xml-general mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Donald_Les...@lotus.com
Subject Re: -1 to Anakia
Date Mon, 16 Jul 2001 18:25:46 GMT

As the writer with primary responsibility for the Xalan
website/documentation on xml.apache.org, It appears to me that we are
talking about three separate issues that should be discussed and settled

1. What document types do we want to use for the XML projects?

2. What look do we want for the HTML we publish on the website?

3. What tools do we use to transform #1 to #2.
Note: if we want to use different XML document types, we also need tools to
transform from the current doctypes to those new doctypes.

1. Currently (with some variations), we are using book.dtd (defines the
"book" for each subproject), document.dtd (defines the s1 document), and a
couple of other doctypes for faqs and lists of changes. Most documents are
s1 documents, containing up to 4 levels of nested sections (s1, s2, s3,
s4).Until we have a good working prototype of one or more alternatives that
we can vote on, I think we should stick with what we have. The current
doctypes have some annoying restrictions, but they are pretty simple to
use, and we have a fair amount of material in place.

I understand there is interest in moving to DocBook or some subset thereof.
Another possibility is DITA, a hierarchical topic-oriented structure (i.e.,
nested topics) with an open-ended specialization mechanism for
extending/inheriting shared structures. See
http://www-106.ibm.com/developerworks/xml/library/x-dita1/index.html. I
have stylesheets in place to transform from our current doctypes to DITA
topics. I also have styleshets to transform DITA to HTML and to formatting
objects, for the generation of pdf. I'm still working on getting decent pdf

2.  As for the HTML, I suggest moving to a frameset with a banner in the
top, a TOC panel on the left, and a body. If nothing else, you can use the
entire width of the page for printing the body without first having to
switch to a print format (which some XML subprojects currently provide).

3. Currently we (most uf us, all of us?)  are using Stylebook and a
collection of stylesheets to transform our "book", including any s1 and faq
documents it contains to HTML. Stylebook packages the transformations,
copies some gifs, and generates a number of jpgs that are used for the TOC
in each document.

Remove Stylebook from the picture and we could generate the same HTML with
Xalan and Xerces (which of course Stylebook already uses to do the real
work), perhaps involving a stylesheet extension to "package" the
transformations -- assuming we did away with the jpgs or came up with
another tool to generate them.  Alternatively, call Stylebook with
different arguments and a different set of stylesheets, and we could
transform the current XML to the same or different HTML  or new doctypes to
the same or different HTML.

I personally believe that Stylebook has been a pretty nifty tool. But
clearly it has its problems, it is not maintained or documented, and it has
made no attempt to keep up with the new JAXP API (except for a minimal
patch to keep it running the Xalan-J2 that I submitted many months ago).  I
would be happy to see it go (and take the sbk: protocol with it!). The only
thing it does that I do not know how to do with Xalan, Xerces, and Ant by
themselves is generate the jpgs for the TOCs. One real easy solution for
that is to switch to text for the TOCs., but I have no doubt there are
other alternatives.

In other words, I think we can remove Stylebook, and continue to use Xalan,
Xerces/Crimson, and Ant, to build our website.

-- Don Leslie

In case of troubles, e-mail:     webmaster@xml.apache.org
To unsubscribe, e-mail:          general-unsubscribe@xml.apache.org
For additional commands, e-mail: general-help@xml.apache.org

View raw message