Re: XPATH query Take 2

[David RR Webber <Gnosis_@compuserve.com>]

Message text written by duane
>The users of ebXML will have to have XML parsers to participate in
ebXML.  It is not hard to write a couple of simple handlers to grab an
XML fragment from an XML DOcument instance.  THink of scalability.  If
GCA adopts ebXML with all 800,000 + of their member companies it will
cripple the Registry if they all ask for one document to be loaded into
the DOM and some processing done.<

>>>>>>>>>>>>>

Duane,

This is totally unrealistic.    Two or three years from now - if you are
lucky
you may see traffic going towards these extremes - and only against
certain pivotal registries.

What we need now is a workable first cut to get 10, 20, 100 larger 
companies up and running.  Believe me - that will take a year just
by itself.

Another thing - who said anything about the DOM?  You agreed with
JP about avoiding ALL talk of backend implementation detail - if
someone uses the DOM to do content parsing - they do so at their
own risk.  Other people will use other means.

We do need to precisely define the frontend service interfacing,
and make sure that is not overly complicated, and also how it
integrates into the schema or DTD parsing interactions.  That
currently is unclear as defined in our spec's as XML Parser is
not one of the actors in the use cases explicitly named.

DW.