Bill and I have been working together for years now. We've pretty much figured out what each other loves doing and what ticks the other person off. We've also gotten pretty good at picking out work together. One of the things we're getting Sam up to speed with is Jen and Bill's Coding Theorem.

These days, clients call up Sam to chat, but in the old freelance days, generally the call would come in to me. They have an old site, and they'd like us to do a redesign.  Maybe there's a small database driven component to the site, or some script, or even some full-blown (but old and creaky) content management system involved.

First, I take a look at the HTML, and the CSS if there is any, on the website. I try to judge the code by the era in which it was written.  If it was a site from 2000, I don't expect to see any CSS at all.  But sites built in 2004 or later, I definitely expect to see CSS.  (Sites between 2000 and 2004 vary widely -- they should use CSS, but many still don't.)

I can draw conclusions about the person/people who built the site by looking at the HTML.  There are dead giveaway things, like the Microsoft Frontpage line in the head of the document (or for that matter, the FrontPage template look), that tell you the person didn't have a clue.  But putting the same font family in absolutely every style in the stylesheet is also a clue. Or severe "classitis", in which everything in the document has a class associated with it.

And if the HTML/CSS is horrible, Bill would check out the back end and find it equally horrible.

Bill and I used to laugh about this as a coincidence, but we come to expect it now. When I call him to tell him about a potential client with awful HTML, I warn him that I expect the database or middleware code to be awful as well.  Inevitably, it is.

So, Jen and Bill's Coding Theorem states:

If the HTML/CSS sucks, the middleware/database also sucks.

Has anyone else noticed this correlation?