xml-dev - RE: power uses of XML vs. simple uses of XML

["Bullard, Claude L (Len)" <clbullar@ingr.com>]

Take it a piece at a time and it isn't terrible.
HyTime was only terrible if you tried to pick it 
up without the help of a book explaining it such 
as deRose and Durand's book.  Because it didn't 
become mainstream, and because the hypertext 
community suddenly expanded a millionfold forcing a 
devolution phase (one reason for devolution is a 
sudden explosion of members forcing conservation 
of resources), it hasn't been given that much 
attention by writers.  Unless XML Schema has critical 
flaws, I expect that it will pick up more support 
once some good books are available, but not before.

There is no *requirement* for XML parsers to 
be small.  That is a goal.   Because XML parsers 
must enable well-formed parsing and might enable 
validation, they can be small.   After that we 
get into featuritis.   Consider the well-formed 
parse as the minimum support, then start working 
on DOM, XSL, DTD, schemas, namespaces, 
Xn systems support, and 
the object can get pretty bloated.   Being able 
to determine which features must be supported 
for a given transaction is a fundamental problem 
for requirements and proposal writers.

Look at MSXML.  It currently supports a lot of 
that list and XDR for schema support.  I don't 
think we know what having support for all of 
XML schema will require.  The question one might pose 
is if all of XML Schema should be supported in the 
same object with the rest of the list.  

Has anyone asked if the set of features in XML 
schema make sense for the majority of applications?

The fact is, XML++ is no longer simple.

Len Bullard
Intergraph Public Safety
clbullar@ingr.com
http://fly.hiwaay.net/~cbullard/lensongs.ram

Ekam sat.h, Vipraah bahudhaa vadanti.
Daamyata. Datta. Dayadhvam.h


From: Michel Rodriguez [mailto:mrodrigu@ieee.org]

On Mon, 10 Jul 2000, Gavin Thomas Nicol wrote:

> The value of schemas will increase dramatically if they are 
> widely adopted/employed (for example, if all XML parsers provided 
> a simple way to use schema validation).

Doesn't this go against the requirement for XML parsers to be small? The
more I read about schemas and the rest of the standards associated with
XML and the more I think that HyTime wasn't so complicated after all!

Seriously, what would be the impact of adding schema validation to a
parser in terms of complexity, size and speed?

***************************************************************************
This is xml-dev, the mailing list for XML developers.
To unsubscribe, mailto:majordomo@xml.org&BODY=unsubscribe%20xml-dev
List archives are available at http://xml.org/archives/xml-dev/
***************************************************************************