Jul
24
2007
Google Desktop for Linux
Posted by luna6 12 Comments
In a splendid move by Google, the company has recently released Google Desktop for Linux. The program provides search results for files residing on your local hard drive and the internet. Similar to the Linux program Beagle, Google Desktop provides more advance search features than earlier generation search programs. Google Desktop for Linux searches not only by file name but also the data within files (PDF, Open Office documents, mp3’s, jpg’s) for more accurate search results. The end result is near instantaneous search results for files and folders that you are looking for.
Unlike some of Google’s Linux version of Picassa, Google Desktop for Linux is a native Linux application not dependant on Wine. Desktop for Linux officially supports Debian 4.0, Fedora Core 6, Ubuntu 6.10, SUSE 10.1, and Red Flag 5, but easily runs on most other Linux distrubitions that has glibc 2.3.2+ and gtk+ 2.2.0+. Since this is Google’s first version for Linux, the Linux version is more barebones than its Windows counterpart. Gadgets and sidebars does not come with the program, just the desktop search application itself. For me I found that to be a positive with less bloat.
INSTALLATION
Installing Google Desktop for Linux was a breeze. Depending on your distro of your choice you have several different ways to install the application. Easiest manner would be to download the Google Desktop deb (Debian based) or rpm (Fedora based) from Google’s website and double click the file once the download is completed.
SETUP
Once installed Google Desktop will automaically start indexing your computer on your next login or the first time you launch Google Desktop. Depending on the size of your hard drive the initial indexing can take hours. Don’t let that scare you away because the indexing happens in the background and you don’t notice much of a performance hit while the indexing is occuring.
Of course privacy is a big concern with any type of indexing program and Google Desktop has easy to set controls for how you want your files indexed. By clicking “Google Preferences” either from your system menu or the desktop icon a well layed out control panel will open in your default brower.
On my Kubuntu system, Google Desktop opened up by the default Konqueror browser. I actually preferred to have the program open in Firefox. The way to switch default browsers for other people wanting to do so would be to select Firefox in KDE System Settings/Default Applications/Web Browser/Default Component. This did open up Google Desktop in Firefox but I found that Firefox took noticeably longer to open than Konqueror so I reverted back to Konqueror.
Google Desktop can search through a myriad of different file types including : Text files and source code, PDF & PS files, HTML, Email from Thunderbird, OpenOffice documents, Image and music files, man pages, file and folder names, and Gmail accounts if you provide login and password the preferences control panel. You can easily adjust what files are indexed via the Google Desktop Preferences. Some examples of how you may want to want to modify the desktop search.
If you have specific files or folders that you dont want indexed : Select “Google Preferences/ Local Indexing/ Don’t Search These Items” and click the “Add file or folder to exclude” and then select the folder or file you want excluded.
If you don’t want to index your web history : Select “Google Preferences/ Local Indexing/ Search Types” and uncheck the box for Web History.
If you want to exclude specific websites from being indexed : Select “”Google Preferences/ Local Indexing/ Don’t Search These Items” and type in the url for the website you want excluded.
If you don’t want Google Desktop to keep track of deleted items : Select ” Google Preferences/Local Indexing/Remove Deleted Items” and check the box for “Remove Deleted Items”.
If you want Google Desktop to temporarily stop indexing : Right-click the Google Desktop icon and seletect “Index–> Pause Indexing” from the pop up menu.
Also in the Google’s Preferences there is a fourth tab for “Other”. The only option in that tab is to enable “Advanced Features”. It should be noted that Google’s Advanced Features are …”Enabling Advanced Features will allow Google Desktop to collect a limited amount of non-personal information from your computer, which is then sent to Google. This includes summary information, such as the number of searches you do, the time it takes for you to see your results, and application reports we’ll use to improve the program.” I chose not to enable the information sharing to Google.
USAGE
Actually using the program was useful and problem free. You can conduct searches via the Quick Launch Box or with your browser. To initiate the Quick Launch Box you can either tap the Cntrl key twice or click the Google Desktop icon. If you would rather use your web browser you can right click the Google Desktop icon and select “Show Homepage” or just go to Google.com and click the Desktop link above the search box.
Either way once the application is launched, the program did work well retrieving files. The benefit of having a program like Google Desktop over a locate function is that Google Desktop can search inside of files (PDF, ODT) for relevant content. Results were comparable to Beagle, but Google offerd a few more options like indexing your Gmail account (also Google Desktop does not use Mono). A useful benefit for indexing Gmail could be if you have a laptop and are not currently connected to the internet. You can still quickly browse your email offline.
Linux is all about choice and its nice to see programs coming from all areas. Google Desktop for Linux does work well and offers features that worked well and unobtrusively. I prefered Google Desktop over Beagle because of Google Desktop’s ability to index Gmail and the fact that it’s not dependant on the dog slow Mono library. Well done G people.
Pros :
Native Linux application.
Easy to control search parameters.
Index your Gmail account.
Cons :
Closed Source.
Program can feel resource intensive at times.
Gadgets/Sidebar are not included (Depending on your point of view could be a positive).
Users Rating: 7.18 
(11 votes)
Loading …
![[del.icio.us]](http://images.del.icio.us/static/img/delicious.small.gif){kind=small}
Comments
12 Comments so far
What is Linux : Red Flag is actually a Chinese Linux distro. Not a typo. If memory serves me correct they might be sponsored by the government to break the Microsoft reliance. I haven’t used it myself, but not a typo. Cheers…
I should mention that while I really, really prefer F-Spot over Digikam, it is so slow on our system that I was forced into Digikam. I was active for a while on the F-Spot list and bugzilla, and tested many patches. However, this Duron 1GB system could not keep up with our ~10,000 photos and growing collection with F-Spot. Digikam works flawlessly. I don’t know if mono is the reason, but fact is that F-Spot barely crawled on this box. Note that I do use KDE, not Gnome.
“Red Flag 5″ (second paragraph)? Isn’t that supposed to be “Red Hat Enterprise Linux 5″?
@luna6: I would say again that probably those performance problems are more directly related to the applications themselves than Mono. I’m surprised to hear you say that F-Spot is more sluggish than Digikam, because the exact opposite has been the case for me. I have several thousand photos stored inside F-Spot which handles them with ease. Even Apple’s iPhoto was utterly unable to keep up with the amount of data in it.
I’ve never really used Amarok so I can’t make a comparison with Banshee, but I will point to a blog post by Aaron Bockover, the Banshee developer, in which he states that he is now able to load a 25,000 song database in 0.8 seconds compared to 32 second previously, and do it using 8.5 MB of memory compared to 81:
http://abock.org/2007/06/27/my-hack-week-the-new-banshee/
As for why we based Beagle off of Mono, in part it was to learn a new technology. Our previous project was written in Python, and while we preferred it immensely to C we had some issues with performance and maintainability. To write a large-scale Python application and maintain it over time you have to be incredibly disciplined as a developer. You have to write unit tests for absolutely everything and have 100% code coverage tests because any changes could have a catastrophic effect at runtime. At least with type checked languages like C or (better still) C#, a huge class of problems can be fixed at compile time.
As for litigation, I have absolutely no concern whatsoever. I work for Novell, but despite the Novell/MS agreement Novell does not have any lawsuit protection from Microsoft… a common misconception about the deal.
The reason why I’m not concerned is because the patent system is a mess. Any non-trivial piece of software probably infringes on somebody’s patent somewhere. It doesn’t matter if Beagle is in Mono or Python or C, it’s probably violating a patent. I don’t know for sure, but since someone was able to successfully get a patent issued for “not equals”, it’s a pretty safe guess.
Microsoft doesn’t have a track record of litigating patents. In fact, they are sued much more than they sue anyone else. And guess what? They *always* lose. They are a big fish with lots of money and they are easy to go after. Sooner or later Microsoft will realize that they’re on the wrong side of the patent system. (In fact, based on recent lobbying in Congress, they already do.)
Moreover, companies like Novell, IBM, Red Hat and others have banded together to defend open source software against patent attacks. MS really has nothing to gain here.
The much greater patent threat to any of us is not Microsoft but patent trolls. These are essentially lawfirms that produce nothing but file and acquire patents solely for the purpose of litigating them. They make no money on manufacturing, only lawsuits. This is what happened to RIM (the Blackberry people), and it nearly bankrupt them. This is what happened with Microsoft over the web page embedding suit. They are the greatest threat to the software industry.
Lastly, I’ll leave you with this: Mono gets vilified a lot for being an open source implementation of Microsoft technology. How is Samba any different? Nobody badmouths Samba for this. How is Wine any different? Mono provides a fully open source environment in which we’ve been able to write some truly great apps like Beagle, F-Spot, Banshee, Tomboy, MonoDevelop, iFolder, etc. I think we’ll be seeing more over time.
A Pendant : Thanks for pointing that out. Fixed.
Marc : Yes Beagle indexes all of the text and metadata contained inside of documents as well file names and such, like Google Desktop.
Thanks Ante for mentioning beagle, I was not
aware of how good it is.
I’m wondering though: Google Desktop also searches
within files right? Anything like that within Beagle?
Thanks
Beagle/Tracker and other tools are way ahead of Google – they are designed for Linux. Google Desktop just brings stupid idea of indexing to Linux. This is something Linux does for ages (slocate anyone?), but Linux also has inotify. Do a little test:
Fire up your Beagle, and search for ‘mamojambo’ (wait for it to return 0 results), then open terminal and do ‘touch mambojambo’. It will appear in the same second.
With Google, you’ll have to wait indexing again. An that’s tomorrow? Or when? Eh…
Just cause of this (and ever lasting indexing) I consider it bloatware.
It sounds kinda cool, but I since there are free (libre) alternatives, I’m not touching it.
Google earth doesn’t use wine it is a native Linux application built using Qt
I’ve heard of some security vulnerabilities when using google desktop on some windows os. I hope linux version doesn’t have that problem.
Hello Joe,
In the past whenever I used Mono applications (F-Spot, Banshee, and Beagle) the programs seemed to load slower and felt more sluggish than native Gnome or Kde apps. Back then there really wasn’t a Beagle equivalent, but Beagle did feel resource intensive, F-Spot definitely felt heavier than Digikam and Banshee more sluggish than Amarok or Rhythmbox.
I do agree with your assesement that Google Desktop for Linux’s initial indexing is slower than Beagle, but gives faster results.
Also. I’ll upgrade to the latest Beagle (I see that its up 0.2.17) and see if it has improved.
Out of curiosity what was the initial reasons to base Beagle off of Mono? Do you have any concerns about any potential litigation against Mono which would affect your project?
Finally, choice and competition is always a good thing, so thanks for your work on the Beagle project. Cheers..
Hi,
I’m the maintainer of Beagle and came across your blog post.
I have one nitpick about the entry. You say that one reason you don’t like Beagle is because it’s “dependant on the dog slow Mono library.” I don’t really understand where this impression comes from. Mono is extremely fast — faster than comparable systems like Java or Python, although it would be slower than a pure C implementation.
If you have issues with Beagle’s speed, that is more likely the fault of Beagle itself rather than Mono. From my own experience with testing GDL, Beagle indexes a lot faster than GDL, although it’s not quite as fast returning search results.
Thanks,
Joe