:
james anderson wrote:
>
> The argument below is spurious. XML with schema supports value domain
> specifications. It also supports tag specifications. XML with "enabling
> architecture" techniques also allows automated tr[a]nslation between tag sets.
>
> Rachel Foerster wrote:
> >
> >
> > Lastly on the issue of using XML -- since XML is extensible that means that
> > the originator of the XML document gets to choose the tag names. Just
> > imagine if you would, a situation where each provider generating a health
> > care claim chose not only their on tags but data content along with the data
> > attributes. We'd be even worse off then now, with only (!) 450 +/-
> > proprietary formats!
> >
------ XML/edi Group Discussion List ------
Homepage = http://www.XMLedi-Group.org
Unsubscribe = send email to: [EMAIL PROTECTED]
Leave the subject and body of the message blank
Questions/requests: [EMAIL PROTECTED]
To receive only one message per day (digest format)
send the following message to [EMAIL PROTECTED],
(leave the subject line blank)
digest xmledi-group your-email-address
To join the XML/edi Group complete the form located at:
http://www.xmledi-group.org/xmledigroup/mail1.htm
