Slashdot Log In
GNOME Usability Study Report
Posted by
michael
on Fri Jul 20, 2001 01:08 PM
from the press-ctrl-alt-shift-k-q-z-to-continue dept.
from the press-ctrl-alt-shift-k-q-z-to-continue dept.
pdiaz writes: "Here is a report made by Sun Microsystems people about GNOME usability. They collected a bunch of professionals (lawyers, engineers, Graphic Designers, etc..) and put them in front of a Gnome desktop. They were asked to perform some tasks and tell what some icons, menus, etc., do. Some quotes are really funny, like when they asked what does the terminal emulator icon launch." Very interesting stuff, and this approach is necessary because once you've gotten accustomed to the system it's no longer possible for you to evaluate how it appears to an inexperienced user.
This discussion has been archived.
No new comments can be posted.
GNOME Usability Study Report
|
Log In/Create an Account
| Top
| 313 comments
(Spill at 50!) | Index Only
| Search Discussion
The Fine Print: The following comments are owned by whoever posted them. We are not responsible for them in any way.
Re:Kind of sad, (Score:3)
I suppose that makes sense pressing that "START" button to shut the machine down? Brilliant.
yep (Score:5)
It is exactly that attitude that makes Gnome unusable to the average PC user. Until that changes, Linux won't make inroads. Taco has already addressed this though. MS spends lots of money doing this kind of thing. The best we have done so far is a cheap rip off of that. We constantly rip on MS for bad software, yet when push comes to shove, we're just trying to give away free versions that look like theirs. It isn't working. This is an important first step in usability testing for Gnome. KDE would be wise to do the same, though they are probably about a decade (maybe a little less) ahead of Gnome in usability. Gnome could/can catch up by doing things like this. KDE can't remain idle, and MS won't remain idle.
did you even read my post? (Score:3)
The problem is that almost nobody takes the time to think about what they're doing to decide if it's a good idea or not. Try navigating on a windows machine without a mouse; after a few minutes it's pretty easy. Try that with GNOME. Switch between apps and see if the keystrokes are consistent. see if the focus goes where you think it will in a complex form. most of the time, GNOME keyboard shortcuts are implemented as a complete afterthought, and it shows. If there's a GNOME standard for this, it's followed poorly.
Additionally, most of the original desktop themes are just plain useless. they're:
- pretty but useless [themes.org] -- how do you maximize one of those windows?
- geeky and useless [themes.org] -- okay, tell me where i click to minimize, maximize, close, or stick the windows.
- just plain dumb [themes.org] -- who the fuck knows how this works? i sure as hell don't, and I have better things to do than figuring it out.
And the scary thing is all these themes have 10k+ downloads, and it took me about a minute to find these examples. I'm sure if I really cared I found find much better examples of what I'm talking about.I just spent 15 minutes looking for a truly good theme somewhere without success. that's a tragedy. that will hurt linux's mainstream acceptance far more than the fact that cmdrtaco was too dumb to buy a supported scanner.
--
agreed (Score:4)
For example, why does everybody copy the design that the 'window kill' button should be right next to 'maximize'? That's horrible design, put window kill on the left, maximize and minize on the right.
Why is it possible to click down on the 'K', move the mouse a few pixels up, release the mouse, and log yourself out. If you have a fast computer, and you use KDE, you've probably done this before.
GNOME allows an application to use the entire task tray, then when you have two applications, it uses half that size.... and it squeezes down. It's efficient use of space, but it's inconsistent and makes it harder to tell with a single glance what's running. KDE makes good use of the space without this annoying inconsistancy.
What the hell are these icons? Stop being cute, start being useful. If you're running KDE, hit the K menu now and tell me what the following icons mean 'quick browser', 'bookmarks', 'toys', 'system', 'preferences' (these last two are way too similar), multimedia or graphics. None of those icons gives you any intuitive notion of what you're about to launch.
Additionally, I doubt I'm the only one who has taken the less-used apps in the menu for each level, made a folder called 'sewer' and stuck them in there. Yes, we're all proud that there are lots of applications now. No, we don't use 90% of them, and having them in our menus just slows us down.
Things are improving, but it's still terribly ironic, the way copying Microsoft is referred to as innovation, yet when Microsoft copies, that's just plain wrong.
--
Yes, MS likes mental domination (Score:3)
MS uses the term "shortcut" where everyone else in the world uses "link". Not just those lame "shortcuts" in the filesystem, but IE uses that term for HTML links. This is from usability testing? Yeah, right.
In FrontPage instead of having templates, like most HTML editors, FrontPage uses shared borders and themes, while "template" is used for something different. And it uses the term "web" where everyone else uses "site" -- the way they use "web" is simply stupid.
I can't recall any others at the moment, but everytime I use some MS product I notice these minor, strange namings. They often forgo convention to use their own odd words. They want to invent a lingo so everyone is confused when they try something new, just like these people were.
Trying to immitate that would be like using the Word .doc format for file saves -- it's hard enough just to import the crap, you can't expect to become the crap.
Paricipant 183: (Score:4)
"Where's Clippy?" (P183)
Anything new takes getting used to... (Score:5)
That being said interfaces which are reasonable can be adjusted to within a reasonable amount of time. Gnome is certainly something that those of average intelligence with the right amount of time should be able to get down... If they are interested in it and there is some reason to adjust to it.
So find a reason for people to use Gnome, and they will. (I am not saying Gnome doesn't have a use...)
Usability studies from Sun? (Score:3)
Re:OpenLook anyone? (Score:3)
As for open sourcing it, Sun did that before they gave up on it... it was kind of their last ditch attempt to outmaneuver Motif. Unfortunately it was too late. Had they done it about a year sooner it might have made a difference.
I used to use olvwm on Linux back in the 1993 to 1995 time period... I imagine the source code is still out there for it, but I don't think it ships standard with many distros these days, let alone is part of the normal installations.
Re:Context is everything (Score:3)
Tying the logo with the name might have helped on the terminal emulator problem too, although they really ought change that to "Command Line Prompt" or something similar.
-sk
Usability vs. Transparency (Score:5)
fundamentally important to not confuse two distinct concepts: usability and transparency.
Usability is directly related to the efficiency of
performing tasks and the ability to anticipate the
user interface for new tasks.
Transparency is the "intuitiveness" of the interface or system. It is primarily a measure of
how easy it is for a naive user to come into the
system and get a something done.
Transparency is intimately related to the experience of the users being examined. In a certain respect, it is a measure of familiarity.
Unfortunately, you will get high transparency
scores nowadays if you simply look and act like
MS Windows.
Usability is a whole other bag of onions. Some of
the features of a transparent interface are relevant in assessing usability, but only to a point. While transparency is something critical for new or casual users, it can be almost completely irrelevant to an experienced user. Once a certain level of familiarity is acheived, usable systems are those that make the most common tasks the most efficient to access and provide easy means of aggregating and controlling common
sequences of tasks. Emacs is an immensely usable
system that has a very low transparency score.
It is interesting to note that the Usability Principles in this study seem to be correctly labelled: they *are* related to interface usability. However, the assessment methodology seems to be primarily measuring *transparency*.
I'd say that this is a basic flaw in the study and
colors the recommendations highly.
It would be nice to see someone do a similar study
but concentrate on the power users and address the
issues around high performance usability.
don't talk unless you've walked the walk (Score:5)
Unless you've sat down and observed your interface getting tested with a usability professional or two who work with regular folks to see how the application works *in the real folks non-geek world* then you don't know what you're talking about.
Really, how can you argue with behavior-based experimental data that "this isn't how people behave"? Oh right - with unfounded 3l337 opinion.
Sure, there are other things we could do to better test usability - like have them spend a week or two with Gnome after this test, then test again to see how much they picked up.
but until you're doing testing with your own projects, until you appreciate that these are real people in the real world (that same world you think should use Linux as a desktop OS) then you're really missing the point.
cz
see www.usability.gov [usability.gov]
IBM Ease of Use [ibm.com]
The perennial Jakob Nielsen [useit.com]
Usability Professionals Association [upassoc.org]
Webword Usability Blog [webword.com]
Listen to the users! (Score:5)
This is not what you should take away from a user interface study. This *is* what the users see when first presented with the program. It really doesn't matter what the programmers/designers of GNOME think. If the user doesn't like it, then he doesn't like it! If he can't understand, then he can't understand.
A long time tenet of communication is that if there is miscommunication, then it is usually the fault of the communicator who hasn't adequately taken into account the audience. If we as programmers/designers aren't using the interface to *communicate* then it is *we* who are failing to communucate, not the audience who is failing to understand.
Why do you think that MS has slowly moved to simpler and simpler language? People don't need techo-speak to understand what is going on with the computer. Understanding phrases like "illegal operation" requires a bit of underlying knowledge about why such an analogy is being used. So why use it. Just say, "your computer just crashed, but it's okay. Just press that little button on the front of the computer so it can restart. Have a nice day!".
Re:Reminds me of my users (Score:3)
That's pretty silly when you think about it. A C: drive, the syntax (C:) etc... It's as weird as anything under Linux. It's just that users have learned this one since the beginning of time, er, MS/DOS epoch, so now they expect the same kind of sillyness.
We need to corrupt our youth at an early age so when they are exposed to the Windows world, they'll be like "Drive letters? How fucking primitive!" :-)
Re:Thank you Sun! (Score:3)
People who are worried that users taught bad habits will actually force them away from being able to write elegant, intuitive systems. Easiest is not always best, etc. (Hardest is not always best either). Seems like GNOME needs a "novice" mode, which like Windows 98 hides all the advanced stuff (like scary black windows my god!), but can be displayed with a click of a button (or by a permanent setting). This way both novices and advanced users can be happy. Maybe every GUI feature can have an experience rating, and the user can set what experience level the GUI should display itself up to.
People who never seen something else... (Score:4)
My girlfriend never used computers before we met together. And on my computer, there's only OpenBSD, FreeBSD and Linux. I showed her how to log in, browse internet, paint, play music, print photos, etc. Her desktop has icons for main applications.
And she's not lost. She can use the computer without any help. With Linux, FreeBSD or OpenBSD.
The last week, she had to work on Word 2000 at her daily job. She was totally lost, found Windows slow and ugly, didn't understand why the word processor had so many complicated buttons and menus, etc.
So no interface is more intuitive than another. It's just a question of what you are more familiar with.
-- Pure FTP server [pureftpd.org] - Upgrade your FTP server to something simple and secure.
I can't wait... (Score:5)
While the reality is is that not only are 'regular' users familiar with windows, but MS has spent significant resources studying exactly these issues and they are common even to those who are not familiar with windows. Many of these concepts (which the Linux community has shunned for years trying to avoid being like Windows) are going to have to be embraced (and extended) by the Linux community if they are going to gain any mindshare in the population.
-Adam
This sig 80% recycled bits, 20% post user.
My Own $0.02 (Score:3)
About Face - a MUST read (Score:3)
Fundamental is the difference between good software engineering, and good user interface design. The automobile industry recognizes this gap - most users don't see what engineering is embedded in the engine, drivetrain, etc, but everyone sees the smooth lines of the body. And the second you sit behind the wheel you can tell if the designers intelligently arranged the controls to be easily accessible and clearly read. NONE of this has ANYTHING to do with the engineering of the car - it could have a revolutionary new suspension system, but that's not what you see. The same gap exists between user interface design and software engineering. An elegant use of pointers is invisible to your user. The relative elegance of software engineering techniques means nothing for user interface design.
Programmers tend to be BAD at interface design - BECAUSE we understand more about how the machines and software work than the average user does. We know the box, so we don't think outside the box.
The book opens with a great discussion of a user's goals, which are usually NOT to recompile a kernel. Cooper says users want: to not look stupid, to get an adequate amount of work done, not be too bored. These goals are clearly not addressed by error boxes that pop up saying 'library x caused a page fault at
Thank you Sun! (Score:5)
Take this ball and run with it, someone! Don't just bitch about how "Microsoft has conditioned everyone to look for a Control Panel!" Who gives a shit? If your intent is to write software for use by the masses, you'd better be damn sure and write it so that the masses will like it and want to use it!
--SC
Re:Kind of sad, (Score:5)
Reading it, the comments seemed to be a lot of things like:
"This is ridiculous! The start button is a foot? What does a foot have to do with a start button?"
Read a little further. They gave the participants the very important hint that the foot is the GNOME logo and then:
Their guesses were all dead on. If you didn't know that the footprint was the GNOME logo, you'd be confused, too. Think of all the associations you can make with a footprint. Traveling, history...exactly what they guessed.
"Whoa? How come the settings are under something called "Settings"?? Where is the control panel?"
Your paraphrase lost the meaning of the original. Try this instead:
They were not confused that the settings were in something called "Settings"; they were confused that the settings were in "Programs". Sounds like a pretty valid complaint to me.
Microsoft has succeeded in making their own screwed up naming conventions the "standard" of computers everywhere.
No. The users' expectations you've quoted were reasonable and not centric to a Microsoft desktop. You found what you expected to. You completely ignored all the information contrary to it.
Re:agreed (Score:3)
If you choose a different window decoration scheme, you won't necessarily have this problem. For example, I like the "Laptop" window decoration. It puts the close-window button over on the left side, leaving minimize, maximize, stick, and the help button over on the right. (KDE2)
But you're right. It ought to be different by default.
--
Re:Thank you Sun! (Score:5)
I often get the feeling that the Linux crowd wants to convert the masses as opposed to winning them over. There's a big difference between the two.
Re:The business of agronomic design (Score:3)
Gnome or KDE interface (Score:4)
Why ? because the strength of the KDE look-and-feel is that it's a close copy of Microsoft Windows, and this is good for 2 reasons :
People who can use Windows are not disoriented by KDE (rah rah, old argument, I know ...)
Microsoft being all about "first user experience" (read glass and chrome on a desktop anybody can use more or less intuitively), they probably spent a ton of money on the design of the Win95 interface, so why not reuse it ? It's far from perfect, but you can be pretty sure it'll be accessible to the mass thanks to M$ money, and KDE reuses all that R&D for free.
M$ is not stupid, and they've been reusing the same old clunky Win95 interface for years now. IMO, that's because they know for sure it's what flies with the users. So, I like Gnome, it seems solid and well built, but I'm sorry to say, it'll probably lose the GUI battle because its single biggest flaw is its non-M$ look-and-feel.
Kind of sad, (Score:3)
"This is ridiculous! The start button is a foot? What does a foot have to do with a start button?"
and
"Whoa? How come the settings are under something called "Settings"?? Where is the control panel?"
Microsoft has succeeded in making their own screwed up naming conventions the "standard" of computers everywhere.
In the future, will all UIs have to have start buttons, control panels and taskbars to be considered usable?
The business of agronomic design (Score:3)
GUI enviroments simply aren't all that intuitive, period. There may be ways to make them more intuitive however this study, while interesting, appears to be more a measure of how similar to MS Windows, Gnome is.
This is not to say the study is without value. Certainly it is valuable, but agronomic design just isn't at a level where a user can sit down and intuit the functions in such a complex devide as a computer operating system. It just isn't possible. This is not a reflection on Gnome so much as a reflection on the study of agronomics.
That said, the study was a good read and did make valid points in it's recommendations. It's just important to keep in mind what was actually being analized.
--CTH
--
In my experience, (Score:3)
There are similar "stumbling" blocks that various users hit when switching platforms -- either from a text-based unix terminal to GUI or from a MAC to WIN32, or whatever.
For most "work" environments, if the user knows how to "click" an icon to run his software, thats about all they NEED to know -- outside of how to operate their software package. It's silly to expect a 50+ y/o client services secretary from company X with no GUI experience to "master" any windows/mac-ish interface. Their "job" is to be able to navagate their software, launch it and shut it down. Beyond that is really expecting too much.
-jhon
An excellent UI design site (Score:3)
A lot of it is common sense, but many X developers would do well to go through the site. Fortunately the GTK pushes developers in the right direction (build the tools, and you can implicitly enforce the standards), but we still have a ways to go for GNOME to be as consistent as, say, Macs were in the late '90s.
Oh, and M$ bashers will have many opportunities for chuckles here. :-)
Re:An Idea for a user interface rfc (Score:4)
- Look.
You see files here. Also, a Trashcan and the Internet
- Take Internet.
You can't do that!
- Drop files.
Where?
- Drop files in Trashcan. /
rm -rf
Done. 261792K deleted.
- Ah! Undo! Undo!
I don't understand that.
- Get files from backup!
I see no backup here
- Get backup from Internet
It is getting dark. You are eaten by a grue.
C:\