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


Help: OASIS Mailing Lists Help | MarkMail Help

ebxml-core message

[Date Prev] | [Thread Prev] | [Thread Next] | [Date Next] -- [Date Index] | [Thread Index] | [Elist Home]

Subject: Tag Languages, UID's etc.

Gentle People,

As has been noted a few times now, the UID is intended to lead to a wealth
of semantic information regarding the associated element, regardless of
specific tag name.  Yet we properly worry about providing some set of
(Oxford) English tag names, since our specifications include XML
definitions. (Sadly, these ebXML specifications DO NOT provide UID's for
these tags, or even account for UID in the DTD's)

It seems our ebXML teams think in terms of elements defined by specific XML
DTD's and Schemas, instead of thinking in terms of 'semantic entities'.  And
little (if any) thought has been given to how the semantic information
presmably pointed to by the UID is to be represented.  Had we been through
that exercise, we would have hopefully recognized that an XML DTD or Schema
in our documentation (e.g., the XML Header of an ebXML transport envelope)
was simply the end result of generation of a DTD or Schema for the 'business
model' that represents that header document, with the target language for
the DTD/Schema tags set to English.  Assuming the metadata included tag name
defintions in French, we could just as easily generate a DTD/Schema for a
the XML Header in French.  Furthermore, I would expect a conforming ebXML
processor of ebXML Header messages to accept and handle both the English and
the French versions of the XML header interchangeably.  That is, the
processor would ignore the tag names, and instead perform its processing
tasks based upon the UID's associated withthe tag names.  And of course, I
would expect ebXML applications to do likewise.  My English based purchase
order application should not care whether the tags names are in English,
French, UID-ease, or whatever.

Naturally, existing ebXML POC processors have hard coded tag name
recognition, since that's all they have to work with at this time.  Were
such implementations subjected to an ebXML confromance test, I would expect
them to encounter a non-English tag set and thereby fail conformance

Even though ebXML has given no more than lip-service attention to UID's to
date, ebXML applications and support software can conform, even if the UID's
are nothing more than placeholders for the semantic information.  They just
need to ignore the tags, and pay attention instead to the UID's those tags

        Bob Miller


[Date Prev] | [Thread Prev] | [Thread Next] | [Date Next] -- [Date Index] | [Thread Index] | [Elist Home]

Search: Match: Sort by:
Words: | Help

Powered by eList eXpress LLC