Markup considered boring
Tuesday, May 9th, 2000
The day is done. I will take a break and do some running now - then I’ll meet Aila at the Royal Swedish Opera. We have tickets for Romeo&Juliet tonight.
Not everyone is interested in XML - I will write about other things, too. But to me even the boring stuff was fascinating today, in it’s own way.
Ah, XML! When all the beatiful meta-things have been said, you must still do an awful lot of markup before there is something to play with!
Here is one real-world scenario:
Take a large, complex non-XML document and convert it to text. Dream up a simple schema. Markup the document by hand. Generate a new schema. Think. Change the schema, validate, change the markup, validate. Try to use the document in your application. Think some more. Change the schema, validate, change the markup. And on and on and on and on…
Before doing all this I had given myself some advice:
- Talk to someone who actually uses this kind of document
- Don’t markup large documents without a basic schema you believe in
- Use a good editor (like XML Spy). Mess up your structure and NotePad can’t help you - you may never get back to wellformed XML!
And I followed my own advice. But that still left me with several hours of boring markup-work… so why do it at all?
Because trying to mark up a real document is the only sure way to find the kind of problems all the sample programs just ignore.