I know that spiders are incapable of filling out forms and the like. But if I have links that have been dynamically generated client-side by some simple javascript, are they capable of following these links?

Unfortunately that is not good news for me. The code I had in mind was the "featured threads" listing in the sidebar of my forum index. It uses http://www.daniweb.com/techtalkforums/external.php?type=js

<script src="external.php?type=js" type="text/javascript"></script>
<script type="text/javascript">
<!--
for (i=0;i<10;i++)
{
  document.writeln("<div class=\"smallfont\" style=\"padding:8px\">" + threads[i].threaddate + " " + threads[i].threadtime + "<br /><strong><a href=\"thread" + threads[i].threadid + ".html\">" + threads[i].title + "</a></strong><br />" + threads[i].poster + "</div>");
}
//-->
</script>

According to the link you sent me, google is experimenting with a spider that can follow simple JavaScript that clearly contains the URL only. Oh well.

On a positive note, I think I cought wind that it is also experimenting with a bot that can read javascript write stuff too. But I wouldn't bet the farm on it.

That looks like something you could do with PHP tho.

That would be nice. In the meantime, I have added <noscript> tags.