All of this is a long way of saying that one thing I have discovered in making my computer more searchable is how slowly I add to my list of files. The combination of Microsoft's lousy search tool in Windows XP, and my own sloppy filing system, had me thinking that there were millions of files here, and that I was adding tens of new files by the day. There isn't, and I'm not.
Unlike the web, the amount of stuff on most desktops is finite and changes very slowly in comparison. The vast majority of it arrives in the form of web pages (and downloads), e-mail (and attachments), music purchases, and possibly the occasional CD-ROM or BitTorrent or file sharing network download.
And from the size point of view, the vast majority of space is consumed by music, images, and video. The density of "searchable data" associated with a 16MB movie clip is very different than that of a 16MB PowerPoint MS Office document or PDF file.
When you think about the problem in those terms, indexing performance on a multi-gigahertz PC isn't the issue it might seem to be. Instead, you probably want to index as much data and metadata as possible.
The real trick is deciphering all those file formats. When you compare Google's toy with something a bit more sophisticated like X1's product, you notice that's one of the significant differences between them. It's no wonder that GDS is free.
It's also no surprise that both products grok Outlook e-mail.
Posted by jzawodn at December 05, 2004 08:23 PM