Saturday, May 22, 2004

Site Search Engines 

As some of my earlier entries have noted, I have been having trouble getting Google site search to work. The problem, it turns out, is that Google has not yet built in the technology to search individual directories--only whole domains. However, I had to take a very roundabout route through Google's FAQs in order to find out this information. The same roundabout route uncovered a temporary workaround. You take the code that they give you at the address above and insert

<input type=hidden name=hq value="inurl:domain name/directory/">

(Replace the bolded text with the appropriate information).

This seems to work pretty well. However, there are a few drawbacks. Although Google has some customisation features, most of them are simple HTML or CSS additions which you could do yourself with a bit of coding knowledge. It doesn't give you a lot of flexibility to exclude password-protected directories, or other useful managerial functions. More importantly, as far as I can tell, it can take weeks to have your site spidered by Google (in laymen's terms--for Google's software to index your site and make it available for searching). If you update regularly, you've got a real problem. I could just about put up with that for my web site, but not for a web log.

Currently, I am trying some alternatives. Right now, I have a form for FreeFind on my web log (see the links column to your right). I like it a lot, and it's very easy to set up. I have a slightly more complicated service called PicoSearch on a draft version of my web site. Both these services work very well. Their main advantage over Google is that you can exercise much greater control over how often their spiders index your site. You can initiate the indexing if you want or even configure the spider to index your site every day. I haven't yet decided which of these services I prefer.

Of course, there are a range of technical issues about which type of service is the best, but these issues are only of interest if you're even geekier than I am. Still, it's a relief to get a site search capability working in some form. The need is not desperate for my web log (yet--although I have already made use of the search capability for something other than testing), but, as my web site has become quite large, the need is rather more pressing. The next step is to place the search form in the design of the web site, which will take some more work. Eventually, I'll want to integrate the technology into the English Department page, which is also becoming quite large.

Comments: Post a Comment
This page is powered by Blogger. Isn't yours?