Style vs. transformation

[Richard Light <richard@xxxxxxxxxxxxxxxxx>]

In message <34FB3F36.1F751837@xxxxxxxxxxxxxxxx>, Paul Prescod
<papresco@xxxxxxxxxxxxxxxx> writes

>I think that this is the correct argument against a two-step transform.
>It is worth noting that DSSSL's first-step is optional, not implemented
>by Jade, and not relied upon by any of the stylesheets I am aware of. In
>other words, it is optional in every relevant sense. It would be strange
>if the DSSSL "subset" for the web made it mandatory.

I take the point about real-world DSSSL implementation being a one-step
process, but it might be worth our being more precise about the DSSSL
situation.  

The DSSSL standard has two parts - a Transformation Language and a Style
Language.  Both parts are equally 'optional': the wording of Section 2
(Conformance) is carefully symmetrical in allowing transformation-only,
style-only and 'two-stage' conforming DSSSL applications.

So yes, DSSSL _as_implemented_ just gives us the DSSSL style language,
or more precisely, the DSSSL-O subset of the DSSSL style language.
This, I would argue, has very limited transformation abilities, which
beings us back to the original point - how do we support reasonably
complex transformations in XSL?  If we do have to invent something, then
I am simply suggesting that we look to DSSSL-transform for ideas rather
than starting from scratch.

(Incidentally, I would see this issue as similar to the need to include
XLL and XSL itself in the XML framework, even though its antecedent SGML
doesn't have built-in linking and style support.  Sometimes the needs of
the Web take us further than we went before ...)

Richard Light.

Richard Light
SGML/XML and Museum Information Consultancy
richard@xxxxxxxxxxxxxxxxx


 XSL-List info and archive:  http://www.mulberrytech.com/xsl/xsl-list