[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

Re: ITP: xslj, a XSL processor (XSL is one of the stylesheet format of XML)



On Saturday 26 June 1999, at 20 h 33, the keyboard of Adam Di Carlo 
<adam@onshore.com> wrote:

> Practically, most of the problems with actually dealing with Jade boil
> down to learning curve (a lot of people are learning SGML or XML at
> the same time), or problems with the backends.  Jade backends are not
> always what they might be. 

It is certainly the main problem with full-SGML. Unlike HTML, where the 
Ordinary User can actually read a few "HTML for pithecanthropes" pages in the 
morning and have its Web server working in the afternoon, full-SGML needs 
twenty years of careful study to be effective.

The lack of good tutorials, the lack of readable reference materials, add to 
the problem.

No wonder XML had such a marketing success (the technical success is yet to 
come). Not only it is free (see Bart Schuller's message) but you can write a 
parser in an hour, you can read and understand the spec, etc.

> CSS is just an annotation model -- you just attach style to rectangles
> of text.  You can't use it to do things you might do in DSSSL or XSL.
> For instance, you couldn't create a TOC in the stylesheet with CSS.

Yes but CSS, like HTML and unlike DSSSL, can be learned in less than a day. 
Whatever its proponents say, XSL in its full power does not seem much easier 
than DSSSL.

> > The only widely accepted XML formatting standard is to use XSLT to
> > generate HTML. Both print and native browser rendering is currently in
> > turmoil.
> 
> Actually, you can use jade/DSSSL to render XML or SGML.

For XML, the Perl/Python/Java hacker will probably find much easier to write a custom parser (which is, in effect, a stylesheet hardcoded into the parser) than to learn DSSSL. With tools like XML::Parser for Perl or XP for Java, it is a piece of cake.

It is probably not realistic for a huge DTD like DocBk, but for most DTDs, it is a very convenient way.




Reply to: