Today's email to Paul Ford at ftrain.com. Posted as proof, if I was to drown today in Islais Creek, that I'd gone totally insane at the very end:

I bet you don't mess with schemas...

It's pretty hard designing a data structure while working with something like XML schemas. They're so flexible that while trying to define my blog data, I felt like Sam Lowry trying to eat his soggy morning toast in Brazil.

I think I was doing alright until the point that I tried to add a definition for A links. I was confused how to make it so you couldn't put A tags inside of A tags. Well, I'm too tired now. I read about XML schemas for about 5 hours and then spent another 5 writing my first schema for the blog. Check it out. I don't know if it means anything to ya. I chose to start with schemas instead of Dtd's because I'm starting fresh and everyone is recommending it over Dtd's. It was actually a little easier for me to comprehend. And what I learn making schemas translates to doing other types of XML.

I really didn't know how to put in markup tags like <b> and <p> and <br>. (or even if I needed to) I just threw them in where they looked good. Didn't even try div and span... I chose to describe things with attributes where I could instead of lots of elements. The books I read seemed to lean this way. Also, didn't see any need for groups with my type of structure. But, if I could get the body text described I would like to make that into a type.

I used HTMLSpy and it made it pretty damn easy to just drag stuff around and get it spelled right.

Funny, I couldn’t get any straight info about how you create a schema to describe regular old xhtml markup. There's a million ways to do "customers...order....addresses" Maybe I just reference XHTML and forget about it. I have no idea. My little brain tried to map out structure of the limited set of tags that I used and quickly got buried.

Here's the report of what I did.

http://primco.org/testblog/schema/beginning_blog_schema.html

davep