Paul M Foster wrote:
On Mon, May 18, 2009 at 04:43:20PM -0700, Michael A. Peters wrote:
Paul M Foster wrote:
That's the same problem XML has. The original idea was that you could,
for example, have an invoice, and because it was marked up with the
appropriate tags, everyone would be able to understand what it meant.
xml provides a standard way of pointing the reader to a reference
specifying how the document is to be read.
I have run into problems before (data from biological databases provided
in xml) that do not define the element and attribute semantics. That's a
problem with the content generator, not xml, which I believe is just a
subset of SGML.
That's true for web documents, but not for others. That's why I framed
the comments the way I did. Things are much more standardized for XML on
the web. But in non-web contexts, there is often no reference to a DTD.
The developers know what it is, but it's not referenced in the XML file
itself.
Even then, let's say one company calls the invoice date tag "invdate"
and another calls theirs "invoice_dt". How does one reconcile this
except by tedious manual examination of the standards at both companies?
Companies that want to share docs need to agree upon a standard.
Third parties that want to take data from multiple sources need to write
translation filters from source xml to whatever standard they choose to use.
My point was that, among other deep flaws, XML typically suffers from a
lack of standardization (except in the web area).
If there is an area with demand for a standard, nothing is stopping
those who would benefit from a standard creating a xmlns / dtd / etc. to
describe acceptable documents that can even be validated against.
Several organizations have already done this for data exchange - from
cooking recipe's to GIS to religious texts. It seems that in many cases,
where a standard is created is where XML is primarily used as a
transport file format between databases, or even used as a database
itself (though embedded real databases like bdb and sqlite are usually
faster than using xml for the storage).
I like xml, but I agree that using it without having a definition
standard declared that defines the elements/attributes and can be used
to validate a document is rather frustrating to the one trying to figure
it out and work with it.
--
PHP General Mailing List (http://www.php.net/)
To unsubscribe, visit: http://www.php.net/unsub.php
