Novell's Releases Linux Usability Testing Videos

sp3298622 writes "Novell is releasing primary desktop research, including over 200 videos and analysis of usability tests, at betterdesktop.openSUSE.org. Vice president of collaboration and desktop engineering for Novell, Nat Friedman: As a programmer, it's sometimes difficult to know how ordinary people with no technical experience are reacting to your software. Linux people tend to know other Linux people. In these usability tests, we selected test subjects who were experienced with Windows, but who had never heard of Linux, and asked them to perform basic tasks using the Linux desktop."

Comments

Batten down the hatches

[ettlz]

You know that ruddy "Linux vs. Windows Usability: The Quake Installation Test" troll is lurking around here somewhere.

Please proofread.

[Anonymous Coward]

That headline is just embarassing.

Re:Please proofread.

[hahiss]

Yeah, I think it should read:

  Novells Release's Linux Usability Testing Video's

Hm.

[Captain Scurvy]

Linux people tend to know other Linux people. In these usability tests, we selected test subjects who were experienced with Windows, but who had never heard of Linux, and asked them to perform basic tasks using the Linux desktop.

Might this only result in the Linux desktop becoming more like Windows?

Re:Hm.

[xtracto]

Burn Karma!...
Sure...
I have seen the page and the different case studies, they seem ok but I think there were 2 or 3 cases that are a lot more common:

1. Scan a picture, create a new document and write something about the picture.
2. Move the pictures of your camera to the place where you save your pictures in the computer.
3. Engage in a multimedia chat with some friend (micrphone+webcam+text)

Of course every linux user knows [although some of they deny it] those are non trivial tasks in a linux distribution ;)

[I can hear the shout of a thousand Linux zealot moderators :) btw I am getting linux instead of Windows in my department computer this week, so nope I am not a n anti-linux freak]

Re:Hm.

[s20451]

The mark of a good user interface is that it is either intuitive (i.e., the function of particular widgets is obvious), or where this is not possible, that it is easy to learn.

Of course the user has to start with a basic amount of computing experience. But you would expect people with windows experience to do well when switching to macs, because the mac interface is well designed, even though it is not the same as windows. So the question of whether the average windows user can figure out the linux interface is a good one.

Re:Hm.

[I confirm I'm not a]

Might this only result in the Linux desktop becoming more like Windows?

Maybe... but...
There was a brief comment in an article in, I think, last month's Linux Format (UK magazine) (I'm at work, so can't get at the article, sorry). Usability testing had been done on Evolution, and it was observed that one volunteer repeatedly used the "send/receive email" when they wanted to create a new email. The testers realised that the traditional "send/receive" button was not particularly intuitive. To my mind, that's the kind of useful information we might well get from this kind of testing - not assistance in turning Linux into Windows 2.

I mention this only because I believe there's still hope ;-)

Re:Hm.

[databyss]

I believe your answer lies here: http://www.betterdesktop.org/welcome/reports/repor t-date-time.html

# Issues encountered:

1. The date and time configuration tool is not easily discoverable from the menus, and is not listed in Personal Settings.
2. Users assumed the root password request meant they had to log in as root.
3. Users wanted the click behavior of the clock applet to be similar to Windows.

# Recommendations:

1. Fix time and date settings to not require root access.
2. Add time and date settings to the Personal Settings window.
3. Make the clock applet behavior be more similar to that of the Windows taskbar clock.


Too bad the tests have such small sample sizes.

It would be good to get some big numbers on these issues, although I acknowledge that it would be very expensive.

Re:Hm.

[zootm]

Putting a CD into a Windows machine pops up a box, one of the options of which "Copy the music to your computer". Your boss must be really dumb.

Claiming it is non-trivial on Windows is fallacious though, it's certainly no more difficult. Windows Media Player is included with Windows (except in some versions in the EU!) and is capable of ripping CDs.

So much for this

[Raelus]

89% sounds like a very good success ratio for the date and time test. However, RTFA and you'll see that only eleven people participated, most of them female.

If you don't have a diverse testing population, you aren't going to produce meaningful results. The idea is fine and all, but the results are mostly useless.

Re:So much for this

[zyklone]

How many you use is not important.

It's how they solved it, you are trying to find problems not produce useless statistics.

Re:So much for this

[Twylite]

The required population size depends on what you're trying to test and how carefully you select your population. If you're trying to test the failure rate of moderately experienced Windows users performing tasks on a Linux system, and you can accurately identify and select moderately experienced Windows users with no prior Linux experience, then you only need a tiny population.

Testing the failure rate is important: 100% of 11 users succeeding at a task can give you at most ~ 90% confidence that all similar users will succeed. 1 of 11 users failing is a far stronger result, telling you that you can expect at least 9% of all users to fail.

Various usability experts suggest that as few as 5 or 10 individuals are required for usability testing, and the remaining usability issues are discovered and resolved via the bug reporting and maintenance processes.

Re:So much for this

[99BottlesOfBeerInMyF]

89% sounds like a very good success ratio for the date and time test. However, RTFA and you'll see that only eleven people participated, most of them female.

Eleven people is a pretty good sized group for a usability test. This sort of testing is pretty expensive and time consuming, it's not like a survey or something. From a group that size, you can get a pretty good idea of how the average person will try to accomplish a task and some problems they may encounter. I've worked on projects where usability tests included only three people to test the interface to a product costing tens or hundreds of thousands of dollars. I'm going to have to disagree that these results are meaningless. The important thing is how did the user who failed try to do the task. What stopped them? What problems did other users have?

Re:So much for this

[Bogtha]

89% sounds like a very good success ratio for the date and time test. However, RTFA and you'll see that only eleven people participated, most of them female.

So only one of them had problems? Sounds good.

If you don't have a diverse testing population, you aren't going to produce meaningful results. The idea is fine and all, but the results are mostly useless.

You shouldn't let the small numbers put you off. Respected usability professionals say you only need five people for meaningful results.

Re:So much for this

[frank_adrian314159]

However, RTFA and you'll see that only eleven people participated, most of them female.

Woot! And we got video of them, too!

Re:So much for this

[Taladar]

You mean the same "respected usability professionals" that publish on websites for which other people feel the need to publish an Opera user Javascript (like Greasemonkey for Firefox) to fix the readability?

Re:So much for this

[maxwell demon]

Did you know that there's recent evidence to suggest that the assumption that print media and online media differ in the effect line length has on readability?

Parse error: Expected verb, got '?'.

Finally!

[Animats]

This has been needed for so long. This is how it's done.

Developers, you don't get to check in code until you've watched the video of users struggling with your program. OK?

Re:Finally!

[Kidbro]

In all the places I've worked, it has never been the developers who have been unwilling to produce user friendly interfaces. It's usually the managers that decide that it's more important to squeeze in two half arsed features with poor user interfaces than conducting usability tests and produce one well coded feature with a good interface.

this is needed

[fak3r]

While I've seen things like this before, I'm liking Novell's approaches to Open Source more an more these days. With the excellent SUSE 10 (still may replace Ubuntu on my main workstation) and projects like Beagle and Hula, they're set to really make a splash if they take this useability idea seriously. They seem to be gelling more on the desktop than anyone else of late, 3 years ago who'd have expected Novell to be doing this? Awesome.

The day the internet choked

[rob_squared]

200 videos + slashdot link = massive conflagration.

Fortitude

[minginqunt]

It's a hard thing is to admit that free software has a usability problem. The natural temptation is to sit and watch these videos whilst screaming "You idiots! You don't click "Send and Receive" if you want to send an email! What's wrong with you?!?!"

It is difficult, but it's vitally important. These people aren't stupid losers- they are fluent in another operating system, where they can achieve whatever it is they want.

The problems on show here are ours, not theirs.

Martin

It's one thing to do an analysis...

[HerculesMO]

but another entirely to start working on a solution. The barrier to desktop Linux lies in simplicity, and without conducting a study or showing you a video, I can explain it easily.

Go to the web on a Linux PC (provided you've got a browser pre-installed), and download a tarball of say, Firefox. You are a Windows user but you're 'elite', so you use Firefox, and since it's available for Linux, you want to have the same browser.

You have downloaded the tarball, presumably to your desktop. You double click on the file, and it gets opened by Archive Manager. And from here, you can bet that 99% of the Windows folks that would like an alternative to their PCs will not make the adaptation to Linux.

It has to be EASY. Apple set the benchmark for this -- and if imitation is the greatest form of flattery, then do it. Who cares about inflating Apple's ego? If Linux makes a breakthru on the desktop because it's as easy to use as an Apple, or even as easy as Windows, how does that hurt anybody? The true geek can rely on the the commandline only distros, or drop to terminal to get their tasks done using regular expressions and grep or whatever they want, while the 'idiots' (and I would venture to say, that I'm one of them) can use the nice GUI that's simple to follow and easy to use.

Then folks, when developers see that they can cross develop applications that work in Linux (with little overhead), and that people will be able to easily use and access them -- THEY WILL. The open source community just needs to see that fact and start making solutions happen. With the extremely fast and accurate nature of Open Source, the feats that have come from it are amazing. It's more amazing, that the basis of Open Source -- Linux -- remains fundamentally unchanged to accomodate the eager Windows users (read: ME) to switch fully to Linux. Until the snobbery stops and changes start, Linux on the desktop is going nowhere fast. And that's upsetting for a Windows user tired of his OS, and not wanting to get tied into another corporate entity (Apple).

Missing...

[TerminaMorte]

Missing Tests:

1) Ooops! Find your kernels source, kill X, and install the drivers for your video card. Oh, and updated XF86Config. Or Xorg.conf. Whichever one you happen to have.
2) Damnit, another kernel panic. Find what obscure change caused it to happen this time!
3) Ah, so now you have a wireless card? Try to get it working! You might need to use ndiswrapper. If you get another kernel panic, go back to #2.
4) Ah, can't get above 800x600 resolution, eh? Yeah... find your monitors horizontal and verticle refresh rates. Google it, and you might get lucky.
5) Figure out how to resolve RPM dependancies. Shit, that package needs Python 2.4.2, huh? Ah well, 2.3.9 is installed. Guess you're out of luck.

All joking aside, this was a pretty intresting study. ;)

Re:Missing...

[MobyDisk]

The parent post was written to be funny, but it is actually very insightful. This biggest problem for Linux isn't where the menus are located, or how the icons look, or the confusion over the meaning of the "Send/Receive" button. Those things happen on all operating systems and all software. The users eventually figure it out. But Linux isn't even ready for that stage yet.

Linux needs to work on getting software and hardware to work together reliably. That means without having to edit configuration files and without going to a command prompt. Simple basic things are missing. We need to work on drivers, resolving dependencies properly, and making packages that just work (including installing icons and adding documentation).

After we get that stuff resolved, then tweaking the UI will become more relevant.

Re:Missing...

[eclectro]

I knew that's what somebody would come back with. Countless computers still have floppies and floppies are still being used. Otherwise, how would I know that the floppy doesn't automount?

You are either a troll or arrogant.

Mod parent +25.

[khasim]

Hmmmmm .... take people who are experienced with performing a certain function on Windows ....

Then put them in front of a different system (like say a Mac) and see if they have any problems performing that same function.

Of course the "easiest" (and therefore the "best") user interface will be the one that is as close to 100% identical to the only one they've used before.

That's great for Novell because they're trying to get a slice of the Windows market.

But this does not provide ANY information that any person could not get just by spending 10 minutes on a Windows machine and copying down menu locations and order and wording.

Re:Mod parent +25.

[TuringTest]

Of course the "easiest" (and therefore the "best") user interface will be the one that is as close to 100% identical to the only one they've used before.

That's false. The best interface is one that reflects the user expectations. The Windows interface doesn't reflect user expectations in many ways, so it's possible to create a better interface than one which is just identical.

That should be the aim for the Linux Desktop design, not just to attract former Windows users but to best serve previous Linux users as well.

Re:Mod parent +25.

[vertinox]

Of course the "easiest" (and therefore the "best") user interface will be the one that is as close to 100% identical to the only one they've used before.

I would have to disagree... I do phone support for a living and help people that have only used windows computers and are not that technically literate and when I ask them to look for a folder in a list they often will look at the files and expect it the entire list to be alphabetical order like on a Mac. I have to correct them and say to look at the top of the listing for yellow folders and look for the folder name there.

The funny thing is that these people have never used a Mac.

Apparently for those who have never used a computer the most logical expectation is that all the files and folders are alphabetical. When I first started using OS X, I found it quite annoying since windows always had it the other way around but take a person who has used windows but never bothered to actually learn it and you'll find they'll expect something else..

The problem of "it's good enough"

[ShatteredDream]

Prior to Windows XP, Windows did so well with the average user because it was "good enough." It wasn't technically the best, in fact 9X was technically inferior in many areas to even Linux circa 1995-1997. So here's the problem. If Linux cannot meet or exceed Windows in every area that matters to a user, why switch to Linux instead of staying with Windows or going to MacOS X? I have a Mac Mini, it could end up being a major threat to desktop Linux for the people out there who are less concerned with having all of their options open and more concerned with getting a system that is cheap, small and just works. If you're not going to use all of the resources available on a new system, why spend $800 for a new Dell system when you can pay $500 for Mac Mini? For the average user there is no reason to pay the extra $300 if they get the software they need.

Desktop Linux needs to grow up in a hurry. That means it needs to be as easy for the average user to use as Windows XP is by the time Vista comes out. I've used a beta of Vista and was incredibly impressed... and I'm a Mac fan first and foremost. Vista is a major threat to Linux and will solidify Microsoft's control, not end it, if things don't change.

Re:S'top the press'es!

[bjtuna]

I usually *am* a grammer nazi so this shit bugs me. I think the Slashdot editors need to put this poster on their walls:

http://angryflower.com/aposter3.jpg

Re:Define basic tasks

[krgallagher]

"Define basic tasks "

OK Here is (I think) the complete list from the article:

• Task: Add an entry in the addressbook
• Task: Make an account on this computer for your friend
• Task: Add a birthday reminder to the calendar
• Task: Change the wallpaper
• Task: Change the fonts on the computer
• Task: Please correct your computer's date/time
• Task: Create a playlist
• Task: Crop the photo of the cat
• Task: Email "A Tale of Two Cities" to arthur@ximian.com
• Task: Upload your photos to your website
• Task: Figure out if the computer is online
• Task: Find all the music by Bach
• Task: Edit the note that you made for yourself
• Task: Find out if the computer is online
• Task: Find the document "A Tale of Two Cities" on the computer
• Task: See if you have enough free disk space for a new game
• Task: Import the photos from the camera to the computer
• Task: Please login to the computer
• Task: Make the photo of the car brighter
• Task: Make a shortcut on the desktop for the addressbook program
• Task: Make a new photo tag
• Task: Make a note for yourself
• Task: Play the selected song
• Task: Remove some songs from the playlist
• Task: Please rotate the photo of the cat
• Task: Assign the "People" tag to all photos with people in them
• Task: Turn the volume down
• Task: Turn off the login music

I found it interesting that eight out of twelve succesfully completed the "Find out if the computer is online" task. I also wonder if these users could complete all these tasks in Widows.

Re:Excellent Idea!

[Smidge204]

What a Novell idea!

=Smidge=

Re:also check...

[kihjin]

That's what subm^H^H^H^Hpreview is for...

Not ready for desktop until usablilty addressed

[Waffle Iron]

It's obvious that Linux has user interface problems that are holding it back on the desktop. For example, most apps take '-x' style options. However, confusingly, many apps have now added '--foo' options.

What's worse, some frequently used apps don't conform to any options standards at all. 'ps' takes a confusing mixture of options, some with dashes and some without, which are mutually incompatible. 'tar' needs some options without dashes, and some with. 'dd' uses a totally different keyword-based scheme like 'foo=bar'. And 'find' has its own little expression language on the command line.

Clearly, grandma isn't going to be able to use Linux until all of these confusing option schemes are made more consistent.

Re:Excellent Idea!

[PsychicX]

Next week's article: Microsoft released a video of people who never used Linux, being asked to use a partly configured LFS system. The users were dropped at a command prompt and asked to simply fire up the windowing system, open the word processor, listen to music, etc. Every single user failed. MS points out that this is 100% accurate and solid evidence that Linux's TCO is 400% higher than Windows.

So ... everything should run like DOS?

[khasim]

It's a hard thing is to admit that free software has a usability problem.

No it's not. It's just difficult to accept that Windows is the best user interface. Particularly because the Windows user interface seems to be changing with every release now. Using that logic, the GUI would never have been accepted because everyone at the time was far more familiar and proficient with the old DOS system.

I have users who were quite skilled with Win2K who are lost with WinXP (until I show them how to make it look like Win2K).

So, which interface should Linux emulate then? Win2K or WinXP? Or Mac? Or something else?

It is difficult, but it's vitally important.

It is difficult and it is important ... but this approach is wrong.

This approach will give you completely different answers depending upon whether the group you select is familiar with:
a. Win2K
  or
b. WinXP
  or
c. Macs

These people aren't stupid losers- they are fluent in another operating system, where they can achieve whatever it is they want.

Yep. And so the "best" interface for Linux would be ... whatever the majority of Windows users are familiar with.

Novell could have saved all that time and money and just spent 10 minutes with a Windows machine, copying down menu locations, order and wording.

There is NO "usability testing" being performed here. No one will learn whether a specific Windows implementation of a menu is less optimal than a different one.

All that will be "learned" is whether those users can find the Linux equivalent and that will always be easiest for them when the Linux menues are 100% identical to the Windows menues that those users are familiar with.

Like fish in a barrel....

[gosand]

89% sounds like a very good success ratio for the date and time test. However, RTFA and you'll see that only eleven people participated, most of them female.

And we all know that programmers have no frickin idea how to satisfy a woman.

like shooting karma-fish in a slash-barrel. :)

I'd go a bit further than that...

[famebait]

As a programmer, it's sometimes difficult to know how ordinary people with no technical experience are reacting to your software.

How about "always completly fscking impossible" for more precision.

Not that there's nothing to gain from training and experience in usability design. Far from it: it will let you skip many obvious problems, and help you resolve others that users find for you in better and more efficient ways. But until your interface is tested on "real people" in at least a couple of iterations, there is no way in hell you can call it "good", "finished", or anything of the sort. If you don't agree, you've probably never done any real usability trials. There are always surprises. Often really big ones.

Your fine tuned detail somewhere may work just as plannned, but it will easily be swamped by problems stemming from inadvertent signals the interface is sending which never occurred to you, or from assumptions you never questioned or even spotted, which utlim ately make people (rightfully!) misunderstand the whole metaphor and do the wrong thing.

There are good news though: If you are willing to really really accept that the user is right (the way people percieve your product is in fact the way they perceive it, and you won't be around to explain to them that their thinking is wrong), and have set aside reasonable time to correct the problems you will find, - usability trials are fun!

Seriously. Fun, enlightening, and humbling (but in feelgood way), and they will broaden your horizons by illustrating just how differently from your daily assumptions it is not just possible but common to think. Do them. You'll like it.

Just resist the urge to explain the problem to the subject (except to be able to move on to test other things). Write down the problem in stead. The trial is for your instruction, not theirs.

Flash abuse

[sfraggle]

Has anyone seen what they've done with flash on that site? Their titles are all individual flash applets just displaying some text in a fancy (but ugly) font. For a Linux site this is rather appalling.

Re:S'top the press'es!

[Zwack]

But it ALREADY is in a semblance of English... I mean it looks like English, I can understand every word...

I just can't parse it.

"Novell's Releases Linux Usability Testing Videos"

Let's take it a word at a time.

Novell's Ah, it's got an apostrophe s on the end, so it's either possesive or contractive. As Novell is an entity I'm assuming that we're talking possessive here. Something belonging to Novell. Good start...

Releases. Well, this can't be a verb as we're expecting the noun that is possessed by Novell, so while it might be nice to think that "Novell Releases" is the start of the sentence, instead we're looking at somethings (it's plural) that Novell owns. So Novell's Releases. Some items owned by Novell that have been released. Excellent, now what about these mythic Releases...

Linux... This isn't so good. Linux is a noun, and not a verb... Three nouns in a row? It's probably not unheard of, but in this case I'm expecting a verb. I want to know what Novell's Releases do... Well, let's soldier on and see if the verb appears later... Perhaps Yoda wrote this.

Usability... Nope...

Testing... Hmmm, perhaps test is being used as a verb and the entire portion in front is being used as a compound noun as favoured by Germans...

Videos. Yes, that's it....

The "Releases-Linux-Usability" (whatever that is) owned by Novell is testing Videos!!! Are they testing VCRs? Video Codecs? Movies? Perhaps if I read the article it would tell me.

Or perhaps they REALLY meant "Novell Releases Linux Usability Testing Videos" NAAAAH!

Z.

Re:Batten down the hatches (Quake Install Troll)

[forkazoo]

So, I wanted to install Quake. I use Linux a lot. I'm no expert, but I'm pretty comfortable with it. I've heard Windows was great for people who want to play games with no hassles. I knew microsoft made it, so I pointed konqeror at microsoft.org, and away I went. Ooops. MS is a company, not an organisation. Microsoft.com! Okay, so I tried to search the website for an .iso, so I could install Windows. Nothing! I realised that it was Commercial Software. I should have known this upfront, but I'm no windows expert. So, for the privelige of *playing games,* I went down to a local computer shop and invested over a hundred dollars in a copy of Windows. (I guess there are a lot of hard core gamers who wouldn't have a problem with putting down hundreds of dollars just to play games, but it isn't something I normally do.)

Installation was pretty smooth. I had to download nvidia binary drivers to get fully accelerated OpenGL, just like Linux. Windows is a supported platform for the drivers. I had to reboot the whole OS after installing them, because Windows won't let you easily drop back to a command line mode and just restart the GUI. No worries - I didn't have a server running on the machine, and it only takes a bit longer to reboot than to just restart a GUI.

Caution - Windows only comes with a special limited feature browser that doesn't support tabs, or anything. It is apparently only provided so you can download the latest version of a real browser after you install Windows. Windows doesn't come with a lot of useful stuff that you expect from a Linux distro...

So, I start reading docs to find out how you install apps on this new OS. I was having a pretty good time. Then, I learned that there is no equivalent of apt-get. If there is free software you want to download and install, you have to do it manually. So, I used the funny miniature "IE" browser to get the Quake source online.

Ooops, bad idea. Windows doesn't come with a compiler. You can download a free version, but the full featured "Visual Studio," costs a lot of money. I didn't feel like investing the effort to understand the differences. I decided to just get binaries. Again, there is no tool to automatically download and install an app, so I had to manually google for windows binaries. Thankfully, Quake is a very popular game, so it was very fast and easy to find, but still, it is an extra layer of inconvenience.

After a flurry of clicking "next" and eventually "finish," I finally had the game installed. Hooray. I tried to run it and I got a "BSOD." (Crash error screen) Of course, I already pointed out that Windows comes with no development tools, so it wasn't like I could try again with the debugger to see what happened. I had no way to see exactly what the issue was. What's worse, I couldn't get back to the system. This *game* had caused the equivalent of a kernel panic. It wasn't just the app that had crashed, but the whole system! this, from a system that is supposedly really great for games! It lets a game kill it!

Okay, so I rebooted into Linux. I already knew of a website with binaries for Quake, so I went there in Konq (Which came installed by default! I didn't have to go and download it!), downloaded a package...

dpkg -i Q

That was all there was to it. This "Windows is great for games" garbage is just horrible propaganda.

Now, if only I could get sound to work in Linux...

GNOME is horrible broken!

[Anonymous Coward]

The problem is not the contributions. The problem is getting those contributions accepted by the maintainers.

Over the years I realized that the request of contributions is just a poor excuse to avoid conversations with the developers or users who want something to get changed.

Some stuff in gnome-vfs for example was so utterly broken that it wasn't touched for a really long time. There wasn't even a maintainer for it (only a guy who kept putting some stuff in there whenever it was needed). Now some other people seem to have taken over the maintainance of it and the process continues.

But within the GNOME development team I found out (due to own experience) that it's quite difficult if not highly impossible to get some ideas through or to convince a developer that a different approach would have been wiser or better. Not to say save a lot of time. But people kept using the broken components for years.

Even now not everything inside GNOME is sane or reliable and a lot of stuff seem to be reinvented over and over again. See DBUS for example or basic things like "specifications" as found on freedesktop.org. GNOME makes freedesktop.org sound like it's a place for developers from GNOME and KDE to met and declare specifications but this is not always true since KDE had solved most of the necessary things that GNOME still urgently needs years before and their specifications and solutions are often by far better thought through and much more mature - and over the years proven that it also works practically and not just as concept.

For example you can compile KDE with a static prefix in say /opt/kde3 and later on you can move this entire directory to /usr/local/kde3 without the need to recompile anything. On GNOME we sill have the issue that every path is hardcoded inside the binaries so you can't move the entire location if necessary. One of the bad concepts of GNOME.

Another bad thing about GNOME is that the developers do have nice ideas at time but they lack the power or durability to make the changes or visions they have complete. GStreamer for example is indeed a nice technology and it somehow made it's path inside GNOME but still it doesn't feel like it's truly part of GNOME since some apps use it, others avoid using it and stick to xine. Now if these apps stick to xine then chances that GStreamer gets fixed and a whole part of GNOME is low.

Another thing is that plenty of the developers seem to have rotating focus on stuff. Today they work on this one, then tomorrow they focus on hacking on Mozilla or hack on 'dead ideas' they have that no one really takes serious so all the resources of working and fixing GNOME get's lost with playground stuff.

We all know that GNOME was meant to be a corporate desktop. But then a corporate desktop needs different resources and a different approach. Serious project leading is required, strict guidelines are required, and people with brains to enable them.

It can not be (now that the HIG as guideline exists for some years) that applications developer still ignore it. I don't care for third party stuff. But I do care for the important and key elements of GNOME software that should be a good example and follow these guidelines.

GIMP, DIA, Evolution, Abiword, Gnumeric only to name a few are in no way HIG conform. Some are, but others not. I filled in a bug for Gnumeric not long ago pointing the developer to the HIG v2.0 where it says that the Toolbar should obey the rules of Toolbar & Menus capplet (which is a core part of GNOME) unfortunately the bug was closed as not a bug and no further comments have been given to it.

Also printing is a necessary importand thing in GNOME imo and it can't be that I load up GThumb to print a *.gif file and it ends up in printing a totally black picture on a white sheet of paper, wasting nearly 1/3 of my black ink cartridge.

It's also inacceptable for a corporate desktop to have a document reader and viewer like Evince that prints a whol

Re:Desktop tasks? or config?

[Reducer2001]

I must commend Ubuntu for making Samba a snap to get up and running using their GUI. Same with printers and disk drives, it's all done via a 'Wizard'.

the driver hurdle

[mcraig]

Personally I think the hardest task for an average user to perform on linux at the moment is driver installation, and lets face it for most people getting all their hardware working is the first step towards adopting a new system. I recently tried installing drivers for my ATi Radeon 9800 Pro a pretty mainstream card from a well known manufacturer, needless to say it's not straight forward by any stretch of the imagination. You simply can't expect joe public to ever learn how to compile his kernel and even messing around with kernel modules is probably asking too much.

Linux is certainly making progress synaptic does a great job of alleviating dependency hell and almost entirely masking it from the end user. I'd like to see the linux community not necessarily looking to emulate the functionality in Windows or Mac OS X but instead looking for what would be the most elegant solution. Perhaps something like an online database of drivers that manufacturers could update, which could be automatically 'pushed' onto your computer overnight and silently rebooted (with your permission in a preferences box) so that you don't even have to worry about having the latest drivers it all becomes automatic would be neat, in the event it failed to reboot it could roll back to the previous driver and notify you in the morning of its attempt.

You could allow users to rate drivers and add the ratings to the database, this way you could specify you only want to automatically update to new drivers that are rated 3/5 or higher for example. This could be like linux's answer to Windows update only better.

Not with the testing they're doing.

[khasim]

That's false. The best interface is one that reflects the user expectations. The Windows interface doesn't reflect user expectations in many ways, so it's possible to create a better interface than one which is just identical.

Not with the testing that they're doing.

You are correct, in theory. You are incorrect in this specific instance because their testing procedure will not yield the information necessary to find a "better" interface.

That is because they are only testing prior Windows users.

Those Windows users have been trained to seek certain items in certain places.

Even if you added a button that said "Complete this test with one click", the users would NOT find it unless they could not FIRST find the Windows button/menu that they were trained to look for of if that button was in that location.

That should be the aim for the Linux Desktop design, not just to attract former Windows users but to best serve previous Linux users as well.

Again, I agree with that, but that will not be achievable through these tests.

Microsoft Word used to have an option to use the WordPerfect keystrokes. This was because the people with the most experience found it very difficult to maintain their productivity while learning a new system. Even if that system was "better" for other people. Back then, the most experienced and productive people had spent years learning WordPerfect for DOS.

Novell has learned nothing in these past years. To make it easy to migrate users, you make it an option to have an interface that is 100% identical to what they are familiar with.

Real "usability testing" requires more people with more experience levels on different systems, including people with little or no computer experience at all.

If you REALLY want to make the system easy to use, you have MULTIPLE options:

# 1. Basic level. Almost no menus and lots of "I want to" included in the icon's name ("I want to send an email to someone" or "I want to look at web sites").

# 2. Emulation level. 100% Win2K look-alike.

# 3. Whatever other interface you design.

The key is to build the interface to the user and what the user expects/knows.

Where's the data? What about KDE?

[10Ghz]

For example, we ran comparison tests on KDM and GDM to figure out which program was more usable for members of our target audience.

Sounds interesting! But I can't find any data regarding that comparison. Sure, there are tests about logging in, but no data about comparing KDM and GDM.

Here you will find over 200 videos of people using Mozilla Firefox, Evolution, Open Office, Banshee, F-Spot and other applications.

3 Gnome-apps, 2 neutral apps. Where's KDE-apps? Looking at the data-section, I see this:

A test that deals with changing the background can involved things like Nautilus or GNOME Control Center

Again: Where's KDE? Going thropugh the test data I see that every single test was with Gnome. Where's KDE? So instead of being called "Better Desktop", maybe this should be called "Better Gnome" instead? then again, what can we expect from having a Gnome-guy running the show? So much for equal handling of the desktops....

Re:Where's the data? What about KDE?

[tigert]

The problems discovered were pretty much valid stuff for either desktop - and we tested both desktops. Once the first wave of geeks has lost interest in the site and it passes the frontpage of slashdot, you might want to go back and discover that there are KDE videos up too, if that is really important.

Both desktops suffer from similar problems anyway. Test one and it very likely applies to the other as well.

Also, more importantly, whatever was tested was a mix. It was a distro with both "KDE" and "Gnome" applications - whether it was "gnome panel" or "kde kicker" that displayed the main menu is pretty much irrelevant when the user doesnt find the right menu entry. The labels are the same anyway since they come from the software packages.

It's about time we started to look at the real world: people use a diverse mix of software - whatever does the job, whatever their sysadmins or distro packages for them. We should care about the whole "free desktop", not just one single project really.

It's No Less True

[aarmenaa]

I should start by saying that I use Windows on my primary desktop. For a while, I really tried to switch to Linux - Unbuntu, Mandrake, Slackware, and Fedora have all been on my box at one time or another. For the most part, getting the OS installed isn't rocket science. What is difficult is working around all the crap once you get a Linux distro installed.

Getting the desktop to look like anything except blurry ass requires an hour of reading about how to install your video drivers. Why? Because after installing your package using the really nice script, it still doesn't work. So you google again and figure out you need to edit that ghastly xorg.conf file. And then Google to figure out why the resolution is stuck. And then Googling again to figure out why the refresh is stuck at 54 Hz and giving you a massive headache. Dual monitors? TV out? You may as well just go cry yourself to sleep unless you're an uber-leet nerd, because that stuff takes hours to set up. That shit is a matter of one click in Windows; my mother can do it.

Then there's networking. Support for your wireless adapter may or may not even exist. If it does, it's probably in one of the generic Prism2 drivers or something like that. Great, but it doesn't help me a whole damn lot - mine says Netgear on the front. Back to Google again. It's also intresting to note that Linux's DHCP client and the server in my Linksys didn't get along real well, even on a wired connection. There's no way someone who doesn't know how that crap works would be able to troubleshoot that.

Of course, there's always multimedia playback, right? The install I liked best so far, Unbuntu, couldn't play anything out of the box. I know it should have been able to, but for whatever reason my install was futzed no matter how many times I reinstalled it. I never could figure out how to make it play videos. There were several settings for decoding and such (as well as about 10 different players to choose from), but nothing seemed to change no matter how I tinkered with those settings. Oh, and Unbuntu comes with several options for audio input and output including ALSA and ESD. WTF is the difference? I've heard of ALSA before so I'll use that one. Oh wait, that one doesn't work, but the ESD one does. Well, as long as I hear sound I don't really care. At this point, I don't even want to Google it.

This is why there aren't more Linux desktops: there are severe usability issues. I find it easier to get a webserver complete with PHP and MySQL up and running on Linux than a desktop. Why? Because I don't need video drivers, audio, or wireless networking. I also don't change my server hardware every month or two. Linux makes a great server, for sure. But as great a server as it is, it's a shitty desktop. And you'll please excuse my anger, I just got finished configuring my Linux install and promptly broke it...again.

Here's what desktop distros should be working on:

• -When it says it's installed, it'd better work (video drivers)

• -Drop the funny names. Yes it's superficial and shouldn't matter, but it does.

• -Make the defaults work. If the driver's there, the comptuer should play sound. And it should always be able to play video out of the box. And at least try to support the mouse wheel. I use mine a lot, and they come on every modern mouse. Why do I need to Google to figure out how to make it work in Linux?

• -Where possible, make it one click. Things like multiple displays shouldn't be so hard. Things like resolution and color depth should be changeable in an applet, not a config file.

• If a luser asks how to do something and your response is to call them a retard and tell to open up a terminal, your software's fucked. Lusers don't ask hard questions, which means that what they're asking for is a fundamental basic. And you just buried it under a shitload of command line.

Re:S'top the press'es!

[maxwell demon]

Well, you need to add to your grammar rules the following: There is no noun which cannot be verbed.
If you apply that rule to Linux, then it's clear:

• Novell's Releases: well, you already correctly parsed that at the beginning of your post.
• Linux: That's now the verb.
• Usability Testing Videos: Those are obviously videos related to usability testing.

Ok, now we just have to find out what it means to linux a video ...

I think I understand what I thought you said

[Bork]

You can not design something without the average user being part of the process.

Where I use to work the software development people sometimes were not engaged in what was happening on a manufacturing side. They "developers" thought they new how to do the manufacturing technician job but it was of those, "I think I understand what I thought you said".

They would start a job trying to get a specification together and so the people they would talk to were the managers of the manufacturing technician. Well guess what - they did not really know the job ether and what was ended up being developed would drive the technician up the wall with how things in there words was "screwed up".

What happened on latest projects was before getting to far into the project spec, they also included the technician in the interviews. Then once a somewhat rough spec was put together and some idea of the direction it was going. The next step was to videotape the technician doing the job as it was currently being down. One month was spent on just taping various people doing various aspects of the job. Each taping session went through a post-mortem review with all parties involved, the spec writer, the software developers, the managers, the technician, and anyone else they could drag into the meeting. The tape would be gone through and question like "Why did you do that? That's not written down anywhere" would be said every five minutes. Even the managers were asking what was going on.

What was brought out in all of this is that unless you are actually doing the day-to-day job in manufacturing, you do not understand the process no matter how many design meetings you have with them. This became the standard method on following projects.

Re:Batten down the hatches (Quake Install Troll)

[Total_Wimp]

Well, except for the fact that your $299 PC didn't have a good enough video card to play the game in question. It looked really bad. You went back to the store to buy a better video card and came home with a nice one for $200. It wasn't top of the line, but it met the specs on your game box.

You followed the intallation instructions to a "T" but the card didn't fit any of your expansion slots. You went back to the store to get a different one, but no one was knowledgable enough to help you out. Finally some kid in the aisle overheard you and explained about AGP and PCI-Express. He steered you to the right card.

After following all the instructions you finally get your game set up, but the graphics look crappy. You complain that your $200.00 card isn't even as good as your PS2. You enjoy bad graphics until Xmas a half a year later when your nephew explains the concept of "Native Resolution." You love your gaming PC now and just think, it only took you half a year to get it right!

TW

P.S. I'm a frequent Windows PC gamer, but I don't have any illusions it's as easy as you make it. Newbies have a steep learning curve.

Re:Batten down the hatches (Quake Install Troll)

[saintp]

I have never been required to re-compile all of my system binaries, patch the kernel and update all my dependancies in Windows in order to play a game, however, in Linux, this is standard practice.

What the hell kind of games do you play? "MTV's Add SATA Support to your Kernel XXXTREME"? "John Madden All-Pro Glibc Updater 2006"? I've never had to do anything close to the rigamarole you describe, and nothing even close for games.

-1, Troll.