[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: A Call for Dialogue on XML Schema Part 1 and 2
- From: Rick Jelliffe <ricko@allette.com.au>
- To: Xml-Dev <xml-dev@lists.xml.org>
- Date: Tue, 13 Mar 2001 10:28:52 +0800
From: Eric van der Vlist <vdv@dyomedea.com>
>This is better than I feared, but even so, why couldn't I use a
>different date format or decimal separator in the transfer syntax ?
>Because it's a bad practice for international exchange formats doesn't
>mean it's a bad practice in all the other cases.
To be clear, it is not a universally-accepted principle of
internationalization
that data should be sent in some transnational format. To the contrary,
there is wide-spread belief that there is nothing wrong if it is easier to
send data across the room (i.e. in a locale-specific format) rather than
across the world (i.e. in a transnational format.) The use of
transnational
formats increases the work that a PDA must do. In a 3-layer system,
it is not inappropriate for the database to use transnational formats,
the client to use localized formats, and the middleware to do the
conversions.
So, from that perspective, XML Schemas is biased either to be the back-end
schema language (database <-> middleware) or against lightweight clients. If
XForms has to include localizing behaviour, I think that is an unfortunate
compensation for XML Schema's over-simplicity (!) in this area.
It comes down to the fundamental issue of what a schema language is for. XML
Schema's view of a schema language is not "how can we express or constrain
idiomatic markup languages?" but "how do we assign and derive (ultimately
storage-based) types?"
Cheers
Rick Jelliffe