| The Client Statistics Gatherer | 1 | 2 | 3 | 4 | 5 | 6 |
The script isn't perfect, of course, but it's pretty close. There is one failure case for which I can't conceive of a solution: if the user renders the home page with JavaScript enabled, then disables it, and tries to follow a link. At that point, the link has been changed to a javascript: link, but now that'll just be a no-op. Novice users, however, don't typically know to turn JavaScript off. And no matter what kind of user it is, I can't imagine why they'd choose that moment to make the change.
Also, right now, users who load my home page and immediately leave without following a link are not recorded. But they're not intersted in staying, so screw 'em — I'm not interested in designing for them. And the statistics are only collected on the front page of my site, so users deep-linking are overlooked. But my site is not on search engines, so it's not common for users to know to bypass the top page, at least the first time. It wouldn't be hard to add the form and load the JavaScript file on all pages, but that seems a bit excessive.
Which reminds me, there are some cases where the swapLinks() routine will need to be fine-tuned. For instance, if your page already has something that updates the links on your page, they're going to be battling for control if they aren't careful. The "Hide Menu/Printer Friendly" option of my own menu is such a case, which is the main reason you don't see that on the home page. (That, and I just never got around to fixing it.)
Finally, it's inevitable that some privacy nerds will whine about all this secrecy. But the fact remains that plenty of other data is being collected all the time, both from your web browser, and by software that monitors your moves through a web site. I happen to think of this data as a hell of a lot more benign than, say, your IP address, or how many porn sites you visited in the last 24 hours (you've been to 8, I see...). I'm also fully disclosing what I'm doing here, and in the home page itself, if you happen to be sneaking a peak at the source code. These people should go focus their efforts on the real issues, and leave science to the scientists, as they say.