Lists Home |
Date Index |
Mike Champion wrote:
> 1/30/2002 4:12:13 PM, "Jonathan Borden" <email@example.com> wrote:
> > Terseness aside, there is something to be said for human
> > readability, and problems with prefixes aside, people are
> > drawn to qnames because they are
> > easy to read, especially if you use a well-known prefix.
> Right. This is one of those issues where you have all sorts of
> options and all sorts of situations under which the options are more
> or less appopriate. I'd remind people of Tom Bradford's "Clean
> Namespaces" proposal http://www.tbradford.org/clean-namespaces.txt
> which more or less reflects the old sml-dev discussion.
Tom's proposal is nice. But seriously imagine the volume of email that would
be generated by the suggestion that the default URI associated with
"org.world:" actually be http://world.org ...
But it really doesn't matter because his example fails the IE5 test, which
means that any XML document which fails to parse in IE5 will generate too
many calls and emails to make anything worthwhile.
No seriously, the reason it doesn't matter is because XML Namespaces are a
done deal. All this discussion would have been fine when the rec was under
development, and I understand there _was_ alot of discussion. But now, circa
2002, there are literally millions of copies of deployed software that
require XML namespace conformance, as well as many W3C recommendations that
require XML Namespaces conformance. In the "let a thousand flowers bloom"
theory of the internet, XML Namespaces has bloomed, and bloomed big. You may
not like it that the flower is purple and sticking right in your face, but
it has been deployed and resoundingly adopted. My philosophy is not to sweat
what might have been, rather deal with what is.
> Sometimes DOM Level 1 or DTD compatibility is more important, and
> something like Clean Namespaces makes sense. Other times, terseness
> and human readibility is more important, and the full power of the
> Namespaces Rec and DOM Level 2 makes sense.
The single biggest problem I have with XML is the lack of compatibility
between XML Namespaces and DTDs. But guess what? XML Namespaces appears to
have won and DTDs have lost. Is _anyone_ really working on fixing DTDs? The
main benefit of DTDs (IMHO) is the dirt simple syntax, but to be honest
James Clark's non-XML syntax is well, pretty simple.
> And then there are the times when you need to use DTDs and namespaces
> and DOM and XPath and Canonical XML, and a career move into a less
> stressful occupation, such as a Middle East peace negotiator or an
> Enron spokesperson, makes the most sense :~)
Yeah I've dealt with that pain DTDs+Namespaces on several occasions, for
example trying to figure out XHTML Modularization ... oy. But no one seems
interested in really fixing that problem, perhaps because hardly anyone is
working with DTDs?.