Fw: Global vs. local

["Mavis Cournane" <mavis.cournane@cognitran.com>]

Dear all
I am forwarding some comments on Local vs Global by Mikkel Hippe Brun.
Mikkel joined our NDR discussions yesterday from the Danish Ministry of
Science and Information.

If you are responding please do so to the list and to Mikkel directly who is
not a member of NDR.

Regards
Mavis
----- Original Message -----
From: "Mikkel Hippe Brun" <mhb@schemaworks.com>
To: "'Mavis Cournane'" <mavis.cournane@cognitran.com>; "'UBL NDR'"
<ubl-ndrsc@lists.oasis-open.org>
Cc: <ksh@sookmyung.ac.kr>; <bikeev@ean-int.org>; <hsedi@attglobal.net.org>;
<bernd.boesler@din.de>
Sent: Tuesday, April 29, 2003 10:54 AM
Subject: Global vs. local


> Hi UBL NDR SC,
>
> Thank you again for letting me participate in our meeting.
>
> I have some second thoughts following our discussion yesterday about
Global
> vs. local.
>
> Gunther Stuhec presented some of his concerns about naming inconsistencies
> following a Global approach.
>
> Our discussion revealed that to the human reader - the Local approach
would
> hide information about an elements type in an XML-instance. Gunthers
> argument was that the parsers and tools would have no difficulty
validating
> the underlying types, and that the local approach would allow for more
> consistent database-designs.
>
> I agree that Gunthers approach is appealing and more consistent, but my
> concerns were that current transformation techniques are unable to match
an
> elements type. This problem will apparently be resolved with XPATH 2.0,
and
> that move reduces my concerns.
>
> The problem at hand is that XPATH 2.0 is not a reality yet, and it will
> probably take some time before the standard is widely supported in our
> tools.
>
> In my view - following Gunthers approach will (for the moment) make it
more
> difficult to implement UBL-messages until XPATH 2.0 is supported. At that
> point it will be "just as easy" as it is at the moment with the Global
> approach.
>
> I believe that with XPATH 2.0 future transformations (XSLT, SAX, DOM) will
> take advantage of "type matching" instead of "element matching". In other
> words - our style of transformation will change to a style, where element
> names do not matter.
>
> Remember that we have to
>
> I suggest that the NDR SC considers the following questions:
> 1. Which is more important:
>        a) Easy transformation but inconsistent naming of elements right
now,
> and easy transformation (based on type) and inconsistent naming later.
>        B) Difficult transformation and consistent naming right now, and
easy
> transformation (based on type) and consistent naming later.
>
> It may very well be the case, that when UBL "takes off" - XPATH 2.0 will
be
> a reality and that we may regret a choice of Global element declarations.
>
> Regards
>
> Mikkel Hippe Brun
>