What do XML consultants actually do?

[David Megginson]

kragen at pobox.com (Kragen Sitaker) writes:

> David Megginson writes

> > What companies hire consultants for is to help them understand how
> > to exchange and process information: 90% of the complexity comes
> > from the nature of the information they're trying to model and the
> > business environment in which they work, 9.9% of the complexity
> > comes from finding, learning, and integrating the software
> > components, and perhaps the remaining 0.1% has something to do
> > with the syntax of the markup layer (but probably not).
> 
> So in a $1 million project, perhaps $1,000 of the effort goes to things
> like implementing DTDs, 

Actually, many don't need DTDs specifically, but when they do, almost
all of the complexity of implementing DTDs belongs to the 90% (the
complexity of the information being modelled) -- sure, sometimes I
miss a closing quotation mark for mispell a parameter entity name, but
the parser brings me right to the error and I can fix it in a second
or two.  The time I spend figuring out how to arrange the information
in the DTD does not belong to the syntax, since I'd have the same
problem with XML Schema, RDF Schema, UML, DDML, or BNF.

> verifying that input is in the correct form, fixing bugs in
> outputting XML, etc.?  Sounds reasonable, if perhaps a little low.
> But maybe the projects you're working on are much bigger and harder
> than the ones I'm familiar with.  In fact, that seems quite likely.

Again, verifying that the input is correct is not specifically an XML
syntax problem -- you would still have to verify the input with SGML,
SML, s-expressions, or whatever.  The only complexity that belongs to
the syntax layer is the extra complexity that you have from XML and
not from the others.

As for output, there's not much to worry about if you're using a
library to output the XML and the library isn't buggy.  Again, you
have to make sure that the right information is in the right place,
but that's not an XML syntax problem.

> I suggest that the most important projects are the small ones.
> ArborText is undoubtedly tremendously significant, and no doubt the
> syntax of SGML contributes very little to its complexity.  But the
> significance of Arbortext is outweighed by the hundreds of little XML
> programs people have hacked up -- programs that would never have been
> attempted if understanding SGML were a prerequisite.

Not at first, I agree, but if XML didn't exist and there were 20 or 30
small SGML parsers in every language and a lot of excitement, they'd
probably be fine.  

After all, think of how many people program with sockets and how few
actually know how TCP/IP works at the low level; or, think of how many
people program with OpenGL or DirectX and how few could actually
explain the rendering pipeline.  

Simplicity gets you the libraries; the libraries get you the
applications.


All the best,


David

-- 
David Megginson                 david at megginson.com
           http://www.megginson.com/

xml-dev: A list for W3C XML Developers. To post, mailto:xml-dev at ic.ac.uk
Archived as: http://www.lists.ic.ac.uk/hypermail/xml-dev/ and on CD-ROM/ISBN 981-02-3594-1
To unsubscribe, mailto:majordomo at ic.ac.uk the following message;
unsubscribe xml-dev
To subscribe to the digests, mailto:majordomo at ic.ac.uk the following message;
subscribe xml-dev-digest
List coordinator, Henry Rzepa (mailto:rzepa at ic.ac.uk)