[
Lists Home |
Date Index |
Thread Index
]
Hi Dare,
Dare said:
I disagree with you and completely agree with Simon on this regard.
Platform and implementation specific technology like how to map XML to
programming language objects is the last thing that needs to go through
the design-by-commitee/lowest-common denominator process that has
characterized the W3C's efforts in coming up with programming language
related standards.
This is exactly the problem with the DOM which you rail against. One
standard API for all languages regardless of whether staticlly or
dynamically typed, weakly or strongly typed, garbage collected or not is
bound to be insufficient many cases. Add the fact that this API will
have to ignore programming idioms, design patterns and naming
conventions in most of its target languages since it will just pick one
means that using it will be counter-intuitive from using other APIs in
these target languages.
Didier replies:
I agree that some languages cannot support any dynamic creation of
objects. This is probably why we ended up with the DOM and static
objects like elements and attributes. However some other languages like
ECMAScript allows that. By trying to isolate the processing from the XML
syntax and not picking at least a prototype language we end up with a
situation where we sit between two chairs. And even strongly typed
languages could be adapted to XML element trees. It seems that we are
trapped in a single paradigm and that paradigm is not what the core
mainstream developers need. This is why we end up with other solutions.
This has nothing to do with all the syntaxic complexity we face today.
It has to do with insufficient tools or tools more archaic that we used
to have. Without any standardization organism, we let monopolies to set
one. Are monopolies working for the users? In part yes but they mostly
work for their stockholders.
Cheers
Didier PH Martin
|
