Re: XML-Data: advantages over DTD syntax?

Sean Mc Grath (digitome@iol.ie)
Mon, 29 Sep 1997 22:58:21 +0100

[Rick Jelliffe]
>
>Because their form of schemas are so complicated and verbose to read
>that you will need browsing tools to manipulate them. This in turn
>gives schemas (even though they are written in XML) the nature
>of binary objects rather than textual objects.
>
A good point. I have fond memories of being able to understand Make
files for example! These days, with "advanced" tools they are still
"text only" they are pretty impenetrable and effectively locked in to
particular tools:-(

On the other hand, in the specific case of XML-Data I would have to say
I am in favour. DTDs are prefectly good "documents". XML's reputation as a
meta-language is, I think, positively served by its use to describe "itself" in
this way.

The approach obviously has its practical limits though. The further one gets
from
"data" the closer one gets towards "algorithm" - the less *practical* a tagged
representation becomes. Full scale Scheme would be pretty inpenetrable in
XML but it would be possible! The fact that it is entirely possible is the
important thing. It means (doesn't it????) that XML can be viewed as the
bed-rock on which all the other required syntactic "short hands" can be based.

So XML could have 8879 DTDs. It could also have a DTD for 8879 DTDs.
Core XML could interpret the latter directly, supporting the 8879 syntax via
a transformation. Future syntaxes, methods etc.; for achieving what 8879 DTDs
achieve could then be cleanly layered on top.

Sean Mc Grath

sean@digitome.com
Digitome Electronic Publishing
http://www.digitome.com