This is a read-only archive. Find the latest Linux articles, documentation, and answers at the new Linux.com!

Linux.com

Feature

Open source usability is a technical problem we can solve on our own

By Frans Englich on July 09, 2004 (8:00:00 AM)

Share    Print    Comments   

Poor usability is a huge barrier to wider open source adoption. Our backends have matured and we consistently achieve technical excellence. Usability is the one area we have not yet mastered. For some reason, we treat it as a mystery instead of looking at it as a problem we can solve the same way we solve all other technical problems.

As a participant in the KDE project (but expressing my own viewpoint here instead of speaking for KDE), the approach I have seen so far to our usability problems is... noise. Ideas are raised daily on the KDE usability email list, but they never seem to generate anything but endless discussions. Developers, users and reviewers all scream that something needs to be done, but apparently no one knows how.

Judging from blogs, email lists, and articles I see, these seem to be the most called-for solutions to our usability problems:

  • Usability Reports
  • Usability Laboratories
  • Usability Experts
  • Companies Specializing in Usability

In other words, many developers seem to believe usability is such a complex matter that the open source community is unable to handle it; that we need outside Companies, Experts and Laboratories to manage this mystical matter.

Programming and "ordinary" engineering are areas in which the open source community has deep experience -- stretching over almost three decades -- but usability is a relatively new matter for us. How we react on usability issues, by seeing them as something mysterious we are unable to handle and someone else can and must master, is similar to how human beings historically reacted to phenomena we didn't understand. Lightning was explained by Thor's Hammer, the plague was a punishment from God, and so forth. In our case, we replace "God's will" with "Companies", "Reports" and "Experts." We don't understand usability, so we push responsibility for it onto someone else.

Too big an issue to rely on outsiders

Even if we decide to rely on outside experts to solve our usability problems, they are going to find it impossible to keep up with us. The KDE project alone has an average of 200 checkins to its code repository each day. There aren't enough outside usability specialists available to correct all the errors that are inevitable with this level of productivity.

One of the advantages of open source is its ability to put the consumer ahead of profit. Our goal is to produce great software while honoring the user's privacy, rights, and freedom. When usability, central to everything in today's software, is outsourced to companies, the open source community's independence and opportunity to achieve its noble goal is compromised. The open source community must be able to handle all its issues -- including usability -- by itself in order for our development approach to give maximum benefit to society and the user by constantly advancing our level of technical excellence.

"Usability Reports" sounds like a possible solution, but somehow, when usability issues are encountered, open source people seem to conclude with a deep sigh that they cannot be solved since we don't have the resources to study usability properly. But we don't need formal studies. We simply need to apply our own problem-solving skills.

Here's an example: Konqueror, KDE's file and web browser, has a menu entry called "smbUmount." I don't need a laboratory with video gear to figure out that this is nearly impossible for non-hacker users to understand.

All it takes is to think once about each little item like smbUmount. If the changes to this and other items that are obviously not user-friendly are made, most of our usability work will be done. We don't need usability reports. We need each developer to devote as little as one single thought to usability.

We need to teach -- and learn -- usability

If I want to learn programming in Python, device driver development for Linux, or object orientation, I can consult countless HOWTOs, online books, FAQs and email list archives. If these sources don't answer my questions, I can ask the community through an IRC channel or email list and almost always get a reply based on the community's decades of collective wisdom.

But If I want to learn how to write phrases understandable by users or what colors to use that still allow color-blind people to use my software or how to best name categories for efficient navigation, I can do nothing but listen to people's opinions in the matter. Where is the open source community's pool of facts and knowledge covering usability issues?

Discussions on our usability email lists are noisy, full of anecdotes and not so humble opinions. We cannot tell each other to RTFM (read the fine manual) because there isn't one. All we have are our style guidelines. Guidelines are "do's" and "don't's," but not rationales. Guidelines are a convenient way of steering usability development that work well in the areas they cover, but as soon as we get outside those areas, development drifts. Then we need knowledge to help us make wise decisions.

Our community has little experience in usability and designing graphical user interfaces, and the way we approach these matters today gives us no chance to teach ourselves. No wonder the result screams for improvement, our discussions are nothing but rants, and we leave the mysterious problem to the Almighty Gods of Usability: The Experts.

We need to build our own pool of knowledge in the usability field. We need online books about usability, published under open source licenses. We need HOWTOs, interviews with project managers, and articles discussing, questioning and driving usability. Then we will be able to work on our problems, since once we have the base knowledge, usability will be revealed as the engineering science it is. Then our usability discussions will no longer be long anecdotes and personal opinions, but will become problem-oriented, and will be discussed in terms of right and wrong.

Once we build a reservoir of open source usability knowledge that rivals our pool of programming knowledge, open source development will not only be the best way to achieve technically excellent software, but will also become the best way to produce usable software.

Share    Print    Comments   

Comments

on Open source usability is a technical problem we can solve on our own

Note: Comments are owned by the poster. We are not responsible for their content.

Grokdoc

Posted by: Anonymous Coward on July 10, 2004 01:11 AM
www.grokdoc.net could be a good resource . . .

#

Re:Grokdoc

Posted by: Anonymous Coward on July 10, 2004 01:26 AM
I joined GrokDoc. Seemed like a good idea but seems way too focused on User Experience than APP/System Usability.
Usability is NOT just important for the GUI's. IMHO its pretty important for NON GUI apps in the sense that they have easy configs & such.

#

Re:Grokdoc

Posted by: Anonymous Coward on July 10, 2004 02:21 AM
Isn't application/system usability based on improving user experience? And isn't a poor user experience indicative of a need for greater usability?

Just a thought.

#

Re:Grokdoc

Posted by: Anonymous Coward on July 10, 2004 10:38 PM
I think what grandparent ment, is that system administrators, programmers and other unreal users (as opposed to real users) enjoy good documentation and sane behaviour too.

#

The Reason why

Posted by: Anonymous Coward on July 10, 2004 01:26 AM
The reason why there is no usability data pool is that there is no data to be had. Useablilty is simply a matter of taste and choice, that's why when you ask a quesiton about useablilty all you get is opinions. Useablilty labs are like marketing labs; they are there to gather opinions from focus groups on useablility.

The closest that open source could come is to let people vote for interfaces etc. The only "best practice" is what do most people find most useable.

#

Re:The Reason why

Posted by: Anonymous Coward on July 10, 2004 02:37 AM
Useability is not a matter of taste -- "taste" is an esthetic, not functional consideration.

For instance, it's not a matter of "taste" or "opinion" whether the menu item mentioned in the article is overly technical and obscure for the lay user. It simply is incomprehensible to to anyone who doesn't know the term. It's a trivial task to give any menu item a self-explanatory name that any non-technical user can understand, and should be a matter of course for all developers no matter what platform they're developing for. Communicating in a language that the user understands is not a matter of taste, it's common sense (and common courtesy). (You wouldn't design a menu in Turkish for American users, would you?)

Similarly, it's not a matter of "taste" when it comes to what color schemes or screen layouts are optimally readable by sight-impaired users: those are proven technical issues. It only takes a bit of effort to search the web for guidance, or to contact organisations that advocate for those with special needs: many of those organisations have already done the useability work and are only too happy to advise developers.

"Choice" (and yes, ok, taste) can of course enter into which design scheme / menu structure / descriptive language to use if there are several choices that might serve users' needs equally well (or choosing to ignore users' requirements for whatever misguided reason).

In any case useability professionals definitely have their place but they're not necessary to design efficient, enjoyable, eminently useable applications.

I'm a developer. I don't have a degree in useability, nor do I consult with specialists when designing application interfaces. When I'm devising an interface I think about:
how to guide a user through the process of completing their task so that they provide the information needed in the order required to get the result they want;
with the clearest, cleanest visual experience;
presenting options and explaining steps in the easiest to understand, plainest language (hopefully with an occasional tasteful dash of humor).

Sure it takes a little extra work, but it actually helps the development process (highlighting trouble spots). Anyone can do it if they choose to put themselves in the non-technical user's shoes. It's not rocket science. Like I said, it's just common sense and common courtesy.

#

Re:The Reason why

Posted by: Anonymous Coward on July 13, 2004 08:09 PM
It's not a matter of taste when you're presented with a list of options, none of which make sense (or all of which seem to mean the same thing), and there's no accessible help button. Or the help button is on the previous screen (and it's useless anyway when you get back there) and the screen you're on is modal so you can't get back without hitting "cancel". (That was my experience last night of K3B, which seems to be totally useless<nobr> <wbr></nobr>...<nobr> <wbr></nobr>:-(

Read ESR's recent rant about CUPS for a perfect example of how to confuse *intelligent* users by actively hiding all the information they need to make a sensible decision<nobr> <wbr></nobr>...

Cheers,
Wol

#

Wrong...

Posted by: Anonymous Coward on July 10, 2004 04:01 AM
Sorry, but this is totally untrue. Could I recommend that you read Card, Moran and Newell's 'The Psychology of Human-Computer Interaction' (1983) or Noyes and Baber's 'User Centred Design'. Admittedly both were written long before the term 'usability' was coined, but they demonstrate a lot of empirical research that is applicable to usability. The field of usability is a multi-disciplinary field composed primarily of psychologists (after all, it is humans that are being measured), but also computer science, discourse analysis, graphic design and others.

I think maybe you confuse usable design with graphic design. It's a fairly common misconception which even large companies make.

#

Re:The Reason why

Posted by: nikolic on July 10, 2004 10:45 AM
Usability is the difference between finishing a task in a few minutes with a low but reasonable learning curve, and spending hours without the ability to finish the same task.

Usability is the core reason that MS Windows pulled ahead of any other OS from 3.x onwards. Microsoft Press has published numerous titles regarding the "Windows User Experience," in order to educate developers and their managers on the importance of a user having consistent and adaptable experience throughout their use of a vendor's software. Compared to contemporary competitors, Windows 3.x was crushing in all aspects of usability and has since been top contender. Regardless of MS business practices, this was their real strength as it was always just a little more convenient to use their products.

Consider MS Office, while it is possible to have some considerable problems to have it do what you want, most users will never know of them as their use is streamlined to their expectations of something like a typewriter. This is supported with interface elements that meet this expectation rather than requiring more direct knowledge of the application. A user needn't enter the em value of their word spacing, or the point value of their tabs. They can drag the options to their needed value and get back to the important things. Sure, you can set all of these things programmatically, but to expect a user to do this is unrealistic. Mention VBA to them and they will correct you saying the fish-fry is actually at the VFW.

This is part of the brilliance of Mozilla Firefox mimicking IE in the placement of Internet Options and KDE and GNOME provoding a Start Menu. Users are far closer to being "up to speed" immediately and are more likely to stay and contribute to whatever application they use.

While improving, Linix is most useful to developers and still abandons many users that would otherwise be interested.

When it comes down to it, people that need something to work --- particularly those that depend upon computing for business --- view initial and even yearly licensing of $500 or more as small change in the face of spending 2 weeks picking up a previously unneeded skill. They may have lost many times that cost by the time that they acheive their needed results.

It should also be considered that many simply can't afford to participate in a MS monopoly. Populations that have the most to gain from Linux are left to learning an entirely new vocabulary on top of an unfamiliar method of production. I recently had a visit from a cousin of mine from the 'old country'. She had never seen a pool table except on the only TV in her villiage. This TV was the richest household in her villiage of about 250 people. I don't expect her to know anything of using any computer --- let alone figuring out how to configure her new Linux kernal.

While man and help pages are wonderful for seasoned *nix users, they are useless for those that are new to the platform. While many a *nix user will boast (almost angrily) that they can enter a command in seconds that commits system-wide changes and that is why it is the best and most usable.

Seasoned users simply forget while saying this that this is exactly what horrifies most users. The command that they misspelled, forgot exactly how to use, or forgot completely can cost them weeks.

In the mainstream, people expect a computer to act as any other appliance would in their home. They don't need to know how fast their spin cycles run in their washing machines and shouldn't mess with it either lest it put them in very serious danger while toying with the washer. A washing machine needs two dials to do the job and users benefit from it.

Jumping back to the thoughts of the requirement in Linux to use the command-line: what holds a developer (corporate or individual) from offering the ability to install a package in the place that a user wishes rather than default locations that are actually numerous, and offering it with the choices of "point and click" progression in place of command-lining rpm, or make, or anything. For instance, offer a command's options in the form of check-boxes at the bottom of a dialog box on screen. At this point a new user wouldn't have to study to again study to use another piece of software, they could simply install it to study the application itself --- which is the point, after all.

Possibly creating a machine-based disclosure of the host machine's packages or hardware, and so allowing the right build every time without any user ever editing a text file; requiring such an edit --- even customizing a script or config file after installation --- should be considered embarrassing. If Linux and other OSS is encourage migration, then the use of that software should be at least, encouraging rather than discouraging.

#

Re:The Reason why

Posted by: nikolic on July 10, 2004 10:47 AM
For instance, the form of my original post failed to act as WYSIWYG removed all carriage returns even though the preview had preserved them...

#

Re:The Reason why

Posted by: roblimo on July 11, 2004 06:57 AM
That's strange.

The default comment posting style for NewsForge is "Plain Old Text," which is supposed to preserve line breaks.

Did you perhaps paste in a document created in another word processor or text editor?

I ask so that we can help our programmers make this site easier to use. This interface was originally designed for Slashdot, which is not intended to be easy for people who are not technically adept.

Thanks,

- Robin

#

Re:The Reason why

Posted by: nikolic on July 11, 2004 08:32 AM
No, although there were some hiccups in the writing process.

I wish I could remember more fully to describe the situatin more closely, however I received an error

I believe that the error suggested a relationship to the use of the back button, newsforge embedded within a frame, or viewing through a proxy --- something that is unavoidable when pressing a link from hotmail unless a site forces otherwise. For instance, I am currently viewing the URL

http://64.4.43.250/cgi-bin/linkrd
?_lang=EN&lah=b60f571e41314d5a04132dbddbd2940<nobr>e<wbr></nobr>
&lat=1089503780&hm
___action=http%3a%2f%2fwww%2enewsforge%2ecom%2
fcomments%2epl%3fsid%3d39129%26cid%3d94772

rather then the page link in the email directly. This is actually troublesome as the proxy isnt using SSL, and so I have to remember to get out of hotmail before offering up any potentionally dangerous info to a site (credit cards, ss#, etc.)

Although the proxy is happy enough with connecting with a secure page, I am sent an insecure one.

Who knows, maybe there is something to do with having my session expire from hotmail's side.

I remember then --- rather than through hotmail --- navigating directly to the original discussion, logging in, and then pressing reply to the same post that I had intended. I can't remember if I rewrote the post or held it in notepad to paste it. I do this often enough for long posts; I may have in this case, honestly.

This second time there were no errors attempting to submit, but the carriage returns were removed after I had posted regardless of their presence in the preview.

I hope that it helps. I would more often submit a bug report, but I am never sure if it is actually me that is doing something wrong or if I have the correct vocabulary to describe what was happening.

If this works, I will resubmit the original post with carriage returns. Thanks for prompting me to send it in.

#

Re:The Reason why

Posted by: nikolic on July 11, 2004 11:44 AM
******************
First, a note:

Interestingly, I received the same error, but this time I can paste it for you:

Invalid form key: 1xQAR1PK6f

Chances are, you're behind a firewall or proxy, or clicked the Back button to accidentally reuse a form. Please try again. If the problem persists, and all other options have been tried, contact the site administrator.

I hope that it helps. By the way, I think grammar should be checked in the above error: Too many commas and conjunctions.

(i.e."firewall or proxy," should be changed to "firewall, proxy," and those are the only required commas in the statement.)

Please note that my last successful post was also submitted through a hotmail proxy, but this is the post that didn't work.
******************

Now for my actual post:

Usability is the difference between finishing a task in a few minutes with a low but reasonable learning curve, and spending hours without the ability to finish the same task.

Usability is the core reason that MS Windows pulled ahead of any other OS from 3.x onwards. Microsoft Press has published numerous titles regarding the "Windows User Experience," in order to educate developers and their managers on the importance of a user having consistent and adaptable experience throughout customers' use of their software.

Compared to contemporary competitors, Windows 3.x was crushing in all aspects of usability and Windows has since been top contender.

Regardless of MS business practices, usability was their real strength as it was always just a little more convenient to use their products.

Consider MS Office, while it is possible to have some considerable problems to have it do what you want, most users will never know of them. Their use of office is made similar to their expectations of something like a typewriter. This is supported with interface elements that meet this expectation rather than requiring more direct knowledge of the application.

A user needn't enter the em value of their word spacing or typeface width, or the point value of their tabs --- even though they are free to do so if needed. Instead, they can drag the options to their needed value and get back to the important things.

In fact, you can also set all of these things programmatically, but to expect a user to do this is unrealistic. Mention VBA to them and they will correct you saying the fish-fry is actually at the VFW.

This is part of the brilliance of Mozilla Firefox mimicking IE in the placement of Internet Options and KDE and GNOME provoding a Start Menu. Users are far closer to being "up to speed" immediately and are more likely to stay and contribute to whatever application they use.

While improving, Linux is most useful to developers and still abandons many users that would otherwise be interested.

When it comes down to it, people that need something to work --- particularly those that depend upon computing for business --- view initial and even yearly licensing of $500 or more as small change in the face of spending 2 weeks picking up a previously unneeded skill.

They may have lost many times that licensing cost by the time that they acheive their needed results.

The opposite should also be considered: while the usability of Windows is preferable, many simply can't afford to participate in a MS monopoly.

Populations that have the most to gain from Linux are left to learning an entirely new visual, verbal, and idea vocabulary on top of an unfamiliar method of production.

At this point usability takes on an economic tone as countries with deep poverty are exactly the places that open source software will prove needed as they develop as an economic fledgling.

I recently had a visit from a cousin of mine from the 'old country'. She had never seen a pool table except on the only TV in her villiage. This TV was in the richest household in her villiage of about 250 people.

I don't expect her to know anything of using any computer --- let alone figuring out how to configure her new Linux kernal or any other PC related skill. I have both Windows and Linux boxes at home. 3 of them. This was an impossible luxury to her and she was very curious to use them.

She was familiar enough with the idea of computing and the internet to get around and start digging.

She has seen PCs on the same TV and there is one --- ONE --- in the medical school that she attends to become a surgeon. She had never used it.

She tried both during the week that she was with me and was simply more comfortable using Windows regardless of the fact that she would never be able to afford it without a fight.

To give you an idea of the general impoverishment from where she comes: her yearly salary as a nurse is just over $230. You heard correctly: per year. That will nearly quintuple when she becomes a surgeon --- imagine that: a surgeon making less than $1500 per year. Even at a surgeon's income it would take her 3 or 4 years to be able to afford Windows licensing --- currently it would take her up to 10 years of heartbreaking saving.

That is not an option for her and will never be unless she is able to come to the US or some other affluent country. This kind of situation is just what makes Linux and open source so important to the world at large.

Regardless, she was more comfortable with Windows: worked more often even with no experience.

For instance, she discovered xKill (briefly: an app that kills a process under Linux --- it's actually really useful tool because you don't have to find the ID of the process to kill). While I was at work, she had killed the KDE taskbar accidentally simply by clicking on it with xKill.

She thought she had irrivocably ruined something and was horrified at what I would think or do. You must understand that this was A_COMPUTER, the same of which her medical school could only afford one. It had the same value as a house from where she comes. She seemed to react with the same gravity as if she had killed A_PERSON. When I got home, she looked as if she had been crying for hours:

1. Restart: problem solved.
2. Comfort the crying.
3. Get dinner somewhere other than at my home office.

Anyway...

(By the way, xKill immediately killing the taskbar? Seriously... have mercy.)

(PS: It's true that you can acheive the same effect in Windows by killing explorer.exe from the task manager, but it does warn you first.)

Back to my original discussioin: man and help pages are wonderful for seasoned *nix users. This is because grep is their friend. (Grep is NOT my friend.) This kind of self-help is virtually useless for those that are new to the platform.

While many *nix users will boast (sometimes surprisingly angrily) that they can enter a command in seconds that commits system-wide changes and that is why this method is the best and most usable.

The problem here is that a user must know explicitly what to do - that point is at the top of the learning curve.

Seasoned users forget while saying this that this is exactly what horrifies most mainstream users. Entering a command that commits system-wide changes is, well, a commitment. The command that is misspelled, forgotten how to use, or forgotten completely can cost them weeks.

Mainstream users expect a computer to act as any other appliance would in their home. They don't need to know how fast their spin cycle runs in their washing machine.

Frankly, they shouldn't mess with it. This is because it could put them in very serious danger. A washing machine needs two dials to do the job and users benefit from it. If a user wished to overclock the spin cycle, they can learn more when they feel the inescapable need to do so.

Regardless, this is an idea for someone other then a househusband/wife to discover.<nobr> <wbr></nobr>...

rpm -ivxzf forTheLoveOfGod

This isn't good enough for ME --- let alone anyone without any computing experience. I would change this torture in my spare time if I knew how, but I unfortunately don't.

Frankly, I personally hate installing anything in Linux. This is mostly because I have to find the thing once it is installed. Time. Searching for it is difficult because there is no common extension for an executable under Linux. If I search for it I receive hundreds of files with the same filename and I have to find the one missing an extension.

This assumes that it is actually missing an extension which isn't always the case.

A user should never be forced to edit a text file in order to get anything to work; requiring such an edit such as customizing a script or config file should be considered an embarrassment to the idea of usability.

If Linux and other OSS is to encourage migration, then the use of that software should be at least encouraging rather than discouraging.

#

Re:The Reason why

Posted by: Anonymous Coward on July 19, 2004 05:03 AM
"Useablilty is simply a matter of taste and choice, that's why when you ask a quesiton about useablilty all you get is opinions."

No, this is simply untrue. Usability is not only a matter of choice. There are rules how a human mind works, how it reflects what one sees and how everyone out there handles his workflow.

I know some interface designers at SAP/Germany, and they are all graduated psychologist - mostly with programming skills, but they know much more about human standards then technical ones.

I think one reason why we fail so badly is because there are no people "checking" the interface design. All big companies that are successfull in interface design *seperate* program logic and interface design, because it's much easier to do one thing at a time.

I always ask out-of-project people when I'm unsure if they can navigate through "my" interface without having to think "where to click next". If this can't be done, there's a flaw in the interface.

One of my worst experiences was xcdroast; this program is simply horrible. It works great, it has all features, but the interface logic sucks as much as possible. There are important buttons on the lower left of a page, you have to SWITCH from the first to the second tab to add things first before you can continue on the first tab. All this is simply *stupid*.

I know very well how this happens, I make the same errors almost every time. But I try to improve steadily.

The first step for all of us should be learning to step back and think "could I use this without knowing anything but the goal?"

Big companies simply throw money at the problem, we have to use our own brains. I think it can be easily done, because we have superior ones. Don't we? Let's 0wn them<nobr> <wbr></nobr>;)

#

Re:The Reason why

Posted by: Anonymous Coward on July 19, 2004 07:45 AM
And you are an Engineer! You won't listen to anyone else. Whenever anybody tries to provide you with advice you will put forth your own arguments why your design is better. You lack the ability to distinguish 'which idea is better' so you consider all ideas junk.

The sorry thing is that if you just took a few minutes to do some research and opened your mind to suggestion you would learn a lot and your software would improve.

As the CEO of a software company we often have this problem with companies we have acquired. We have even fired lead engineer for some projects and taken his 'baby' away from him because he refused to make the product user friendly. When the engineers in the world realize that without the marketing guys their salaries won't get paid software industry will be a better place.

By the way. I am an engineer and a marketing guy.

#

Re:The Reason why

Posted by: Anonymous Coward on July 20, 2004 08:18 PM
Useability is for wimps and morons?
To all those who replied, "Useability comments come from whiners, morons etc [insert your favourite disparaging term of contempt here]" I suggest you check
a) Hall of Fame/Shame already mentioned
and
b) Jakob Nielsen's <A HREF="http://www.useit.com/" title="useit.com">http://www.useit.com/</a useit.com>
for hard-nosed 'software acceptance' research plus commentary on the benefits of improved useability.

If you want to win the war against the dark side, you have to make your software more attractive to use.

Technical competence (or their umpteen engineering patents which demonstrate same) doesn't sell Mercedes Benz cars - the desire to own one *more than to own a humble GM/Ford/etc* is what does it. Their technical competence is necessary but *not sufficient* to generate the requisite consumer demand.

Technical competence beyond the ken of MS has never been an issue: even Bill is not interested in technical excellence, he just wants world domination. Mediocrity will get him there as long as the latest version is [or appears to be] "easier" than whatever people were doing before.

So my recommendation: do not attempt an open frontal attack on the MS hegemony; rather do an outflanking job - make it so that using OSS becomes a no-brainer because it is just so nice to use.

That's what has kept Steve Jobs and the Apple-wagon rolling...

It also happens to account for a large slab of the equally faithful OS/2-eCS user community staying with a platform which has been declared dead (by MS) annually since 1989<nobr> <wbr></nobr>:-).

John Angelico
OS/2 SIG
Melbourne PC User Group Inc
talldad@kepl.com.au

#

Usability

Posted by: Anonymous Coward on July 10, 2004 02:13 AM
Some of the simplest rules for usability are:

1. consistency (learn once, use many times)
2. focus on most typical use (advanced needs are for advanced users)
3. elegance (minimal steps to do typical tasks)

It doesn't take a lab to design according to these rules. It takes a whip to ensure the rules are obeyed and it takes restraint to avoid turning something elegant and powerful into a monolithic Frankenstein's monster.

Labs make sense for polish but a lab is as simple as loading the software for someone to play with, someone who hasn't got used to quirks, and watching them try to do everyday tasks.

#

Re:Usability

Posted by: Anonymous Coward on July 10, 2004 02:56 AM
Interesting thought here - has anyone considered a tool which could (VOLUNTARILY) be installed on someone's computer when they use a new program for the first time, that would watch what they do and send "video" (think VNC rather than xine) of their mouse movements and screen contents back to the developers?

It would need to be VERY prominently displayed onscreen that everything you do is being recorded, but it could be useful (especially in conjunction with a list of tasks to try to accomplish, which is apparently what professional usability studies do) to let the developers see what the user is having difficulty with.

The same software could perhaps be used for parental computer-use monitoring for small children (not-so-small children would figure out how to get around the monitoring just like they figure out how to get around other parental restrictions...)

#

Re:Usability

Posted by: Anonymous Coward on July 14, 2004 03:56 AM
This could be similar to the Debian popcon package. It checks the popularity of packages installed, used and upgraded. It is then used to decided what applications need to be on the first install disk.

applied to UI it could check to see what people do the most and would allow developers to concentrate on those things for useability and optimizations. This would be almost like a user profiling thing.

#

Re:Usability

Posted by: Anonymous Coward on July 19, 2004 04:35 PM
Interesting though but not practicable. It is nearly impossible to come from the mouseclicks and the screen to the intentions a user had while performing these interaction steps.
The most powerful instruments in usability labs is the "speaking aloud thought protocol". The "test user" is speaking aloud his thoughts while performing different tasks. So you have to provide an audio track.
But the biggest problem is the amount of data. You need hours to analyze a task solved in minutes.

#

Look harder

Posted by: Anonymous Coward on July 10, 2004 02:38 AM
ever heard of the Gnome HIG? The same principles could easily be used in KDE apps, and produce very usable apps!

#

Re:Look harder

Posted by: Anonymous Coward on July 10, 2004 06:18 PM
There is some stuff in the Gnome HIG that really shouldn't be there. The Gnome HIG is also a lesson in how people create some good guidelines, but then try to follow them in a static and formulaic manner. It's as if people think that by preaching HIG compliance it will somehow make everything usable.

Read any good book, such as the excellent Elements of User Interface Design, and you'll find that this is bad usability. Look harder.

#

Bull.

Posted by: Anonymous Coward on July 19, 2004 11:12 AM
I used to be a Gnome fan. Its so-called "user-friendliness" push has made it unusable and a pain to deal with. Havoc Pennington needs to take a 4-year break in a room with nice padded walls until he gets his mind back and goes back to real coding rather than pushing bad UI rules and convoluted frameworks all of which he must have made up while on crack.

#

useability is relative

Posted by: Hillbilly on July 10, 2004 02:43 AM
my first computer had win98 it took a little time to learn to be proficient using it, after trying Linux it took a little time to be proficient on Linux, this is only natural with learning anything new, whether learning to speak a foreign language or learning a new trade, etc...etc...

i been using Linux for 100% of my computing needs that if i get on a Windows computer i feel awkward on Windows now...

some people just have a hard time using a computer no matter what OS is installed, i have a few family members that own a computer and they are just as clueless on Windows as they would be on Linux or any other OS...

the ability to adapt is crucial to learning anything...

#

Re:useability is relative

Posted by: Anonymous Coward on July 10, 2004 05:13 AM
the ability to adapt is crucial to learning anything...


But we should do what we can to avoid putting obstacles in the way of learning.

#

Re:useability is relative

Posted by: Anonymous Coward on July 20, 2004 02:36 AM
Part of the problem is the idea that 'they just aren't used to the system'. Too often it is used an excuse to not change something. "What? You don't know all the keystrokes for vi? Well you just need to learn the difference between vi and Word."

Ever run across a 'feature' that just really annoyed you? How about "are you sure you want to exit" when you have done nothing to create or change data. For a casual user, it's not a big deal. But for a daily user it can be real annoying.

Ever worked with an application that kept forcing you to move from keyboard to mouse and back to keyboard? Data entry, navigation... Web pages are notoriously bad at this.

And from the article, there should NOT be a menu item called 'smbUmount'. As a user, I don't care what smb is Umounting. Why not 'Disconnect this network location'?

The article sends us on some random detours. Usability has nothing to do with code check-ins, privacy, or OS independence.

It's pretty simple really; take a group of target users, have them point out things that are unclear or difficult, follow through with making the changes. This is the process we need instead of: 'well, there's a learning curve, you'll get used to it'.

#

Re:useability is relative

Posted by: Dmitri Zdorov on July 20, 2004 06:51 AM
Unfortunately it is more complicated than that.



OSS suppose to be a superior technology, but it does not work out of the box and in many cases it takes a lot more time and effort to make it work. This is true for commercial software too, but often in a different way. In commercial software it is more often a technical problem. Software is written using non-standard proprietary methods with the $$$ in mind instead of "Better for users".



OSS usually does not have this problem, but it is still difficult to use. And the reason is USABILITY (or lack of it)


OSS usually does not have this problem, but it is still difficult to use. And the reason is USABILITY (or lack of it)

#

Usability for the lowest I.Q.?

Posted by: smurfnsanta on July 10, 2004 02:51 AM
The problem here is the same problem we experience with marketing: Write a commercial, news article, or movie with words and concepts that an average citizen can understand, and you rapidly loose everyone under that IQ level. So most media is written at about an 80 IQ, with an occasional sophisticated or enlightened quip to keep the intelligent from shutting you off.

At what range would you place the majority of developers? Certainly above average. If they are just average, their IQ and familiarity spikes in logic and programming.

Asking developers to increase usability is roughly equivalent to asking them to dumb the interface down. That's a big hurdle, but one that's important to broach.

Certainly decent How-To-Make-My-App-User-Friendly tldp's would be useful. But a simplified method for newbies to submit 'I don't understand feature X on version Y' messages to developers or perhaps an intermediary clearing house like bugzilla might yield promising, realtime results.

#

Re:Usability for the lowest I.Q.?

Posted by: Anonymous Coward on July 10, 2004 03:19 AM
Usability is not about dumbing down. As with all design it is really difficult. It is about understanding what the user wants and making real choices, based on understanding of the user. This often results in simplicity and really has nothing to do with stupidity.

This dumbing down mentality is really keeping KDE back IMHO. It takes courage to make real UI choices.

#

Re:Usability for the lowest I.Q.?

Posted by: Anonymous Coward on July 10, 2004 03:51 AM
Think of usability in terms of an iceberg: for the computer novice/newbie, there is a simple, stream-lined interface with relatively few options on the surface. These users are less likely to get confused and (hopefully) will be happy.

However, for the power users, there is another, greater, deeper layer of options lying underneath the surface of the GUI. Here is the expert's playground, the place to tweak to one's heart's content. Even better, the expert can find and understand this layer because he or she is, after all, an expert/power user.

This is the way GNOME is trying to approach this debate. GNOME certainly does not have everything right in this regard, but I think it's a promising direction. (Hopefully, with searching enabled for gconf-editor in GNOME 2.8, things will be easier to find under the surface.)

KDE has the flexibility and power (in spades). It just needs to provide the simplied interface on top, and let the power users play beneath the surface.

#

Re:Usability for the lowest I.Q.?

Posted by: Anonymous Coward on July 10, 2004 05:20 AM


If my users look dumb to me, doesn't that mean that I am simply not trying to look at things from their point of view? Remember, Eric Raymond has trouble configuring his printer sometimes.


Usability makes things easier regardless of your "IQ", because the less mental power you have to devote to dealing with the mistakes the author of your program made, the more time you have to do what *you* want to do.

#

Re:Usability for the lowest I.Q.?

Posted by: Anonymous Coward on July 19, 2004 06:53 PM
Yes,

http://usability.kde.org/activity/usabilityreport<nobr>s<wbr></nobr> /usability-report-guide.php

btw bugzilla is a good usability candidate.

It's not only usability, it is also: how to make the workflow faster. I.e. why do I need x pages in order to submit a bugreport and so on.

Mailman is a good candidate for usability improvement.

Gcc is a good candidate for usability improvement for knownots.

#

Re:Usability for the lowest I.Q.?

Posted by: Dmitri Zdorov on July 20, 2004 07:00 AM
This is a very big misconception, and it comes from not understanding of what usability is.

The same way:

People want to have safe streets and parks not to promote lower physical fitness. Even best karate masters armed with machine guns prefer safer streets. And they have mothers and daughters that prefer safer streets.

#

Re:Usability for the lowest I.Q.?

Posted by: smurfnsanta on July 20, 2004 06:36 PM
This is a very big misconception, and it comes from not understanding of what usability is.

Looking back on it, I really just wanted to incorporate the idea of an easy to use 'I don't get this' reply method/app for the users who find certain 'features' incomprehensible.

Unfortunately, there isn't a usability project interface for what I had in mind, and you don't want thousands of emails from newusers spamming dev lists with these types of problems. I think categorizing user problems in real time would have a tremendous value for dev teams however, and inherently address usablility issues.

#

Coolest comment of the year!

Posted by: Anonymous Coward on August 01, 2004 05:50 PM
That's the coolest comment about usabilty i have read in a long time!

Can I quote you?

marius.

PS: coudn't find the find the submit button for a sec...
What does it say about my IQ?
Or... what does it say about its absense... i'll go and click preview i guess. (like it says underneath)

PS2: You could say that i'm obviously stupid cause I didn't instantly know what to do even though it's written.
I say there's very obviously a problem when you have to explain.

#

howto

Posted by: Anonymous Coward on July 10, 2004 03:36 AM
find someone who uses the microsoft version of whatever program you're copying and then sit them down in front of your project. watch them and see what they can't figure out. remember that they have no fucking clue that their mouse has two buttons much less how to use the right click.

also remember that they have no concept of drag and drop, and only a vague understanding of saving files. actually, you might as well say that file and folder locations are just as lost on them as right clicking.

keep in mind that they will only use about 3% of the features of your particular program, and all those extra buttons you had to have little tiny icons made for will go mostly unused.

don't forget this either... even though your job as an open source programmer is to make your application look EXACTLY like the microsoft version, even microsoft doesn't always do it the best way. you could actually improve your program's usability by NOT making it look exactly like a microsoft application

#

Re:howto

Posted by: Anonymous Coward on July 10, 2004 08:12 AM
> don't forget this either... even though your job as an open source programmer is to make your application look EXACTLY like the microsoft version...

Looks like Open Source is doomed if Microsoft goes under!. Help keep microsoft alive - buy Windows XP or Longhorn NOW!

#

Re:howto

Posted by: Anonymous Coward on July 10, 2004 10:22 AM
don't forget this either... even though your job as an open source programmer is to make your application look EXACTLY like the microsoft version, even microsoft doesn't always do it the best way.

What the hell are you talking about? Since when are we, open source developers, trying to copy Microsoft. I have no intention of doing that. It's is an idiotic and pointless activity. If we are just cloners, then we might as well just just MS products. I work with open source because I feel I can do something better.

Good lord!

Daniel.

#

Re:howto

Posted by: Anonymous Coward on July 10, 2004 11:54 AM
I was totally being sarcastic with the last paragraph, although I was thinking about Ximian Evolution when I said it. But sarcastic or not, there are too many similarities between so many windows apps/interfaces and so many open source projects. You can tell that in many instances the quick/easy was out was taken, and the interface is more or less a ripoff of an existing app.

Actually, I remember the first time I saw someone demoing Evolution. I remember him saying "see, it looks almost JUST like outlook!" And they said it like it was such a great thing.

#

Well, some of them know about two mouse buttons

Posted by: secrity on July 10, 2004 11:17 PM
I fully agree with all of your points, except for one: I have met users who do know that there are two mouse buttons and could maybe use it. They may use it by accident, but they could use it.

#

Re:Well, some of them know about two mouse buttons

Posted by: Anonymous Coward on July 10, 2004 11:47 PM
Yes, but what about three buttons?<nobr> <wbr></nobr>:-)

#

Re:Well, some of them know about two mouse buttons

Posted by: nikolic on July 11, 2004 11:59 AM
Hey, I have 5 buttons on my mouse and I know how to use them, but only in Windows --- they are even configured to act differently within different applications.

Unfortunately, I still can't get Linux to use them. Any help?

#

Re:Well, some of them know about two mouse buttons

Posted by: Anonymous Coward on July 12, 2004 01:53 AM
http://koala.ilog.fr/anyboard/MouseWheel/posts/21<nobr>3<wbr></nobr> 0.html

The end of the url should read "2130". Newsforge mangles it.

#

Re:howto

Posted by: Anonymous Coward on July 20, 2004 03:52 AM
Of course, maybe if you spent a little time understanding the expectations of the users, you wouldn't have such a low opinion of them.

Why would you write OS software for people you apparently despise? I could understand if it was your job, maybe, but only for a short time.

You need to realize that users are only interested in performing tasks. They don't care if some uber coder built an application with so many features that 97% of them will never get used.

You need to figure out who your target audience is and what they want, and then design with that goal in mind.

#

No, no, no!

Posted by: Anonymous Coward on July 10, 2004 03:46 AM
Perhaps you remember Mr. Cronenberg's film "The Fly" and how a certain dialogue went (don't know the exact words):

-- "It's missing something. I'm almost there, but I don't know what goes wrong..."
-- "Maybe some final detail... I know you'll find it."
-- "I don't know... I walked through everything, checking all possible errors... still..."<nobr> <wbr></nobr>:-/
-- "Hey, maybe what's missing is that granma touch, you know, they kind of pick up a baby, look all over him, hug them and pronounce them ok."
-- "Well, I'dont... yeah! You maybe into something there..."

That's what I think. Developers understand a lot about computers -- but not about people! If it feels like noise, it's not because it's just noise. It's difficult to understand what users say, just like whales. One has to master other areas, not pure IT, to make sense of them. Psychology, human factors, aesthetics, art, common sense etc. all play a role.

For this to work a sculptor is needed, not a project manager or an engineer. We need guys that do beautiful things... guys like Everaldo, though he's a designer, not a GUI expert... but who knows about the future.<nobr> <wbr></nobr>;-)

Or guys like this one: http://www.ferryhalim.com/orisinal/ . We need people who can cook delicious plates, beautiful and tasty. We need someone who wants to please. Just like KDE guys know about software excellence, we need someone concerned about visual/audio/immersion excellence.

I guess we need a grandmother...<nobr> <wbr></nobr>:-)

Take care.<nobr> <wbr></nobr>;-)

#

Re:No, no, no!

Posted by: Anonymous Coward on July 19, 2004 06:59 PM
What's needed is a formal usability review.

And substancial reports like this one
http://usability.kde.org/activity/usabilityreport<nobr>s<wbr></nobr> /usability-report-guide.php

What's not needed are insubstancial idea flames by newbies on the mailing list that are excited with each others ideas. Nobody has time for this. At least a usability maintainer is needed who writes studies or a usability list weekly report.

German Linux magazine started a Usability column, no analyst trash talks/opinion paper but a substancial article about an application.

#

Re:No, no, no!

Posted by: Anonymous Coward on July 21, 2004 07:35 PM
One has to master other areas, not pure IT, to make sense of them. Psychology, human factors, aesthetics, art, common sense etc. all play a role.

Quite, the arrogance of the original author is shocking, as if programmers are omniscient. And the mini rant about how usability experts are external specialists, and how that compromises openness is beyond me. The guy obviously has a chip on his shoulder. He should shut up, read some of Nielsen's stuff, and then come back.

#

Why re-invent the wheel?

Posted by: Anonymous Coward on July 10, 2004 03:50 AM
Apple has always been praised as a leader if not THE leader in software usability. They have devoted countless hours and dollars to making their systems some of the most user-friendly ever.

No, they aren't perfect and no, you can't please all of the people all of the time. But...they are by far and away the furthest along.

So, why not just "borrow" tricks from their HIG? I *DON'T* mean clone Aqua or OS/X. I mean review and consider the sections on such things as how menus and windows operate; where to place buttons; color scheme recommendataions; when to use transparancy; audible and visual feedback; etc.

A perusal of the Gnome HIG would also be a good idea, but I wouldn't get too hung up on compatibility with either guide.


  -Charles

#

Re:Why re-invent the wheel?

Posted by: Anonymous Coward on July 10, 2004 05:29 AM
This may be a good idea if we would not run into claims of "IP" theft but please don't give me a one button mouse any time soon.

A Nony Mouse

#

Re:Why re-invent the wheel?

Posted by: Anonymous Coward on July 19, 2004 05:05 PM
Because this way is very expensive and that is because there are no "tricks".

Apple uses a lot of time and money to test their software in usability labs. This is very expensive. I do not think this is a practicable way for open source software, expecially for smaller projects.

#

Apple don't get it Re:Why re-invent the wheel?

Posted by: Anonymous Coward on July 20, 2004 06:34 AM
I bought my parents (in their 70's) iMac (top of the line at the time of purchase) and they still don't 'get it' on its usage. I even bought an iMac intro book with lots pictures but it still doesn't help.

(I've purchased a total of 4 Macs over the past 20 years, so it's not like I'm not familiar with it... I currently own and use mainly wintel boxes, but that's a different story<nobr> <wbr></nobr>:-)

#

GNOME is an example of why not to do this!

Posted by: Anonymous Coward on July 10, 2004 07:52 AM
GNOME did some usability studies, funded by Sun, and found that they need to remove flexibility from GNOME. This is all well and good, if the result is a good UI design for both novices and power users. However, as their move to "spatial" Nautilus, and their villifying anyone who dares challenge the most obvious wisdom of spatial Nautilus (despite it being tried and discarded by Mac *and* Windows), shows, usability studies aren't always accurate but open source developers aren't always pragmatic enough to look at the usability studies with a critical eye. Some would rather bask in their self-labelled brilliance than deal with the fact that they might be WRONG.

#

Re:GNOME is an example of why not to do this!

Posted by: Anonymous Coward on July 10, 2004 10:26 AM
If you don't like spatial Nautilus, change it.

Spatial browsing is supposed to help the low-level users (e.g. my grandma). It makes sense to make the default settings match what is best for the least techie users, because techies are precisely the ones whom you can expect to change a setting they don't like.

Cheers,
Daniel.

#

Re:GNOME is an example of why not to do this!

Posted by: Anonymous Coward on July 10, 2004 02:07 PM
What it's supposed to do and what it does are two different things. I introduced the user from hell to gnome, he knows absolutely nothing about computers. He is in his late 70's. He did all his writing on a typewriter because he didn't know how to use a computer. He screwed up his windows machine he got for a present, he "accidently" deleted the windows system directory! He had so many pop-up ads previously when he tried to go to a web page, he didn't even want his computer fixed. I put him on the linux machine so I could log in remotely, he has cable broadband and so he couldn't delete anything serious. I tried him for about a month first on GNOME. Then switched him to KDE because he was complaining alot. He hated gnome's filemanger! He hated having windows all over his screen. He is now using evolution,Firefox and MS Word under codeweavers and KDE. I no longer have him rushing to my house every day with a computer problem.

#

Re:GNOME is an example of why not to do this!

Posted by: Anonymous Coward on July 12, 2004 07:05 PM
Well, I have no problem with the default settings being set to what is deemed best for clueless newbie users.

What I have a serious problem with is the removal of the possibility to change the settings, something which seems to have become popular at Gnome recently. It seems that some people think that we should dumb down preferences dialogs to the barest minimum of options, as including too many options makes the application too confusing.

This defeats the object of *having preferences* in the first place. The truly clueless don't change preference settings anyway, since they are too scared to touch anything.

For anyone above that user level, even a relative newbie, there should be no difference (difficulty-wise) between two options and twenty options, as long as they are all clearly and sensibly labeled and categorised. A rich preference set helps people adjust the application's behaviour to the way that they like to do things, thus making it easier for them to use the program.

If you don't know what something means, or you don't care which way it behaves, you can just leave it alone.

I am extremely frustrated by the dumbing-down of Galeon's interface. An option to put tabs on the left or right is hardly an arcane power-user requirement, yet gconf is still the only place where you can set this option, as apparently the developers think that normal users can't handle the confusion.

If they take out the gconf option, or mouse gestures, I'm bailing for Firefox. I don't know whether the millions-of-extensions model is wise in the long run, but in practice it's doing a better job of giving me what I want.

#

Exactly.

Posted by: Anonymous Coward on July 13, 2004 08:33 PM
Yes, yes, yes! Gnome's arrogant "WE know what YOU want" attitude just sucks.

It is wonderful to see that I am not the only person on this planet feeling that way.

#

Re:GNOME is an example of why not to do this!

Posted by: Anonymous Coward on July 19, 2004 08:21 AM
You are so wrong its not funny. The spatial metaphor is based on theory that has no relevance in the real world. People do not want many windows all over the place. They want to rearrange each window for that particular purpose at that moment in time, not have it appear in the same place. They want on screen tools like folder views and navigation buttons to help them move around.

This is based on my experience as an advanced user who mostly uses the shell, and a sympathetic viewing of every novice I have ever seen use a computer, and everyone in between, including, apparently, the entire Window and Mac user bases, who abandoned this model.

Tbe fact there are a bunch of hot keys in the Gnome spatial implementation to fix its many issues shows how flawed it really is. Sure its for the novice user. What a joke.

This all shows one of the big flaws of open source, especially when there needs to be an effort like this to make a consistent desktop. I imagine working on Nautilus pays. One hot shot after another steps up with their theories, and fascistically implements them. Then they realize it's not actually going to work, and more importantly is boring, so they walk away. So we end up with a bunch of "solutions" that aren't well rounded and never really finished. We end up with a bunch of "solutions" that are designed for theoretical users that don't really exist. The techies won't use it, and the novices are lost in the woods, especially since they can't ask the techies, who have all turned the spatial model off, or use the shell.

The part that does make it work is designing component oriented software. Then at least the next person to come along can build on other people's work, and ultimately Nautilus is fairly useable because of this. Once you turn the spatial model, which is designed for nobody, off.

#

Spatial Nautilus hurts novice users the most

Posted by: Anonymous Coward on July 19, 2004 11:20 AM
You must be nuts if you think novice users are benefitted by being pushed into such a terrible unusable interface. If you'd ever tried asking a statistically significant group of *REAL HUMAN USERS* rather than just deciding that "this must be right for them" you'd know spatial nautilus is TERRIBLE for new users, more so than for anybody else.

#

NO DON'T imitate gnome PLEASE

Posted by: Anonymous Coward on July 10, 2004 05:08 PM
I like KDE and IceWM. I'm productive in them. The interfaces are consistent from app to app. The interfaces are consistent across upgrades. Gnome is a nightmare of radical changes and brokeness in each release.

One example: Galeon used to be the Perfect Web Browser. I adore the Smart Bookmarks Toolbar, it's a fast and convenient way to do searches. In early editions of Galeon, you could easily edit the Smart Bookmarks Toolbar by running a wizard. Fast and easy. Now, at version 1.3.15, I can't even begin to figure out how to change the damned thing.

And the freakin toolbars are HUGE, with no way to reduce their size that I can find. This is NUTS. And Galeon is no longer themeable- you're stuck with the Gnome theme.

And when will Gnome discover right-click? right-click menus are a huge timesaver.

I don't know about these studies that Gnome is supposedly following. But they are wrong and bad. Removing functionality and hiding configuration options is really foolish. That's not simplifying, that's vexing.

#

Re:NO DON'T imitate gnome PLEASE

Posted by: Anonymous Coward on July 12, 2004 01:48 PM
And when will Gnome discover right-click? right-click menus are a huge timesaver.

Not sure what you mean - Gnome does have right-click menus, and always has. The only restriction the HIG places on them is that they should never be the *only* way of doing things. What's your problem?

#

DBus, HAL

Posted by: Anonymous Coward on July 12, 2004 02:06 AM
Speaking of GNOME the DBus / HAL development is going pretty well. Yet another improvement for the usability, user friendliness. Think of it as "Configuration -> Hardware". See http://www.freedesktop.org for source, screenshots, fearure list.

#

usability

Posted by: Anonymous Coward on July 10, 2004 08:31 AM
Ergonomics my friend. It can be quantified. It can be done. Measure how long, how many mouse clicks, keystrokes, and mouse movements it takes to perform a task. And name software, so that people know what it does. Gimp? Konqueror(SP)? etc...

#

What a silly editorial

Posted by: theantix on July 10, 2004 09:55 AM
There is an immediate tradeoff between useability and features. The KDE and Gnome project have different stances on that tradeoff -- Gnome preferring simplicity and useability before gee-whiz features, and KDE preferring the opposite. Complaining about the usability of KDE is as useless as complaining that Gnome doesn't have the latest gee-whiz feature. I say to both of them: examine where you sit on the useability/features balance and choose the desktop environment that best suits your personal style. And yes, having either as an option is a good thing, because there are legitimate reasons why different people have different sets of preferences.

But please don't complain about the makeup of the projects that help define them -- that's just plain silly.

#

"Usability vs Features" theory is false

Posted by: Anonymous Coward on July 11, 2004 12:49 AM
The "tradeoff" between usability and features is for the most part a myth, or should I say more of a "self fulfilling prophecy". When a piece of software is designed well, simple basic functionality remains easy to use even when more advanced functionality is also available. There is no reason for a basic task to become harder just because extra functionality is also available. It seems to happen like this in the real world because designing usable software is difficult, and designing a bigger piece of software (read: more features) is even more difficult. So what people do is dumb things down by removing functionality. So that when they screw up the design, the software is less screwed up compared the even greater mess they would have made if they had to put the extra functionality in there as well.

Simply put, it is harder to screw up the design of a program with limited functionality compared to a program with lots of functionality. But this doesn't mean that the way to make usable software is to dumb things down (better design is), after all a dumb and usable program isn't much help if the feature you need just isn't there.

--
Simon

#

Re:What a silly editorial

Posted by: Anonymous Coward on July 19, 2004 07:01 PM
Oh, I think KDe is more usable than Gnome. And Gnome lacks the options at the right place although they are there. They chose the wrong guidelines. Sad but true. KDE is a usability dream compared to Gnome.

#

Whine,Whine,Whine.........

Posted by: Anonymous Coward on July 10, 2004 12:41 PM
Is that all the "Useability" crowd can do these days?

Just how seriously should people who's basic goal in life seems to sit around whining about something rather than actually getting off their fat lazy asses and actually doing something to improve the world.

I hate to break this to you so-called "Usability Experts", but Open Source wasn't created for you losers to build your so-called careers on.

Most people in the Free Software/Open Source community are *NOT* interested in being suckered
into working on you half-baked "suggestions" only for you to come along later and claim all the credit the while the Free Software/Open Source community did the actual work.

And let's face it, this is what 99% of the whining
concering Free Software/Open Souce is really about.

The "Usability" community is upset that they can't get FS/OSS developers to do the dirty work of actual development of their "suggestions"

#

Re:Whine,Whine,Whine.........

Posted by: Anonymous Coward on July 10, 2004 01:20 PM
Look at http://developer.gnome.org/projects/gup/ , GNOME Usability is there to look at

#

Building Usability

Posted by: Anonymous Coward on July 10, 2004 01:04 PM
Articles of this nature keep popping up over and over again. This one has some merit, though.

If you look at Apple's old HIG - the one that corresponds to the likes of System 6 and System 7 - you'll see all their specifications about the placements of elements on the display, and about the way in which the computer should respond to user input. But what you'll also see at the beginning is a series of statements, a "HIG Prime Directive" if you will, being the principles behind the specifications laid down in the rest of the document.

Usability isn't something that a group of creative people suddenly will into existence; it's something that's built, from the ground up. You start with the foundations, of course, which is why the comparison between solving usability issues and solving technical issues is apt; obviously not the same skills are used, but the procedure is not dissimilar.

The foundations - the principles I refer to above - have already been well-laid elsewhere, so re-inventing them is not necessary. However, OSS should feel free to set new directions; no-one is beholden to make everything look and act like it would if it were running under the Microsoft Windows platform.

One principle that would make a big difference from the get-go: in Apple's old HIG, it states that "The computer should not ask the user for anything it doesn't already know.", meaning of course that you tell the computer something once, and it shouldn't ask you again. This in turn means a deal of information sharing between the OS and Applications and between different Applications, more than occurs now with popular applications such as those that run within the LSB/XWin/KDE or LSB/XWin/GNOME environments.

#

a big problem...

Posted by: ammoQ on July 10, 2004 02:39 PM
is whether to make UIs consistent with already exisisting UIs, which can be a trade-off in terms of being intuitive, or make them great on their own, but different from other UIs.

#

a common example..

Posted by: ammoQ on July 10, 2004 02:46 PM
... is the floppy disk icon which means "save" in many applications. The problem is that most files are not saved on a floppy, many PCs don't even have a floppy disk drive, some new users do not even know what a floppy is.

#

No smbUmount menu selection

Posted by: Timo Pirinen on July 10, 2004 02:54 PM
Here's an example: Konqueror, KDE's file and web browser, has a menu entry called "smbUmount."


No, Konqueror on my SUSE 9.1 does not have an entry smbUmount. In the Tools menu there is an entry for 'Smbumount current share'. Note the additional two words?


One could ask why is the entry placed in the Tools menu with all those other entries pointing to external (for Konq) tools? Now could it be that it also uses a tool external to Konq, could it? It could, wow!



I don't need a laboratory with video gear to figure out that this is nearly impossible for non-hacker users to understand.


The entry is only activated when it is applicable. It is greyed out and can not be chosen if it is not. Like when there is no 'current share' on which to apply it. Easy to understand, no?

#

Re:No smbUmount menu selection

Posted by: Anonymous Coward on July 10, 2004 05:31 PM
No its not easy to understand. I'd like to see you find any average user who knows what smbumount means or what curent share means.

#

Re:No smbUmount menu selection

Posted by: Anonymous Coward on July 10, 2004 06:32 PM
Exactly. It's easy for us to know because we know what 'smb' means, what the history of 'umount' means and how the 'n' got dropped. But for a person that doesn't understand 'mount', let along 'smb', it could be truly confusing.

#

Re:No smbUmount menu selection

Posted by: Timo Pirinen on July 11, 2004 04:16 AM
No its not easy to understand. I'd like to see you find any average user who knows what smbumount means or what curent share means.


Yes there should be an easy way to share files in Konq. There is! Just right-click a folder and choose 'Share'. Smbmount and - umount are something completely different, you just mistakenly think they are Linux filesharing.



You (and the writer of the original article) fail to see that SMB is not native for Linux. Native Linux filesharing is extremely easy via a context menu selection called, you guessed it, 'Share'. Why oh why should the non-Linux SMB be a preferred way of sharing files in Linux?

#

Re:No smbUmount menu selection

Posted by: Anonymous Coward on July 11, 2004 02:48 AM
I think "Disconnect current share" would be better.
Or better still, if ABC is the current share, "Disconnect share ABC"

#

Re:No smbUmount menu selection

Posted by: Timo Pirinen on July 11, 2004 04:01 AM
I think "Disconnect current share" would be better.
Or better still, if ABC is the current share, "Disconnect share ABC"

No. The verbs 'share' and 'disconnect' should be - and they rightly are - reserved for Linux's native filesharing.


You are missing the fact that both smbmount and -umount are tools for something (SMB) that is basically non-native for Linux. For chrissake, it is even called Windows filesharing, isn't it? That is the reason it is placed in the Tools menu.


There are Linux native tools for sharing folders like NFS, which surely should have the first call for words like 'share' and 'disconnect' in a Linux file manager. And they do, which is as it should be.

#

Re:No smbUmount menu selection

Posted by: Anonymous Coward on July 11, 2004 04:08 AM
Oh, right, like the user knows the difference between a Linux native share and a Windows share?

Jesus, THIS kind of thinking RIGHT HERE is WHY usability issues never get solved!

Even the term SHARE is braindead! What the HELL is a "share"? Do you mean to say "FILE THAT IS AVAILABLE FROM ANOTHER COMPUTER", maybe?

Geek Morons! Jesus Christ! They're absolutely BRILLIANT BUT CLUELESS!

#

Re:No smbUmount menu selection

Posted by: Timo Pirinen on July 11, 2004 04:22 AM
What the HELL is a "share"? Do you mean to say "FILE THAT IS AVAILABLE FROM ANOTHER COMPUTER", maybe?

No I don't. Haven't you ever heard of a printer share? And why do you think two computers are needed?

#

Re:No smbUmount menu selection

Posted by: Anonymous Coward on July 11, 2004 11:32 AM
maybe NFS should be defaults for sharing. However NFS is no more 'native' to linux than SMB is, so that's a fairly weird argument. The underlying mount protocol should be hidden from the user, just like umount.

#

Re:No smbUmount menu selection

Posted by: Anonymous Coward on July 19, 2004 06:58 PM
See, replies like this make me feel developers are sticking to their own structural logic....not unified token logic, which is closer to user understanding....

#

Get proffessional help!

Posted by: Anonymous Coward on July 10, 2004 05:20 PM
There are already loads of usability docs on the MS website which somebody could try reading, not to mention loads of books on the subject. The real problem is that programmers and geeks are trying to be something they are not.
A recent study found that the majority of open source programmers were pros who were going home and programming in their spare time. when the open source community needed better icons and graphics some artists came along and helped. There MUST be some professional usability people using the software who don't realise they could help out.
The best solution would be to ASK FOR THEM TO HELP. make a new project and advertise on newsforge and other sites asking for people who already know about this stuff and then the problem will be solved.
Integrating apps and communicating with other projects wouldn't hurt either<nobr> <wbr></nobr>;o)

#

usability / design :: science / art

Posted by: Anonymous Coward on July 10, 2004 06:05 PM
usability is a subset of design. the problem is a lack of design. to trivialize design as "useability" is a mistake.

design requires a different culture and hierarchy than is common to engineers. engineers who get visual culture are few and far between. Steve Jobs is one rare example.

open-source software will never have good visual design unless more artistic or creative minded people are drawn into open-source culture. usability without good visual design is worthless. you might as well hire jakob nielsen design your UI.

#

Re:usability / design :: science / art

Posted by: Anonymous Coward on July 19, 2004 05:27 PM
sentences without a beginning capital letter are difficult to read. especially when running mozilla on a 1280x1024 15" crt at default text size. the periods at the end of sentences seem to end up a single grey pixel.

#

Ditch the Concept of User Friendly

Posted by: eliwap on July 10, 2004 06:25 PM
Summary of a real life event...

I knew a very smart lady. A computer novice studying to become a doctor. She is one of the smartest people I have ever met. I tried to teach her about computers and showed her about menus and how to open a file. After showing her how to do it, she was nearly in tears when I asked her to replicate the process. The moral of the story is that there is no such thing as a user friendly computer interface. Maybe... user friendlier.

Applications are just that they are applications. Forget about consistency across programs. A multimedia player is not a spreadsheat application. And a spreadsheet application is not a graphics application.

Look at the modern living room. A VCR, DVD, TV and Stereo. Each device (application) has there own series of buttons. Each has their own menu of options. And each thing does something slightly different. And each thing takes some learning to be able to master.

Same thing with computer applications, and there are a lot more of them on a computer then there are in your living room. Instead of thinking about a consistant UI across applications. What is needed is task oriented rather than interface oriented thinking. And once this is done, think about it again, and then again. The people programming an app has already learned how to use the app. So... for them its easy. It does not mean that it is easy for a person new to the application. In fact, it might be damned hard, or nearly impossible. Like, menus that hide infrequently used tasks in favour of those more frequently used. And when a user needs to use a function that is infrequently used, where the heck is it, or more to the point, because the user has never seen it, it doesn't exist.

Thank MS for turning the average user into an idiot. Which brings me to another point. Stop treating the average user as an idiot. There is a universe of difference between a naive user and an idiot.

The best example I can remember of task oriented thinking was Wordperfect 5.x for DOS. Hit F3 for help. And the help system asked the user to mimic the keystrokes required to perform a task as the user searched for more information about it. It took a bit of learning, but everything was easily learned and once learned, the task was mastered. Its not "where do you want to go today," but rather, "what do you want to do now" that facilitates an easier interface. And that is dependant on the application. More specifically, the task in the application. Think of each task whithin an app as a mini application, just like a car (ignition, windshield washer, turn signals, brakes, gears, accelerator, stereo, dashboard, etc). What is the easiest and best placement and organization of the tools required to perform a task and once that is done, every element in the interface will fall in place.

#

Re:Ditch the Concept of User Friendly

Posted by: Anonymous Coward on July 19, 2004 10:53 AM
Yes, yes a thousand times yes. As a producer I work with both programmers and designers. Programmers do amazing work -- and almost always think in terms of efficiency for the system -- as opposed to efficiency for the person using the system. User task-oriented thinking is definitely the way towards more usable functionality.

And more usable functionality for open source can only benefit those outside (and inside) the technical community who might use it... and therefore advance the cause of Open Source...

M~

#

Re:Ditch the Concept of User Friendly

Posted by: Anonymous Coward on July 19, 2004 07:01 PM
You talk about computer organization, task-oriented user interfaces and what more. The problem: I can't read the whole comment without getting a headache.

Do you expect people to take you seriously on these topics when you fail to structure your own comment decently?
Use whitespace, insert some linefeeds here and there, _apply_ the principles you're advocating.

#

Nice text

Posted by: Anonymous Coward on July 10, 2004 06:25 PM
Thanks for this text. It really asks some good questions. Some of them have been (partially) answered by some development teams (e.g. GNOME), but I think the real issue is audience shift here. Linux has traditionally been a domain of technically educated users and system administrators. For the first time, developers of open source software are faced with Joe Average.

As the time goes by, the developers will solve all of the issues. I have no doubt. It'll be a long a tedious process, but we don't have to be afraid of it. We simply need to start viewing thing from a slightly different perspective - that of your average user.

#

Usability? Start with File Manager

Posted by: Anonymous Coward on July 10, 2004 08:47 PM
As a recent convert to Linux one problem I see is the lack of a windows explorer-like file manager. Attack me if you must (or better yet, point me to a windows explorer-like file manager), but I want to see my data and drives a certain way, and every FM I've tried falls short. The Xandros version seems to be closest to what I want, but even it is less user friendly than windows explorer. This would seem an easy fix. Am I wrong?

#

Re:Usability? Start with File Manager

Posted by: Mike Johnson on July 11, 2004 12:05 AM
Yes.<nobr> <wbr></nobr>:-)

All the clueless Windows users I know (and there are plenty of those) have no clue *where* their files are. When they save a file in Word, they click save on the first directory it shows, they don't browse to another folder. We get frantic helpdesk calls when Word decides to open to a different folder. "Help! All my documents are gone."

Forget filemanagers, people don't use them like you do. What's really needed is a unification of save and open dialog boxes. They should all look the same and all default to the same folder, ala 'My Documents'. Regardless of the toolkit or desktop environment, everything from WindowMaker to Gnome to KDE should agree on a common folder and set of shortcuts to display.

That way the filemanager would be mostly irrelevant like most users expect.<nobr> <wbr></nobr>:-)

#

Re:Usability? Start with File Manager

Posted by: Anonymous Coward on July 19, 2004 06:56 PM
Computer users who don't get the concept of file location within the filesystem, should be educated. That's the only viable alternative, instead of dumbifying the UI to suit these users.

Look at it like this: if you don't know where to put the key in, you shouldn't be driving a car either.

Don't get me wrong, I'm all for a more user-friendly and intuitive interface, but what you're saying is simply going too far.

#

Related idea

Posted by: Anonymous Coward on July 10, 2004 10:20 PM
I had a <A HREF="http://www.livejournal.com/users/arvindn/3554.html" title="livejournal.com"> similar idea</a livejournal.com> a while back. My proposal is specifc to GNOME, but it could be easily adapted for other projects.

#

Iterative is the answer

Posted by: Anonymous Coward on July 11, 2004 02:51 AM
I agree non-English words like SMBUMOUNT should never appear in a GUI for normal users. But after obvious stuff like that is done its time to let some normal and advanced users try it. Apply what you learn from then and lets some normal and advanced users try it again....etc.

#

usability

Posted by: Anonymous Coward on July 11, 2004 03:07 AM
I beseech the kde development community to think before they act when it comes to usability. The gnome community decided that they wanted to increase usability, and the result was a massive feature-removal spree that ended when the gnome desktop environment was severely crippled.

Over the years, I have seen many wonderful increases in usability in the kde world - the ability to use single menubar in a panel rather than a seperate menubar in every window, the mouse-gesture/macro functionality, multiple tabs and panes in konqueror, etc.

I find kde extremely usable, but more importantly it is featureful. I dont think usability needs to be addressed any more than it already is. Any attempt to directly improve usability generally leads to feature removal.

There are more important things that the kde development community should be focusing on if they want to further improve KDE - decreasing the memory overhead for instance, or adding new features.

I think that designing new interface paradigms would be an extremely valuable use of time, but not if "usability" is the primary motivation. A suggestion: how about adding a configurable, ion-like pane mode for kwin?

#

Lose The Jargon

Posted by: Anonymous Coward on July 11, 2004 06:39 AM
Every specialty has its own unique set of code words and In-Group terms. It seems to be universal human nature. Members of a particular group can identify each other by whether or not someone 'gets' the meaning of a particular term.

Consider the example given in the article about the smbUmount menu entry. Here's a hypothetical conversation between a newcomer and the somewhat more experienced end user that he calls in for advice:

Newbie: Uh....what is this thing labeled smbUmount? What does it mean?

Vet: It unmounts the smb shares.

Newbie: Say what? What are smb shares? What does smb mean? What are shares? Why do I need to unmount them, are they sitting on something? And what happened to the "n" in umount? Is this a typo?

Vet: No, not a typo. Programmers are in love with abbreviations and acronyms. Just a quirk. And smb is an abbreviation for samba. So the menu item unmounts the samba shares.

Newbie: So what the heck does this command DO anyway?

Vet: Samba is a kind of network protocol or something, not real sure myself to tell the truth. But what it actually does is disconnect your file browser from the file system of whatever network computer you are hooked into that happens to be running Samba, usually a MS Windows box.

Newbie: You mean that in human language, it disconnects my computer from the windows systems?

Vet: Yup

Newbie: Then can you tell me why, in the name of all that is blatantly obvious, didn't they just label that menu item as 'Disconnect From Windows Filesystem' or something equally sensible?

Vet: Dunno. Ask them.

#

Re:Lose The Jargon

Posted by: Anonymous Coward on July 19, 2004 06:33 PM
Well, the obvious reason not to have "Disconnect From Windows Filesystem" as a menu item label is that it's too long. What's really needed here is context-sensitive help. I recall that some of the old DOS programs, including MS EDIT and Borland's C++ compilers, had the old "press F1 for help" convention on any menu item. Some Windows applications have a feature where if you leave the cursor over a button or other control, a little window pops up and tells you a tiny bit about the item. I think it would be good if a lot of X apps had a similar feature, but where you could then press a key, or maybe the right mouse button, and get more detailed help. Maybe some X apps do this, but mozilla sure isn't one of them.

#

Ask the user.

Posted by: Anonymous Coward on July 11, 2004 06:48 PM
How about you give the users of your software the possibility to tell you what they like and what they don't like? One possibility would be an extra button in the window frame, which composes an email out of a screenshot (draw a frame), a title and a text to the developer. In usability, there are no stupid comments, because if the user has a problem, it's either a bug or a usability issue.

#

Re:Ask the user.

Posted by: Anonymous Coward on July 19, 2004 07:01 PM
Good idea, and I've recently come across something similar in the world of online publishing. Bruce Eckel's online book Thinking in Java 3rd edition has a feedback link at the end of every paragraph. Apparently some of Bruce's friends cobbled together some code to help handle the system. Maybe he'd be willing to share it with the open source world.

Now we have to come up with a name for the new feature, preferably something catchy. Obviously, it should be activated by a menu item in the Help menu, something like Report-A-Problem, perhaps.

It should be configurable to either send the report to the local help desk or to the program's developers/maintainers, so that organisations can choose to keep their problems in-house if they want. The help desk might escalate the report to the next level. The average home user won't have a help desk, so (s)he'll probably send it straight to the top.

#

Poor Babies!!!!!!!!

Posted by: Anonymous Coward on July 11, 2004 09:29 PM
Populations that have the most to gain from Linux are left to learning an entirely new visual, verbal, and idea vocabulary on top of an unfamiliar method of production.
>
>
I guess people like you will *NEVER* get very far
in places where you need to learn the *NATIVE* LANGUAGE*
spoken there like Japanese in Japan, French in France,ect,ect,ect.

I'll say it again.

The *ENTIRE* usability "community" is composed of a bunch of freaking *MORONS*!

#

Re:Poor Babies!!!!!!!!

Posted by: Anonymous Coward on July 12, 2004 10:17 AM
But they probably know latin abbreviations much better than you!

#

Re:Poor Babies!!!!!!!!

Posted by: Anonymous Coward on July 12, 2004 08:53 PM
You are so clueless it hurts.

"Populations that have the most to gain from Linux are left to learning an entirely new visual, verbal, and idea vocabulary on top of an unfamiliar method of production."

THEY DONT CARE ABOUT IT. They dont like computers
at all. They just use it.
Like a dishwasher.
They expect things to work with a click without
needing to understand very much of it.
And they have the right to feel that way.

#

Re:Poor Babies!!!!!!!!

Posted by: Anonymous Coward on July 19, 2004 05:44 PM
It's not a question of stupidity, it's a question of differences in domain knowledge. For example, educated users of the English language know that the common abbreviation meaning "and all the rest" is "etc." (from the Latin "et cetera"), not "ect." (which I assume has something to do with ghosts, slime and 80s music).

#

Good comparison

Posted by: Anonymous Coward on July 19, 2004 06:41 PM
Chinese is a very easy language, unfortunately not for Westerners.

#

Bzzzt. Wrong answers

Posted by: Anonymous Coward on July 12, 2004 01:51 AM
Microsoft: Security? We'll throw enough money at the problem and it will go away.

Open source: Usability? We'll throw enough eyeballs at the problem and it will go away

#

Let's put ideas to work

Posted by: Anonymous Coward on July 12, 2004 06:51 AM
All coments are welcome, but why not to start working on it?. I propose (don't ask me where and how) to create a "tuxgui.org" website, let people (developers, artists or whomever want it) to post its designs, and then let brainstorm on them. I think it's easier when we have something visual (not visual basic thanks god) to start with. It would be nice also, to have tools for a rapid gui prototyping, maybe something like Karamba or gDesklets?.

#

Let's get the software working first

Posted by: Anonymous Coward on July 12, 2004 11:19 AM
Well let's see... Certain precedents have been set, and we should stick with them (i.e. things like right click for a menu, an x button closes, etc). As for making it ultra uniformed... why bother, Linux is about choice, and I feel something designed for the lowest common denominator may be irritating and useless for people wanting a little extra punch. Also, although I haven't touched a mac in years, when my jr high finally upgraded from the good old commodore 64's, to lc550's... I was totally lost. I was a “happy” Windows user at home, and I found everything on the mac totally unintuitive and couldn't really figure out how to navigate well. Same thing happened years later with the iMac's at my university. Anyway, when everything else is done you can worry about people not figuring out what an icon of printer does, or what File save as means... You want a good example of UI predating this stuff? Look for old sierra games from the late 80's<nobr> <wbr></nobr>;)
Yours,
BoyOfDestiny

#

Useability is in the eye of the beholder

Posted by: Anonymous Coward on July 13, 2004 10:36 AM
I wholeheartedly agree with some previous points made. With just 2 years as an administrator of an I.T. department I can tell you that the following points are true.

1. Users get confused by too many ways to do one task.
2. Users do not like having so many things coming at them that they do not understand whats going on with the computer.
3. Users would have just as much pain learning a Mac as learning Linux, it's the pain of change which keeps most users from truly trying to learn something new.


        I agree with making tasks simpler for the user as the way to win more over, but I also believe that user attitudes toward something new must be changed. If your users do not want to use something new they won't try and learn.

So, the rules to usability.

1. Leave advanced options for advanced users.
2. Keep It Simple And Stupid.
3. Provide meaningfull help to users to complete common everyday tasks.
4. CONSISTENCY must exist along with number 3.
5. Win your users over with the benefits of Linux.

I think this is achievable with our current and future open source developers. I also believe we can have competing desktops which each provide their own unique way of accomplishing this. All we have to do is keep the big picture in mind while forming the details<nobr> <wbr></nobr>:-)

#

Usability is already defined

Posted by: Anonymous Coward on July 15, 2004 05:34 AM
All you need to do is pick up a book titled "Human Interface Guidelines" written by the good folks a Apple computer about 20 years ago. It is still the defacto standard for usability. If you want to know how to build useable GUI interfaces, look no further than Apple computer. Play around with MacOS X and then you'll know how to build your interfaces.

#

What people like, they dont know yet.

Posted by: Anonymous Coward on July 19, 2004 05:04 AM
Usability isn't a problem.

Character, and feel of the software is.

I wish programmers for Linux gui related software would scrap their whole idea that Windows, with menu's and text and icons is the only way.

I would prefer a GUI that is more device like.

I think that alot of people would prefer it aswell.

I think also that your typical 'computer illiterate' person doesn't actually know what they want.

It's a Graphical User Interface, not a Human User Interface, I think that it should be a graphical interface that maintains the feel of the machine it is running on.

The Linux console maintains this, which is why I love it so much.

#

Make Usability Teams and Projects join these teams

Posted by: Anonymous Coward on July 19, 2004 06:11 AM
The problem about usability is that in a project there is not much to be done about usability. for a small application there is normaly only a few screens and for a bigger one maybe 20 to 40. This means to little work for highly specialized usability experts.

But these experts could join in a team and for a usability group. They can have there own experts on each language, even there own expert for several languages. making it easier to make applications multilingual from the start. They can make their own standards, usability reviews and tests, and educate their members.

Also standard component designers can put up their usability team, like for example wxwindows could as an example. So if you use wxwindows you should invite this usability team for help, cause they know their tool, its strong and weak points.

Well these teams then can look for projects they want to make USABLE and standard. They can ask the project leader if they can do the usability/screen designs. Or project leaders can ask a team to alocate force to their project.

These teams can also make money as they could be hired by small companies who lack the size for usability experts and hire these specialized people.

Anyway it is just a concept, but i think it could work.

john.vanderpol@wpg.com.py

#

A major role here for OSDL

Posted by: Anonymous Coward on July 19, 2004 07:01 AM
Firstly, it's great that this issue is on the agenda. It's desperately needed.



Surely the Open Source Development Labs (OSDL) [made up of the who's who of the Linux industry, including Linus himself] is the main vehicle that should be driving the useability issue? After all, it does have a Desktop Initiative as one of its main activities. <A HREF="http://www.osdl.org/" title="osdl.org">http://www.osdl.org/</a osdl.org>



The GNOME initiative has been a great start, however the challenge is that both GNOME and KDE are going to be widely deployed, and it would help the Linux cause if there was some standardization regarding useability, without taking away the flexibility to customize.

#

Re:A major role here for OSDL

Posted by: Anonymous Coward on July 19, 2004 08:36 AM
The point stands - everyone is saying it is not our problem.

Accept it and start on fixing it.

Yes the gear selector does not need the same interface the stereo does in your car - valid point.

Ever figured out those icons for fog lights?

Don't dumb stuff down, do think about what the user wants to do and refer to it as such.

Acknowledge someone has a point - prioritise the problem - decide what to do and when.

People are flexible - many use the start button on their way to stopping a computer.

How many control centre apps, print managers, sound managers, weird ways to get multimedia keyboards working do we want?

At least now we've progressed to caring about the bloke next door using Linux.

A slick interface is all everyone ever wants - it is important enough to demand humility

#

Ask your Grandma

Posted by: Anonymous Coward on July 19, 2004 08:03 AM
to operate your program. If she fails, it's no good

#

Re:Ask your Grandma

Posted by: Anonymous Coward on July 19, 2004 10:29 PM
In that case, I can ditch my computer altogether...

#

stfw!!

Posted by: Anonymous Coward on July 19, 2004 08:28 AM
It seems that a lot of responses are the kind of this comment's title. I don't think that those assholes are representative of nothing, fortunately.


There are a lot of material out there that can be useful. I liked <A HREF="http://c2.com/cgi/wiki?InterfaceHallOfShame" title="c2.com">Interface Hall of Fame (and shame)</a c2.com>. Very useful... and funny.

#

Why require mounts anyway?

Posted by: c++ on July 19, 2004 11:49 AM
I see the question of what to call "smbUmount" as less important than the following question, which strikes at the heart of UNIX design: why should I ever have to mount or umount a share? Why can't I just "vi<nobr> <wbr></nobr>//server/share/dir/file", respond to the username and password prompt, and be done with it? This is lots easier to use and to explain to new users. (the share syntax, not vi<nobr> <wbr></nobr>;) When training people on Windows, I avoid letting them know that they even have the option to map shares to drive letters, or to a mount point somewhere on C: (yes, NT can do that). Such information would just confuse the hell out of them.

Yes,<nobr> <wbr></nobr>//server/share/... would require rewriting vi and every other app, so how about<nobr> <wbr></nobr>/outside/server/share/... instead? vi (and emacs) would never need to know. If $DISPLAY is set, and the username and password to the other machine isn't stored in the user's local account information yet, the OS could pop up an X dialog. Otherwise, it would be forced to return an error to apps that don't understand the file sharing mechanism.

#

Re:Why require mounts anyway?

Posted by: c++ on July 19, 2004 11:57 AM
Just in case someone misunderstands, I wasn't talking about shortcuts when I said NT could mount; that's a totally different feature. Users love shortcuts. Unlike mount points, they hang around forever, or at least until that user deletes them.

#

start with Learning

Posted by: Dmitri Zdorov on July 19, 2004 12:03 PM
I am usability expert and for me it is obvious that:
1. Problems of usability in OSS are HUGE.
2. Over all it probably now is THE problem.
3. Problem can not be solved without higher level of education and knowledge (in this field) be most developers is OSS community
4. There are tons of books and websites on the subject, just go ahead and read some of them. There is at least one website that should be read from top to bottom by all interface designers (useit.com) and there is at least one book that is the MUST - Design of everyday things by D. Norman
5. Usability is not a matter of taste, the same way is decision to put pull or push label on the door.
6. I like this phrase: “Those who ready to sacrifice usability for performance will loose both and deserve neither”. You should like it too.

#

Re:start with Learning

Posted by: Anonymous Coward on July 19, 2004 06:26 PM
First of all. I do think that usability is a problem of open source.

but: it is not a problem of KDE, it is more a problem of the Environment: Howtoinstall new software, drivers and so on. Also writing software at Unix plattform is difficult. Inconsitent or non-existing Documentation is a real problem.

Books and websites: You don't have to adress the issue from a psychological perspective. You just have to write reports and look what can be improved. You also don't need to have defined HIGs.

When certain components are composed to a software it often happens that one does not fit to the other. But this will be improved.

There are usually problems of the kind:
* information sinks (who is responsible?)
* howtosolve a problem? A guide tell you to log in as root. So you have to know what a root user is, howtodo it.
* skill barriers
* non-documented issues.

#

Re:start with Learning

Posted by: Dmitri Zdorov on July 20, 2004 06:38 AM
Usability problem is a lot wider than Linux and KDE or environment or how-to or drivers.
It is still pretty bad in Windows world and in OSS it is at a disaster level.
Difficulty of writing on UNIX/Linux is not an issue.
The issue is that vast majority of OSS developers do not know much about usability at all. Thus they do not understand importance and do not understand where to start and what are the real problems.

Main problem of usability in OSS is priority.
That is set wrong in heads of all developers. They are trying to improve wrong things most of the time.

#

KDE leads in Usability

Posted by: Anonymous Coward on July 19, 2004 06:15 PM
From my perspective KDE leads in Usability. The "noise" is from the KDE newbie proposal/discussion list, the sandbox of usability discussion. Usability research studies like that one of Relevantive were taken serious. Persons who write deep analysis and don't make just nice proposals to turn KDE into WindowsXP and discuss with each others. For usability research it can be important to make usage statistics. For usability issues there is a format: http://usability.kde.org/activity/usabilityreport<nobr>s<wbr></nobr> /usability-report-guide.php

KDE can improve a lot, can do better. But I believe it is the best DE in Usability out there. Windows XP is bad in usability for semi-professional users, because you are restricted in an non-intelligent way. Gnome removed all the features... Many gnome apps clone the Win-interface. Mac OS X is an high-price OS that is nice but not as good as the Mac crowd says when it comes to usability.

there are a few usability thing:
* a Theme manager in needed and will get shipped with KDE 3
* automount
* Plastik instead of keramik

#

RE: Software Usability

Posted by: Anonymous Coward on July 20, 2004 12:14 AM
This is a topic has been discussed at length in our little LUG.


    It seems to me like UI design and more traditional coding are different skills that tend to attract different people. It would be very nice if it was practical to separate/abstract interface details from the underlying applications in a standard and cross-UI way.


    Assuming it is possible and practical (who knows?) it seems to me that working this way would fit the Unix philosophy (small tools, modular design) much better than current monolithic approaches.


    As a thought experiment I've contemplated how this could work for an application developer, a UI developer and a user :


    Application Developer : The application talks to the outside world through abstract hook objects allowing the app developer to concentrate on the nuts and bolts. An interface is automatically generated when the application started so it is usable in most GUIs and UIs, and sometimes the autogenerated interface is even created in a logical format! If the application developer dislikes the machine generated interface produced in his preferred GUI environment he customises it using something like Glade, but leaves customisation in the multitude of other environments to the UI experts (if they feel so inclined).


    UI Developer : It's a joy to not have to rely on apathetic developers to use the UI as it was intended to be used. UI's can be kept consistant, and new experimental UIs can be road tested with production apps. Freedom!


    Users : Applications are available no matter the interface so a UI choice can be made on the important things (features and usability). The apps can be used in ways the developers didn't even envisage ie. used through mobile phones, speech synthesis and/or translators, web interfaces, the commandline, neural interfaces (why not?)... an interface can even be exported for use with your favourite scripting language. Perhaps applications could be graphically grafted onto eachother (the GUI equivalent of piping etc...) Applications are no longer made redundant in new UI environments as they pick up new interface elements automatically, though not necessarily very intelligently so may require some manual tweaking..


    In summary I think it should be said that we're Free Software people, and shouldn't be looking to constrain anyone to a single environment.as in the proprietary software world. I don't know if what I suggest is even possible, but even if only a small subset of applications have an interface simple enough for this kind of treatment it could save a lot of time and effort, make UI and app developers more independant of eachother and make new things possible.

-Mark


   

#

Re: Software Usability

Posted by: Anonymous Coward on July 30, 2004 07:57 PM
http://www.uiml.org/

I think what you're proposing is what the User Interface Markup Language could possibly achieve.

-JensR

#

Comic Strip about this article

Posted by: Anonymous Coward on July 31, 2004 12:24 AM
<A HREF="http://www.ok-cancel.com/archives/week_2004_07_30.html" title="ok-cancel.com">http://www.ok-cancel.com/archives/week_2004_07_30<nobr>.<wbr></nobr> html</a ok-cancel.com>

#

Usability? What is it at all?

Posted by: joaobr on August 01, 2004 08:39 PM
like Frans said here: lots of discussions and no result but also claims we need to hear and learn and he tell us we need more data but not from outsiders?

This confusion mirrors the real situation.

Instead of defining what we need we should first define better what usability in fact is.

This endless discussions we have are good but analising them well they are all technical discussions or soon turn into it.

Usability is absolutly relative.

Usability therefore is a very difficult issue, much more complicated as the most complex code because it is rated by THE user.

And the user is a relative thing. His opinions are based on a mix of skills, habits, knowledge, intelligence, tasts and may be in first place interests.

So then we need to know what he wants. Means, usability can have a different definition for the Internet surfer and for the letter writer, for the developer and the gamer.
This perhaps could lead us to develop profiles which could be chosed by the user. As an example the Internet surfer get Konqueror, Kmail, Konsole icons on the panel and the bureau slavy get the koffice buttons for quicklaunching his apps.

Ok, this obviously does not change anything to how easy an option could be found and how easy this option could be set.

Now we could believe this is a technical issue but this is the point - I do not think so: Usability is not a technical area, Usability should define what is good and needed and the technician only cares HOW to achieve it.

This brings me to the idea that Usability people should not be technicians or developers.

Usability people are the interface between the user and the technicians. When the usability person is also the developer he ever will be guided by his logical and technical manner to think about. But sure, this also is not a so bad thing.


We can learn Usability basics here: The best loved car could be rejected in Australia because people are used to sit on the other side<nobr> <wbr></nobr>... and that is probably the best example for how to think and what we need to do.


Overall we should not forget that KDE is a feature rich system with so much options and not only a nuts screen where the user can change colors and fill the desktop with icons. Who learns KDE will like it more each day and becomes natural a better user. Soon we have better users we will get more users and this brings more hints how to improve usability.

Who does good work do not need to look for critics which come automatically. I think that we can ignore a little bit with patience the usability problems and honor more the great work which is done so far.

#

Re:Usability? What is it at all?

Posted by: Anonymous Coward on August 02, 2004 08:49 PM
This editorial and the subsequent comments are the most damning words I have EVER heard from the mouths of OSS developers. "Fuck off users, fuck off usability specialists, we will continue blindly copying MS and Apple UIs, producing crap that is unpleasant or too technical for anyone but developers to use."

Enjoy that zero desktop market share growth, guys.

#

Why usability specialists ARE necessary

Posted by: Anonymous Coward on August 02, 2004 09:14 PM
Your essay exemplifies why usability specialists ARE necessary. Usability is NOT about font color and what phrase to use. If that were the case, then yes, existing standards and HOWTO docs would be sufficient.

Usability is about working with end-users to understand how users work and think. New development is often targeted towards solving novel problems that do not naturally map to how users think. A HOWTO document will not describe novel tasks. Someone must sit down with users to understand how they think about these tasks before an interface can be built, which leverages their existing mental models of the task. A usability specialist is not necessary for all application development. But if you are creating something very complex, innovative or targeted to specific user populations, then understanding your users is critical… and again, that is something you will not find in a HOWTO document.

#

Read about it if you don't want to hire someone!

Posted by: Anonymous Coward on August 09, 2004 09:27 PM
Wether you like it or not, usability is a specialism, just as coding is. Programmers can learn about usability (And I think it would be extremely useful for their skills as well as their perception of the business) but it takes time because it requires a way of thinking. And definately beyond labeling alone.

So if you don't want to consult an external firm - and why should you hire, there are probably a lot of usability experts who are willing to help you for free - then read about it and broaden your horizon.

A few -easy to read- books that give you some ideas about the psychological and project aspects of usability are:

Jesse James Garrett: 'The Elements of User Experience'
Donald Norman: 'The Design of Everyday Things' and 'The Invisible Computer'
Steve Krug: 'Don’t make me Think'
Alan Cooper & Robert Reimann: 'About Face 2.0: The Essentials of Interaction Design'

Some of these are for internet only, but the basic principles apply to all interfaces. Usability isn't something elitist but simply a way of listening to your users and helping them use to your product without making them rtfm<nobr> <wbr></nobr>:)

It is not a matter of us and them. Not where development is concerned, but it is about making the obvious better software the most popular and easiest as well.

(I'm into usability and coding, but have tried in vain to install linux)

#

This story has been archived. Comments can no longer be posted.



 
Tableless layout Validate XHTML 1.0 Strict Validate CSS Powered by Xaraya