Posts RSS Comments RSS 394 Posts and 1,279 Comments till now

GNOME 2.6 Usability Study

User Instinct recently published a GNOME 2.6 Usability Study and Review. This article was also featured on Slashdot the other day.

(As a side not, let me just comment on how tiny the text is in the UI documents)

Overall, this is a very good document stating to good and need-to-be-improved things about the new GNOME interface. FYI GNOME 2.6 is the second version to embrace the newly formed HIG. I am not sure what audience they had in mind when they wrote this article, new or experienced users, but the recommendations they suggest would accomodate new users very well. Some of their suggestions would in turn make more work for experienced users than necessary.

There are several points I do not agree with in the article:

  • Several times they reference the Windows ™ way to be the Right Way ™. I agree that including manners and actions similar to Windows will increase usability to users who are familiar with that interface. However even Windows has interface problems, and using the Windows interface as a basis for your guideline wont necessarily improve the overal state of User Interfaces.
  • Their suggestion of removing the terminal window from view (by default) to maintain the abstraction the simplified interface is trying to gain. Unix is a very CLI oriented operating system. Although there are many sets of sophisticated GUI’s to help users maintain system functions traditionally only availabe on the command line, they are in no way a sufficient replacement. Until there are sufficient usable interfaces that experienced system administrators feel comfortable with, a terminal should not be removed as a focus point. (Note: they do suggest keeping the link in the applications menu)

One of the most interesting and important parts in this article were the user surveys.

Interestingly, the users had difficulty locating the web browser. I wonder if they had difficulty finding it in the application menu (organizational and naming issue) or if they expected a link to be on the desktop or toolbar.

All test users had difficulty installing applications. The author suggests abstraction so that this can be achieved, but “Users” on a UNIX box cannot install applications on a system in most situations. Creating a more abstract way to install software would limit the control most UNIX users are used to.

Not suprisingly, most of the users had difficulty opening and reading a text document (they proceeded to double-click on the file in a graphical file viewer). File association can be set up in GNOME. Some of the users complained Open Office didnt look enough like Microsoft Word. I have no sympathy for them. Open Office has a very competent interface, and this is a good example of users habits interfering with their intuition and judgement in an interface (Sound familiar? The same problem exists with window manager users. To them, it is a very comfortable interface to use, no matter how unintuitive it is. Most of these users have been using these interfaces for so long, they no longer remember the learning curve when they were first introduced to them).

Gaim had the same problem as Open Office, most of the users had been using the official client and were unfamiliar with the location of the buttons in the new interface. Other problems included the lack of wizards to set up the buddy list and the default option of “tabbed” windows, causing users to miss incoming messages.

Playing media was a little difficult, however I blame that on the application choice for the test. One application was too unstable in the environment and proved unusable. The other was difficult to use because it did not come with default radio stations. These test users were picky, but I guess the majority of users are like that.

In conclusion, the article states that GNOME 2.6 was a solid interface, but needs some work. This is fair. It would be interesting to do the same study on GNOME users in a Windows environment (there do exist UNIX users who have not used the Windows interface since 3.11 or NT4).

GNOME’s Human Interface Guidelines
GNOME’s usability efforts.

3 Responses to “GNOME 2.6 Usability Study”

  1. on 29 Jul 2004 at 9:57 pmobso1337 dot org

    Open Source Usability Article
    I recently came across an article on Newsforge.com which was posted in the Open Source Usability (registration required) community.

    Overall it doesnt really give the user much information about the state of usability in Open Source. The author did …

  2. on 21 Sep 2004 at 2:09 pmobso1337.org - Weblog

    KDE 3.3 Usability Study
    User Instinct recently posted a usability study on KDE 3.3. Several months ago, they had also posted a usability study on GNOME 2.6.

    Overall, the article indicated that KDE is “good”. It recognized several groups who have been actively improvi…

  3. […] Also check out my post on the GNOME 2.6 Usability Study. […]

Trackback this post |