I'm a prior offender: In my case it's due to a "light" understanding of HTML and XML. I was under the (obviously mistaken) impression that empty tags (p, br, &tc.) should have the "closing" flag for completeness. But as you can see, the only completeness is in my ignorance! 8^)

I don't develop websites, I just generate a few static pages of data. I tend to use boilerplate (e.g., XSLT) and fill in the blanks. So I've seen a good bit of *ML, but can't claim to "know" it. Thus, when I write from scratch, I type what I *think* works, and fix bits until the receiving program quits bitching.

...roboticus

P.S. Now that I've (just) modified my PM settings to be as picky as possible on the HTML, I just had to learn how the consarned p tag works. (As you can see, above I thought that p was one of the empty tags.)