Re: [docbook] Whatever happened too CSS+XML?

["Scott L. Holmes" <scottlholmes@yahoo.com>]

--- Jirka Kosek <jirka@kosek.cz> wrote:
> So you think really that people will be
> satisfied with documents where:
> 
> * are no links
> * are no images
> * are no tables
> * ... and dozen of other features

I would caution against assuming what people
would be satisfied with. At the speed of the
internet, we may all be using some tool like
Flickr or Del.icio.us in two years because that
is about how fast things can change.

The beautiful nature of semantic markup is that
it may be rendered in various ways using a
variety of tools to produce a variety of affects.

> With just XML+CSS you can not achieve things
> which I mentioned above. By 
> achieve I mean that it will work at least in a
> mainstream browsers like 
> IE, Mozilla and Opera.

Things looked hopeless for a while with Netscape
4.x as well but market forces and other pressures
put an end to the necessity for supporting that
user agent.

Based on my research of the xhtml 1.x specs,
xhtml rendering is firstly an xml parsing
process. I'm pretty sure that these browsers are
all proto-xml rendering engines at their core -
each with their own set of capabilities.

So the clumsy point I'm trying to make is that
our emphasis should be on delivering semantically
rich content to a user agent in a form that fits.
Some content can be xhtml and some xml, some CSS
and some xsl, some javascript and some vbscript,
some svg and some flash.

Here's an excellent example of stuff:
http://www.kanzaki.com/works/2004/imgdsc/0106.html

> I will celebrate day when common browsers will
> be able to display 
> DocBook XML directly, styled with CSS and with
> working links, images and 
> tables. But until this happen (read "for the
> next ten years") I will 
> downconvert DocBook to good old HTML for crappy
> browsers.

The technological achievement of modern browsers
(IE included) is positively astonishing. The
layers and layers of functionality is undeniably
some of the finest software engineering yet
devised. I would hardly call any of them "crappy
browsers.” Any organization that even attempts
to implement the w3 box model has my deepest
respect (and sympathy - poor buggers).

They're user agents. They render stuff.

> I'm not saying that I'm happy with this
> situation, but here we are. 
> Pretending that some technology works, while it
> is not true is the worst 
> thing you can do for such technology. Premature
> and unsuccessful launch 
> is not good marketing strategy IMHO. Play with
> this stuff in your lab, 
> persuade browser vendors and let us know once
> technology is ready for a 
> real use.

Oops. Given this logic, I'm guessing we should
close down sourforge.net.

One attractor for docbook is its clear emphasis
on semantic markup making it an ideal dialect for
sourcing other documents. What I believe would be
useful is an interim dialect (xhtml basic or even
xhtml 2.0) that is not necessarily intended to be
rendered in a user agent. Then the problem domain
of the user agent would be vastly easier to
subsume.

Cheers!

Scott