The toman diaries

Filesearch ala Google

lørdag, oktober 16, 2004

Google has relesed a beta of their take on filesearching, called Google Desktop Search.

It works by indexing all files on your computer, that is, it scans all files and put them in an index, a database. This lets you find the files a lot faster than you otherwise would. It scans continously, so keeps the database up-to-date enough so that you can find what you browsed in Opera or IE just a minute ago.

This [indexing] isn't new of course, MS Windows 2k and XP already has it built in and its called the Indexing Service. You turn it on and off in the disk' properties and search the database from an awkward interface somewhere completely else. Unix also has a similar utility called locate which searches a database normally created every 24 hours. Other similar programs exist for windows as well, but their names slip me right now.

Google Desktop

Whats really different about Google's approach is its inclusion of your local database in normal websearch. It appears just like google news does, above your normal search results. Exactly how it does this I dont know yet, but it works equally well in IE as in Opera, so I'm guessing it is hijacking some networking component in windows itself. A little spooky if you ask me. This is also probably the reason why its incompatible with so many programs, including the Microsoft Firewall Client on the ISA Server, some anti-virus (nod32, Panda), and, notably, the webhancer dll, which also hijacks windows networking. (Some of these incompatabilities can be fixed though, but thats not the point...)

Another nice(?) thing about Google Desktop Search is that it finds Outlook (and Express) Mail, in addition to having categories for spreadsheets and word documents etc. All windows-centric in other words..

Now let me see if I get enough time on a windows computer to really like this app.