W3C home > Mailing lists > Public > w3c-dist-auth@w3.org > October to December 1997

XML DTDs

From: Yaron Goland <yarong@microsoft.com>
Date: Tue, 7 Oct 1997 23:45:21 -0700
Message-ID: <11352BDEEB92CF119F3F00805F14F48503DFC767@RED-44-MSG.dns.microsoft.com>
To: "W3c-Dist-Auth (E-mail)" <w3c-dist-auth@w3.org>
It's at times like these that you find out who your friends are. They
are the ones who send you personal, not public, e-mail saying "Shut up,
you're making an idiot of yourself."

The actual language in most of the letters was a bit stronger but one
does try to maintain a certain level of decorum.

That having been said, the basic issue seemed to be - Does one require a
DTD to define an XML document? 

The answer is not necessarily. Many XML documents can be parsed without
ever having to touch a DTD. In fact XML documents which require a DTD to
be properly parsed have to put a declaration to that effect at the
beginning of the document.

However this is not the real issue, the real issue is the use of DTDs in
defining XML elements. If one is going to define an XML element, say in
a standard, then one should use DTDs to provide the definition. This too
is not strictly necessary. One could define the XML element in Sanskrit.
However using a DTD means that you can dump the definition into an XML
parser and determine if your syntactical ducks are all in a row. This is
useful both for validating the standard and for run-time parsers who
want to make sure that the incoming XML document is syntactically
correct.

Of course, the sad part about this whole flame war is that no one ever
argued against putting DTDs into the DAV Document. In fact, putting DTDs
in DAV has been on the  DAV "to do" list for quite some time now. The
argument was strictly over timing.

The reason why we did not move ahead with it is because many people can
not read DTDs but everyone who can read DTDs can read BNF because DTDs
are based on BNF. As such it made sense to use straight BNF so that
everyone could read the spec. It just seemed a bit much to say to HTTP
people who are generally familiar with BNF but not DTDs "We would like
you to review our spec, oh and please read this 65 page spec first." 

Eventually we would have to transfer over but we had hoped to do that
later, once the semantics of DAV were settled and issues like syntax
became more important.

The issue is now academic. Someone has come forward to help us do the
translation.

Nothing gained. Nothing changed. A lot of screaming and yelling where a
little common sense could have solved everything.

I have taken a lesson from Sankar's post. Rather than screaming and
yelling he calmly asked if anyone was going to help do something
constructive.

I just wish it had been me. 

I apologize for my part in this mess.

		Yaron
Received on Wednesday, 8 October 1997 02:45:43 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Tuesday, 2 June 2009 18:43:43 GMT