Google Desktop Search for Linux
Google has just released the first version of Desktop Search for Linux.. Get it at http://desktop.google.com/linux/.
I downloaded and installed it on my desktop machine, which I switched to Ubuntu just yesterday. You can read about the switch in progress here. I also installed it on my laptop, that has Beagle doing the indexing already.
Initial impressions are positive. The install process was easy, similar to a Windows install. I downloaded the .deb file, and it was a double-click install after that. After installing, it’s a bit weird, because there is no indication that it is installed. There is no icon in the menu’s, nothing on the desktop, nothing that tells you to restart, nothing at all! On the desktop, I started it up manually, after looking up the bin file. On my laptop, I logged out and logged back on. When I did that, I got the window shown below.
I read the Privacy Policy, and chose to enable advanced features. After all, doesn’t Google say, “Do no evil”? Hmmmm
Aaanyway, after doing that, the familiar little icon shows up in the status bar, and the indexing starts off. The defaults are interesting, as you can see below
It defaults to searching your home directory, as well as the /man directories. This is a nice feature, since the man pages contain a wealth of information, and for new users, it’s often difficult to use the man command. Once my index is completed, I’ll use this a bit, and see if I prefer to use this over man.
The other preferences are the usual stuff. You can decide what to index, add specific folders that might be mounted in other places, or specifically exclude some folders that you don’t want indexed.
You can also choose to search you Gmail, but since Gmail includes search itself, I didn’t enable this.
To search, you can hit “Ctrl” twice, and this brings up a little search box.
You can modify the behavior of the search box from the Preferences link, which brings up this page:
Pretty standard stuff again. You can also check the status of your index by right-clicking the status bar icon and going to Index>Index Status. This brings up a page that shows you what’s up.
Indexing does not seem to use much CPU or memory. As I write this, the indexer is consuming about 8.8 MB of RAM, and CPU usage is at 18%. It automatically speeds up when you leave the computer alone, so your normal usage is not impacted much.
Results are shown in your browser. This is different from Beagle, which comes with it’s own client. I don’t mind either, but you may prefer one approach over the other.
It indexes websites browsed in Firefox by default. This is nice, since Beagle does not do this by default. I could not find an option in Beagle’s preferences either. You need to install a browser extension. Google wins on this, definitely.
EDIT: Apparently, according to Joe in the comments, this is a limitation of my choice of distribution, not Beagle. Still, from the users point of view, Google does it, Beagle doesn’t. However, my criticism of Beagle has to be tempered now. Joe is Joe Shaw, the guy who does the maintaining of Beagle for Novell, so I think we can take his word for it, eh?
As far as the search efficiency goes, it seems to be pretty good. The index is not complete yet, but the results I am getting are pretty consistent with what Beagle gave me at the same stage of the game.
Conclusions:
Google Desktop Search for Linux is a solid tool. It’s a native implementation, none of the Wine stuff they use for Picasa. It’s fast and intuitive, and does not use too many resources. The results are presented in a web browser, which some may not like, but it should work for most people.
Indexing your web browsing is a good feature that differentiates it from Beagle, the big dog in the Linux search space. (Only on Ubuntu and Fedora. In SUSE, Beagle has this enabled.)
The Google Beijing labs people did a good job, and I’d recommend you take a look at it.