[
Lists Home |
Date Index |
Thread Index
]
His example shows that using XML to support a single function
is pretty awful.
It is. Not new news. As the guy who talked long about
coarse transactions, he knows most folks don't do that.
He knows that they do that repeatedly. Painful code to
enter and maintain, and brittle. XML is not object-oriented.
He is showing that using paths to identify data
for use by a function is verbose.
It is. Not new news. As the guy who talked long about
loose coupling, he knows what the tradeoffs are in data
driving with a syntax unification approach based on
a hierarchical format. Path identification should
go away in the program but how else will we deal
with selectors? XML is not object-oriented.
He's right that the code examples are real ugly.
So the alternatives are?
Ok, we go back to the old Xerox approach to
markup: elements and text nodes. Full stop.
Straightforward binding and no intricacies
from those messy attributes, PIs, DOCTYPEs,
and so on. IDs? Interoperability? Oops.
While we're at it, let's fix that semantic
problem. Let's bind data to code and send
it all as one neat package. Security? OOPs.
Good discussion topic. XML is not object-oriented.
Well... duh!
len
From: Sean McGrath [mailto:sean.mcgrath@propylon.com]
Article entitiled "Speaking XML" by Adam Bosworth:
http://www.fawcette.com/xmlmag/2002_12/magazine/columns/endtag/
I'm intrigued by "it is becoming increasingly necessary for developers to
directly access and
manipulate XML documents."
My reading of the article suggests that "direct access" here means "lets
make the XML a direct
serialization syntax for objects":
import Customer
Customer c = new Customer ("Customer.xml")
If this is the line of thought I'm fairly squarely against it for more
reasons than I have time
to articulate at the moment. Just looking for opinions on my interpretation at
this point!
|
