Posted on 10/11/2005 12:56:54 PM PDT by N3WBI3
Welcome!
Better Desktop is a project dedicated to sharing usability data with Linux developers. Over the past year, we have conducted many usability tests on different parts of the KDE and GNOME desktops. We created this site to serve as a place where developers can watch videos of these tests. Here you will find over 200 videos of people using Mozilla Firefox, Evolution, Open Office, Banshee, F-Spot and other applications. All of these can be found in the data section of this site.
The video repository on this site is growing. We will continue to add video as we produce it. Please visit us again soon to see what is new.
Better Desktop is sponsored by Novell. It is part of the OpenSUSE project. Why We Do Usability Testing
We do usability testing at Novell because we want Linux to be easy-to-use and easy-to-learn. Usability testing provides us with a reliable way to measure how well different software designs meet these goals. We do usability tests for three main reasons:
1. To find out which parts of a given design work well for our target audience and which parts don't. For example, we ran general tests on the F-Spot photo organizer because we wanted to gain an overall understanding of what its interface did well, and what it did poorly.
2. To compare the effectiveness of various designs. For example, we ran comparison tests on KDM and GDM to figure out which program was more usable for members of our target audience.
3. To verify that our specific usability goals have been satisfied. For example, after running several iterations of tests on the Gnome sound capplet, we proposed a set of changes to its user interface. After these changes were made, we ran a round of usability tests on the capplet to ensure that our changes did in fact help us to meet our goals for that tool. Data
Our testing methodology allows us to collect many kinds of data, including behavioral data, qualititative data, and quantitative data. Each of these types of data helps us to build better software in a different way:
1. We collect behavioral data simply by watching our test subjects use our software. By paying close attention to what they are doing and saying as they go through our tests, we learn about their habits, their workflows and their expectations. This data helps us to gain a more complete understanding of how people in the real world relate to computers.
2. We collect qualitiative data by videotaping our subjects as they think outloud about our software. This data helps us to make educated guesses about how to improve our designs. It helps us to understand how our test subjects feel about the software they are using. By paying attention to our test subjects' expressions, tones of voice, gestures and pauses, we can learn to improve the quality of their experience. We use qualititative data to help us make using our software pleasant.
3. We collect quantitative data by keeping track of statistics like how many test subjects completed a given task correctly, the average time it took to complete a given task, the percentage of test subjects who gave up on a given task, etc. Over time, we will add more statistical data to this site.
It's all about the amount of effort put into HCI -- Human Computer Interaction.
Due to its nature, there was never a good central concerted effort to get a cohesive interface and general "way of doing things" in Linux for the average user. It's not quite out of its geek stage yet. And their best stuff just attempts to copy Windows.
Microsoft has spent the years basically just trying to copy Apple, and blown it for the most part due to the different metaphors, and thinking they can do it a bit differently and have it still work right.
With both Linux and Windows, usability guidelines for third-party developers aren't that good.
With the Mac, Apple has from day one worked on usability. They don't always get it perfect, but there is a general cohesiveness, with the computer trying to help rather than confuse you. They also have had for a long time very good and detailed guidelines for developers, down to what should go on dialog buttons (descriptive action language instead of yes and no) and how they should be placed, to the pixel.
As with Windows' Ctrl-Alt-Delete, the Mac's biggest former UI boo-boo (dragging a disk to the trash to eject) was only meant as an internal shortcut for developers. However, they've even solved that, as now the trash icon turns into an eject icon when you move a disk's icon.
Their biggest continuing (although getting better) boo-boo was copying the Windows Taskbar as the Dock in order to woo Windows users -- although they did implement it a lot better.
Does Mepis have a version for x86-64, like an Athlon64?
However, once you have learned the software, then how would you rank them? In other words, compare an expert-level Linux user to an expert-level Windows or Mac user. Which software would then be easier to use?
Given the choices available in the *nix world(s), I would say that Linux would be easier to use.
As far as GUI is concerned, I'm talking ease of use too. For example, individual menu bars on each window, that's bad news from an HCI standpoint.
Which software would then be easier to use?
Let's just use the menu bar as a starting point. The HCI concepts of motor memory and infinite height mean you will always be able to get to your menu items faster on a Mac.
Another example: the interface guidelines for buttons on dialogs helps keep the experts from clicking dialogs without reading the tiny text in it. On Windows you see "yes, no, cancel" and you often click too fast, clicking for what you think the dialog says. You may only look at the button. On a Mac you'd be more likely to get "Save, Don't Save, Cancel" with the dangerous button ("don't save") put out by itself. That and other visual cues help save us from ourselves.
The link is now bookmarked and the next time I go on my Linux PC, I will check it out.
Thanks.
bookmark ping. I know the thread is old, but I now have an unused PC and maybe can get to the Linux install I have been planning to do for a couple of years...
Disclaimer: Opinions posted on Free Republic are those of the individual posters and do not necessarily represent the opinion of Free Republic or its management. All materials posted herein are protected by copyright law and the exemption for fair use of copyrighted works.