Re: [ubl-sbsc] Re: [ubl-hisc] Naming of the XML expression of XPath information

[<stephen_green@seventhproject.co.uk>]

> I won't go as far to say to get 
> rid of the XML format/instance, because I'm sure others have good uses 
> for them.

How about as a format for 'normative' definitions of a subset? 
EDI obviously uses text for codelists etc but this seem a little
weak to me and obviously XML wasn't around twenty years ago when
EDI might have used it.

It seems to be easier to thoroughly define XML with encoding, schema
and the exact 'tags' you need to be unambiguous and it so easily
allows the normative definition of a subset together with the necessary
uri(s) attributed to it. In particular, I guess one can't provide a 
schema to thoroughly define the text or html formats. Psychologically
too I guess XML formats are what developers would be expecting and
prepared for.

Of course the other quality of the XML format is to allow extension
with extra data, etc as and when needed (e.g. will there be UBL 1.2?).

It seems to get XML to text (and of course to HTML) is easy but not
so easy to get back to a standard XML format without the schema if
it is lost

In short I guess I'm saying I'd be very nervous about making the XML
secondary to the text and html formats and even more nervous about
providing a normative potential standard just as text or even more so
as html (which I'd regard as less robust). Yes EDI does it but surely
we're moving away from that now for the above reasons and probably
others.

All the best

Steve