[
Lists Home |
Date Index |
Thread Index
]
Don Box wrote:
>
>...
>
> Relax NG has a much looser model (a la Perl) in which two things are
> compatible if they share structure independent of a common, named type
> definition.
I think we should be careful to distinguish between structural typing,
dynamic typing and loose typing. There are statically typed programming
languages that are structually typed. For instance, you could say that
of the C++ templating feature. RelaxNG is ultimately a language for
statically constraining interfaces and thus has more in common
philosophically with C++ templates than with Perl.
> Each approach appeals to its own community of users in very deep ways.
James Clark, an editor of RELAX is a Java programmer. And Henry
Thompson, an editor of Schema is a Python programmer.
> This explains the almost rabid hyperbole from both sides, especially
> when someone makes an off-the-cuff Star Trek-TNG reference during a
> keynote talk ;-)
I think that the rabid hyperbole arises because people don't like to be
forced into solutions that they feel don't fit. Software vendors always
want to project the sense that the industry has established consensus
and everything is moving along fine. Confusion delays software
purchases.
As more or less a bystander in the schema wars, I'm looking for a few
disinterested third parties to emerge as examples of people who find
that XML Schema appeals to them in a deep way. John Cowan claims that
people who spend a reasonable time doing a "taste test" prefer RELAX.
I'm curious if anyone without a stake in the argument will stand up and
be counted as a counter-example.
Paul Prescod
|
