Lists Home |
Date Index |
On Monday 28 October 2002 03:40, Martin Soukup wrote:
> > But THEY don't even want XML; they probably don't find wandering a DOM
> > tree
> > any more friendly than calling whatever passes for Perl's "pack" and
> > "unpack"
> > in VB. They are the people who want to just have magic serialisation
> > data structures to strings of bytes.
> Good point. There are products (like J2EE 1.4 from what I hear) that
> provide OO serialization to/from XML for "free" though.
Yep, but at that point it's less relevant that it's XML; it's just some
unspecified magic that turns object here into object there, which is why many
people here distrust them and say it's no better than how we used to do
[marking information as to how software that doesn't recognise it should
> That sounds like a rather wise thing to do. Any idea why the XML spec
> doesn't include mechanisms for meta-model information?
Because XML has a fragile data model, designed for publishing stuff to a
browser rather than transfer between applications? In HTML you just ignored
unknown tags, which was fine because the text inside would still be rendered
just maybe without the desired styling. With XML they made everything more
fragile, it had to conform to a DTD, but your document supplier could
suddenly start supplying documents with a different DTD and as long as they
XSLT pointed at by <?xml-stylesheet?> was also changed to work with the new
DTD it'd still work OK. It's aimed more at displaying data to people than for
interchange of information between bits of software... I'm still trying to
find out where the 'XML data' idea first arose.
A city is like a large, complex, rabbit