The standards/design community has taken notice of the new Hotbot, and with its bold design statement being carried out in XHTML and CSS, it’s certainly worthy of comment. Unfortunately, there’s a slight problem with it. If you visit the skinning preferences page in Mozilla, Compuserve 7, or AOL for OS X, you get the following message:
To choose a new skin for HotBot, you must download a browser that supports Web standards.
Visit the same page with Netscape 7.x and you’ll have no trouble at all. All of these browsers use, essentially, the same rendering engine. They have the same standards support, give or take a few bug fixes. The source of this roadblock seems to be a poorly written detection routine on the server itself.
Fortunately, this is a problem that’s easy to fix. All the HotBot folks have to do, as my co-worker Arun wrote so cogently, is spot the Gecko, and here’s hoping that they do so soon. If you’re doing UA detection of any kind, either client or server side, then you ought to read his excellent explanation of how to detect the whole Gecko family at once, rather than client by client. It’s liable to let you avoid a whole lot of headaches. You’d avoid even more if you did object detection instead of UA detection, but one thing at a time, I suppose.
It’s just occurred to me to wonder if anyone’s written an AmIHotOrNotBot. The search parameters in the advanced interface would no doubt be very interesting.