XML/XSL versus BML

XML/XSL versus BML
Posted 11:14 PM

The HTML standard, which the W3C was built around, was written by carefully observing what people were doing with HTML and then concocting a standard around that. This is how good standards are created - by wrapping up a lot of people's hacks into one neat documented package.

The W3C, taking their success as a sign that they needed to mess with everything, decided to launch a barrage of new standards, the most egregious of which is the XSL/XSLT set.

So what are people hacking today?

It's all about BLOCKS. Everything in information management and reporting apparently revolves around creating templates and then inserting variables into those templates. Most major pieces of web development software has a different BLOCK syntax, from SQL Server to Wikimedia to Kuro5hin. I wrote an open source macro insertion language 8 years ago - and still use it today. (CVS access) It seems that writing markup languages is a neverending hobby of many developers.

One of my favorites is BML. Brad Fitz has, in a few lines of perl, created a block manager that outstrips the W3C. By exploiting CODE blocks of perl, he avoids the need to reinvent a text manipulation language for the more complex transformations. The result is that BML is just as powerful as XML/XSL - without the overhead or the complexity.

This points to an overriding problem with XSLT, CSS, XHTML (egregious) and all the new "standards" coming out of the W3C. They are not being written with any idea of what standards are.

Standards are a documentation and consolidation of the best and most common industry practices, not the result of a committee that wants to dictate what the most common practices will be. A good standards organization spends most of its time observing, documenting and commenting on existing protocol - producing standards documents only when one becomes obvious.

Labels: web development

[View/Post Comments] [Digg] [Del.icio.us] [Stumble]


Bracing against the wind
www.documentroot.com