A Plea for Schemas

[Paul Prescod]

Tim Bray wrote:
> 
> At 12:07 AM 11/2/99 -0600, Paul Prescod wrote:
> >By the time you have a schema for your domain you have mapped out your
> >communal understanding of that domain.
> 
> This is a surprising and counter-intuitive assertion that needs some
> supporting evidence.  I disagree.  I think large parts of the the
> "communal understanding" live in wetware between human ears and in messy
> procedural code.

I agree. But there is also a huge amount of communal MIS-understanding
that lives in wetware. It is flushed out through the process of
formalization. For instance we all talk about "links" but it is only in
trying to formalize XLink that it became clear that we all used the word
to mean radically different things. 

You can also use UML, set theory or Zen meditation to flush out
misunderstanding but schema writing has the big advantage of providing
immediate, interactive utility. Some customers say: "You mean we have to
go to all of this work to understand our problem domain before we
implement the code? Who has time for that?" The more saavy ones say:
"finally we have an excuse to map our our problem domain without feeling
like we are wasting time drawing pretty pictures."

> >In many areas that *is* cracking
> >the nut.
> 
> Where I work, people aren't satisfied until there's running code that
> does useful stuff, and the schema-ware is an essential but fairly small
> component of getting there.

I often start with two piece of running code and finding the common
language between them *is the problem*. In other cases, of course, the
schema is done before I come in and they want help with the code.

> > As a practical example, consider word processing. By now there
> >are dozens of word processors and desktop publishers. It would be a
> >massive effort to come to understand their commonalities and define the
> >"universal DTP language."
> 
> Huh?  Weren't we talking about schemas? -T.

Sure, we're talking about a schema for the common language between word
processing programs. Once you have such a thing, writing converters into
and out of it is relatively tricky ... its the figuring out the
commonalties between the different models that is the hard (perhaps
impossible) part.

 Paul Prescod

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)