Slashdot Mirror
User Feedback and Open Source Development
Earl Shannon writes, "With the new release of Sendmail I was looking over Sendmail.net and
came across this essay called It's the User, Stupid. Author Mike Kuniavsky states a very important question if Linux wants to make inroads on the desktop as an open source solution. How can the open source development model obtain the necessary user feedback to development interfaces that the user will intuitively able to use?"
I think that the author's point was that a GUI is more than just the graphical icing on the cake, it's the cake and the icing. Part of what makes Windows easier to grasp is point and click installs and plug and play, etc etc. If this is what the author meant, then you're in agreement (as am I).
METAPHOR is the power at work behind this ease of use which we are so bold as to label INTUITIVE. Applying the word intuitive to something as copmplicated as a computer--and following the strictest denoted meaning of the term--is of course, an overstatement. However, to be really strict, one can debate the intuitiveness of anything. Once we consider all terms as relative and not absolute, Linux is back to where you can't wiggle away from the fact that without a comlete GUI, Linux is a highly "non-intuitive" system to learn.
When you sit a completely new user, a child let's say, in front of a Linux Console and a blinking prompt, that child has NO CLUE, HINT, OR INDICATION what to do to start learning the system. In front of Xwindows environment like KDE or Gnome, they will see that there are objects that can be manipulated. There are things to read, which may be read with no more effort than clicking on them. They will hit the help files literally by accident. It's even likelier that they will find them on purpose as they are marked by a question mark. How long do you think it would take for a naive user to randomly stumble onto the combination, "man man"?
This is the beginning of exploration. Moving files from this place in the filesytem to that is simply easier for new users to do and feel confident in themsleves doing when they visualize the task metaphorically, substituting a file icon and a motion for a command in which they would have to spell out the filename precisely and observe command syntax without error.
With enough exploration comes competence, then later on mastery
The second child can begin learning for themselves with virtually no prior knowledge beyond their native language, basic human experience of manipulating objects, and proceed from step to step, and even, at the level of mastery, understand and manipulate the system without any help from the guiding metaphors at all .
The first child, staring at a black screen, would not make any progress to any level of understanding whatever, much less mastery, without massive assistance and constant intervention of an external helper.
There is, in short, a reason why personal computing has flourished in the "GUI" mode, and has almost completely abandoned and wiped away vestiges of pre-gui modes of computing in our cultural ideas about computing is. (Don't believe that, just ask kids to draw you a computer "at work". When I was a kid, you would have seen drawings of a large box with "das blinken lights" and large tape reels, maybe a monitor with text interface stuff and a cursor block. Nowadays you'll see a window with buttons.)
This reason will not change for linux gurus (not to be insulting, but yall's birthrate is just not high enough to affect the demographics of the situation) and in fact, if they do not clue up they will be completely beaten into the ground by Windows2000, which they may choose to regard as inferior. Businesses will continue to choose Windows at the desktop and the server for the simple reason that the product promises that relatively low-skilled workers will be able to handle the basic administration of users,apps, and common network configurations, and that these workers will be found in abundance; ie: quick and cheap. It won't be a quality thing, it will be a salary thing. It's so easy to see that only a Linux guru could misunderstand it.
Good luck to you guys--if you're still debating the "intuitive" thing, you're gonna need it!
You miss the point he's making. _You_ are not the typical desktop end user. You are what he refers to as either a programmer user, or a sophisticated power user.
The result is, that OSS project GUIs are designed without reference to the _typical_ desktop software end user because:
i. Most typical desktop users don't use OSS
ii. Most OSS projects don't do user studies _except_ to get the opinions of those power users and programmers who are involved in the project/mailing lists, etc.
Just because _you_ are involved in the GUI design of the GIMP, doesn't count as doing studies of typical desktop software end users. When the designers of the GIMP go into a real graphics or publishing shop (the kind of place where everything is done on Photoshop on a Mac, in other words) and actually study how real graphic designers work, then you can claim that they've done real end user studies, not before.
i don't really think anything can be intuitive. to understand even windows, you have to have some degree of experience with computers. if i was to buy a computer, with no prior experience using windows, i would be as lost as someone coming from windows to unix. more so, probably. example: this little plastic thingie comes with my computer, i heard it called a 'floppy disk' at the store. they say i need put it in the floppy disk 'drive' to install the drivers? something like that. ok, it's in...now what? oh, i click on the A: drive? .... you see, that's not intuitive. and how would one make this intuitive? you can't. it requires some learning to even do that simple task. no user interface can really be intuitive. computers just aren't that simple, yet. not much is. what can one really do succesfully with no experience? i can't think of anything off the top of my head. anyway, i think people should get off the intuitive thing. intuitive and ease of use aren't really the same thing. i believe people should concentrate on writing better documentation for their programs. even the most complex program can be deciphered if there is good documentation. :)
Yawn. What a non-issue. Frankly, I couldn't care less if some weenie incapable of reading documentation or editing a configuration files decides they want to use some other piece of software or switch to Mac or MS Windows. If you want a nice user interface either sponsor the programmer/s to put it in place or quit yer pointless whinging.
Or like clicking on the "Start" button to shutdown your computer.
:)
Makes sense to me. Click Start to start shutting down your computer.
There are some 21 ways to leave your application...
Alt-F4 almost (>90% of the time) always works.
The average user does not care HOW the computer works. The average user wants it to work without "General Protection Fault" and other stupidities. MacOS is much better than windows at simply running. (My experience.) A computer OS needs to be like a good car: you can run the ever living hell out of it and still depend on it running the next day. Like I say most every day; if cars ran as well as windows we'd be walking. But the car has got to look good too. There's no single "intuitive" interface. There is more than one way to skin the cat, though. Look at one of my favorites: MacOS (And windows.) Command-c is copy, but so is copy from the edit menu, or from the copy button on a tool bar. What is intuitive to you is gibberish to me. Hell I'm still lost when DOS gives me abort retry fail. I'm less of a programmer and much more of an end user. I spend most of my time arguing with techies who are very technical but haven't the foggiest notion of what I really need the computer for. I need it to run. I need it to run 24/7 with little maintenance and no frigin' crashes. I need it simple enough to be explained over a cell phone at a truck stop in the rain to a guy who barely graduated highschool. Look at it from a secretarie's point of view: we put a guy on the moon with a computer that wasn't much more powerful than a pocket calculator. But 33 years later she can't get her 500mhz Winduhs machine to talk to her printer.
But on the Mac you can drag a layer from photoshop or an email from Eudora onto the trash and it will delete the appropriate thing in those programs. Thats where "intuitive" really shines!
Not that I'm eager to defend Windows, but much of what you describe on the Mac is true under Windows 98, too. I can delete e-mails in Outlook by dragging them to the Recycle Bin. I can delete a chunk of text in word by highlighting it and dragging it to the recycle bin. I can reorganize my Start menu by dragging items around on the menu, and can delete items by, you guessed it, dragging them to the recycle bin.
>>Windows is horribly unitutive and relies on so >>many just plain stupid techniques.
>You mean like dragging a floppy disk to the >trashcan to eject it?
No, that's no good. Copy it to 50 remote hard drives and hold the floppy to a magnet.
""
Dragging a sheet-of-paper icon to a waste basket to delete a document is intuitive for me and probably for 90% of the Western world.
""
Right you are. On that particular example, Windows and UNIX stops there.
But on the Mac you can drag a layer from photoshop or an email from Eudora onto the trash and it will delete the appropriate thing in those programs. Thats where "intuitive" really shines!
I use a mac as my desktop and telnet to linux boxes because there are so many great features
that combine into a great desktop environment.
Theres a short list of really great idioms in each of the windows and mac wand os/2 and beos and amiga and nextstep worlds that would propel the open-source desktop into new levels of usability.
Here are my top 5. Four offive are the mac - i have used Windows and GNOME and KDE for years and months so its weighted towards MacOS through experience not fanaticism.
Pop-up folder windows. On the Mac, you can drag an open folder window to the bottom of your screen, and it will become a tab. This is a great place to put icons that launch applications, or saved games or folders that you save documents to or utilities to drop files onto. They open as you drag a file over them and you can get 100 to 1000 icons accessible with two clicks very easily. Drawers on the typical UNIX desktop come close but dont match because pop-ups can have different displays and can be a 2-d display.
Single-menu-bar. I did a task-centered comparison of a single menu bar as found on the mac vs. the menu-ber-per-window as found on windows and the mac one by far. KDE does this ok! Kudos to KDE.
Sub-menus that can be opened on the sub menu. OS/2 and the Mac have this. Lets say you have a
sub menu called "Recent ->" under "File". The idiom is, click the Recent instead of clicking one of its sub-menus and the contents of the sub menu will open up.
Tray icons. Windows uses this most extensively but the Mac has this too - basically applications can add a small icon to a tray in the taskbar or on the right edge of the menu for macs, that can offer a menu of quick, omnipresent actions, and display feedback of what is happening.
Hide other windows. Well every gui has this but some do it better than others... window-shading is superior to minimizing because the user can remember what window is what more easily. Given the woeful titles that X-windoows windows usually have in a the taskbar (all my netscape windows if I have 20 open will say Netscape: and thats it)
we should consider the taskbar a failure and find a better way of switching applications. Being able to hide applications, and hide all other applications, is especially useful. And would be even on the multi-page desktop standard to UNIX.
All of these idioms are well within the reach of the KDE and GNOME and Enlightenment geniuses out there. Part of the problem for us in the open source world is interoperability however. Because theres not a common way of doing certain things those things just don't get done. Here is my list of badly needed idioms across the open source desktop world.
1. cut and paste and close window keys are in complete disarray. Its a pain to copy a simple URL between netscape and a terminal or get out of an application! Horrible!
2. installing an application should add a goddamned icon into the GUI automatically. If I rpm -i an application, then it should appear under a menu somewhere, and it should be relatively easy to copy that icon to the place I launch applications from. Installation on UNIX is still a ten step process (much better than the 50 steps it took before RPM) but it should really be 3:
- click URL
(web applications downloads RPM)
(RPM manager installs automatically)
(Icon gets added to the GUI)
- start program and learn it
- make another icon on desktop or panel
where the user likes it
3. Clicking an document icon should launch the most appropriate application, not just the application associated with the mime-type/extension. This is a broken idiom in UNIX because of the lack of meta information available with each file. There are some standard extensions, and its possible to get use the magic numbers using 'file' but there are so many kinds of text files out there and I want to edit some of them with emacs and some with notepad. Basically a system-wide mapping of extension to application does not work well in UNIX because customizing it is not fun (system-wide tables tend to get too long to be manageable) and applications dont have a central self-registering doctrine like on Windows.
Thats my rant. Hope I made sense. The bongloads really don't help...
A true test of the GUI is how usable is it when you are stoned. UNIX is just not fun on weed. Mac is great! windows is tolerable.
Well... If you're gonna get picky. The browser is NOT part of the KDE window manager. It is a part of KFM. The KDE window manager (KWM) runs on top of KFM.
I'd like to add to my comments: Mac users use the apple menu much more than Windows users use the Start menu. Reason for that probably is that any administration for Mac must be launched from Apple menu; in Windows control panel is accessible from "My Computer". Also, users like (and use) aliases to programs from Apple menu. In Windows so much crap is in the Start menu (my guess), they leave it alone because they can't find what they want quickly. Unless someone prunes it all back for them. in response to some other poster, Windows taskbar is a version of the Next dock. Apple's equivalent is the vertical list of windows under finder, which is a little slower to operate since it requires opening of the list, but if many windows are open in WIndows, the vertical list works better since it's less confusing. I don't know of a perfect way for this function to work. Most X environments (where i spend most of my time) imitate the taskbar idea. I find myself not making much use of that, so I'm not sure if this is a Windows-ism I'd like to see in Linux environments or not, but ,personally, I have banished the "start" button. Frankly in WIndows and in Linux environments too, it's often used as a way to prove how much shit you have on your box, with the assumption that more is more impressive.
Personally I think the placement of the taskbar is trivial but at least you have a choice of where it goes. Try moving the menubar on the Mac.
People growl at the thought of having to edit a text file to make an adjustment or configuration. Geeks say "Awesome - text file" and whine to people to just learn how to do it.
This is exactly how I feel. At the college I go to, the main person who teaches multi-media for the computer science department uses a WYSIWYG(what you see is what you get) HTML editor called Dreamweaver. I've been using HTML for the past 3 or 4 years and have gotten fairly good at doing things in notepad/vi/whatever. Whenever she tells me to do something she will go "Just open up Dreamweaver and yada yada yada..." But I cannot stand the program. To me, it has a complicated interface, that is severly non-intuitive. I can just go into Notepad, and be done with it. Yet she teaches her students how to use Dreamweaver, and just enough HTML to fix things when Dreamweaver goes wrong.
To her and all her students I'm a psycho doing it the old fashioned way even though I can probably out code them and create web pages twice as fast then they can in their intuitive program.
One of my favorite quotes from RMS was when a guy asked him if he had ever used a Microsoft program before, and he replied "I think I have, once. I was in a hurry to type a letter, and used the computer that was supplied in my hotel room. The program had all sorts of pictures and buttons. It was all very complicated."
Intuitive is often what you are used to. In the beginning, pictures are more intuitive then plain text, so that is what you get used to. I still have trouble using Linux even though I am fairly adapt at DOS. I'm sure if somebody sat down and explained it to me, I could catch on very quickly, just like I can explain to most people HTML in less then an hour.
I'm not really sure where this conversation is going, so I guess I'll end it here.
Too lazy to log in.
Bullshit, and bullshit. I read the entirety of the gimp-devel and new-httpd lists every day. Those lists spend entire days hashing over tiny changes to the interface. Your entire premise is that these projects are coded without attention to user interface, and that shows that either you've never been involved in an open-source project, or you've been invovled in some that work in a totally different way than the ones I've seen.
You miss his point. He did not claim that there was no attention to UI issues, but that there was no feedback loop to end users. There's a difference. MS and other commercial software companies have formal usability testing programs. OS, on the other hand, depends on the feedback from a self-selected set of users. Non-power-users don't tend to volunteer comments for this sort of thing - that's why MS & co. very deliberately seek out average users who wouldn't otherwise be heard, and pay them to participate in usability research. And rather than depending on a user's feedback on his or her subjective impressions, real UI design involves empirical testing - a person's introspection about what parts of a UI they find hard or easy is often flawed.
Where do the installed components go? With RPM and Linux they go all over the f*n' place. Some will staunchly defend the way libraries and binaries and statically linked binaries are split up. Some though call it an administrative headache.
Thank you, Carp, for articulating a lot of good ideas here.
It's human nature to copy the best ideas and tools of others. In the good old days, when iron came along, bronze tools were abandoned quickly. And so it is today. But remember, the model for good UI is not Windows -- it is and always has been the Macintosh. So, open sourcers, roll up your sleeves and study the best ideas of the Mac if you want to put a nice GUI on your program. When you really look at a Mac, you notice that it's not only nice buttons and windows, but really well thought-out dialog messages and other nice details.
One more thing -- the goal of Open Source is not just to conquer the Evil Empire (that will happen in good time just as happens to all empires ...). And the goal of Linux is not just to conquer the Desktop either, although that may happen too. The fact is, Open Source is achieving a whole bunch of goals, and we should clarify what these are.
First of all, Open Source is using that still- revolutionary idea of free speech to erase the mistakes of years of proprietary software development in the corporate world ... 30 or more years of proprietary confusion in the Unix world for starters (How long has Microsoft been around, anyway?). We need to all thank Richard Stallman for changing that. If GNU tools and Linux succeed in providing cheaper, better, and more stable solutions for the server and workstation markets, that already has HUGE implications for the globally networked planet of the 21st century. If the same GNU/Linux engine is a good foundation for desktop solutions, then that's an added bonus.
But to achieve success on all fronts, GNU/Linux developers and distributors might need to remind themselves to which audiences they are addressing their solutions ... and tailor solutions to the right audience.
As it stands now, it is fair to say that most Linux distributions on the store shelves are sort of muddled affairs that are attempting to address too many different audiences at the same time. The problem I see has much more to do with the documentation and installation scripts included with distributions than it does with the software.
Because of this lack of focus on specific audiences, the graphical or menu-driven installers for popular distributions really fall short of the mark for ALL audiences. They are WAY TO HARD to figure out for newbies, and they make a lot of awful assumptions that advanced users can't stand.
I would like to see a Linux distribution whose installation process started with a series of questions about me as the user and my needs, sort of like the way tax preparation software now comes with a series of guided questions for the typical tax "newbie". If my responses to the guided questions showed that I was a total newbie, then the installer would make a bunch of decisions for me -- from partitioning to choice of window manager -- and presto! I would get a really slick desktop installation without worries -- and without all of the unneeded stuff. I am convinced that with proper, well-written documentation, any newbie can learn to use a KDE or GNOME box just as easily as he can a Mac or Windows box. Some really well-thought out documentation could go a long way toward showing a mouse-bound newbie how cool and powerful a Unix command line interface is too -- something they can't get in a Mac or Windows box. The CLI needs to be marketed as one of the really cool features of using a computer, and not something for the newbie to shy away from.
Just like tax software, for more advance users there would be an easy way to break out of the guided installation to let Unix Gurus install and configure everything manually, which is what every advanced user is going to do anyway.
Please don't forget about accessibility in your designs. Windows may be full of multiple (and arcane) ways of doing something, but if you have any eyesight or physical dexterity problems they are the only company (save IBM w/ OS2 but they never did get to their apps) that have made the full spectrum of software (theirs and ISVs)configurable (in one place) to address all the needs. Including simple stuff (but impossible to do without for some) like de-bouncing keys and mice. And, no, command line is not a solution. These (strange) key combo-s enabled serially are a godsend. You don't have to copy their implementation, but don't forget why they are there. Ari
Pinball IS a 3rd party app. Microsoft licensed the game from Maxis and had nothing to do with its development.
Professional UI designers don't work on "Free Software" projects (Linux) because all design applications exist only on "pay software" (windows and mac). Yes, I've heard of "the gimp," but it's STILL not Photoshop. (Freehand, etc etc.) Why would a designer spend time working on UI, gratis, when the end result benefits an OS that doesn't support graphic/UI design to begin with?
So which apps are those that you can only configure or use from within X that are making you suffer ?
Open Source is about sharing code with the rest of humanity that can understand it. In order to topple Goliath the rest of humanity, that doesn't understand the code, needs to be able to use it quickly and easily. One AC to another.
You can't eliminate "learned stuff" completely. Be realistic, here. True, using a mouse is learned, but so is learning how to use a keyboard. (hey, why are the keys lined up QWERTY, again? heh.) The mouse has a much shorter learning curve than a keyboard. Just try explaining command lines to a preschooler. Try the same with the mouse (especially a one button mac job). Which effort do you think will prove more successful?
Linix is often called an "OS" for convenience, but I've seen it, and I don't think that's quite true. Linix is a shareware program that runs on top of DOS, as far as I can see. It extends the DOS command set in some interesting and useful ways, and it also provides some limited multitasking capability and even a crude text-only networking model. For shareware, it's a really sharp piece of work, and I have to hand it to the guys who wrote it. They did a nice job.
If Linix had been released ten years ago, when DOS was the state-of-the-art business operating system, it could have been very successful. Of course, they would have had to put it on a business footing and get rid of the silly "shareware" stigma, but there was a real place in the market for that product.
Unfortunately, Microsoft obviated the need for Linix when they created Windows 3.1. First, they perfected true multitasking, which Linix just can't match -- and who needs an add-on anyway, when it's built into the OS? Second, they rendered the clumsy, slow command-line interface obsolete in one stroke.
It's a shame, but Linix is an excellent product which was obsolete before it was even released. But they do have a chance to redeem it, by adding a decent user interface. There's no way they can get ahead of professional software, because they just don't have the resources, but they can sure keep up if they're willing to work on it. So let's see a GUI built on top of Linix. I think it could work.
Not to be a Mac-bashing jerk or anything but when you read up on Linux for PPC (like perusing Installation Notes and FAQs) you did find out that there are no non-destructive HFS+ partitioning tools to be found anywhere in Mac-Land or outside of it, right? and that either a backup and reinstall of the MacOS would be necessary on a resized partition or an additional disk was in order, right? Partitioning on Mac is just not the cakewalk it is on INtel. I don't know why that is, so I'll refrain from making a snotty remark about it. (Does anyone know if there is a reason why there may not be a non-destructive partitioning tool for Macs?)
As for the Linux filesystem, there is just so much more flexibility, security, bomb-proofing, network-capability built-in there that you don't understand yet--sure it looks totally crazy now but you'll come to understand the whys+wherefores over the course of a couple years of exposure/use. Literally you'll be 2 years into your Linux adventure and still finding out reasons why it would be a penalty to "streamline" the filesystem the way you probably have in mind based on your Mac experience.
.dotfile becoming corrupt. What's probably the most common problem administering mac users? Preferences file corruption? Where else should .dotfiles go ?
~/.dotfiles? What's the problem? I have mine you have yours I can't get at yours, no matter how I long to mess with them, and my launching gimp is never a proximate cause of my realplayer
Why does everyone want linux to be a standard desktop OS? When I first got into it back in 96, it was because I found it interesting. Something I had to tinker with in order to make it do my bidding, full source so I could do real tinkering and screw things up bigtime, the smug feeling you get(got) when you told your coworkers you use linux and they reply with an inquisitive look of awe.
/dev/null and what have you. The desktop is defined as Windows to me, and if you where to get linux up to desktop standards, it wouldn't be fun anymore, just another uninteresting piece of bloatware.
Now we have this SuSE guy stating "Linux isnt ready for the desktop". Did anyone expect it to? Linux is for geeks, your grandma doesnt give a hoot about kernels and compilers,
Tis probably why I prefer tinkering with *BSD these days =P
There is a problem of literacy; because there are the highly literate technologists like us -- the demographic of people that read /. -- and a demographic that is less literate.
In the same way that Comer & Stevens are the de rigour of networking for the high technologists, there is always a niche for the Dummy series as well. Perhaps that is something to consider in terms of UI's, where they are always going to serve different demographics, whether within one 'universal' UI, or multiple UI's.
Perhaps the art is in constructing an UI that has a suitable balance between intuition and power. MSDOS was never intuitive, and Windows was somewhat more closer but had infrastructure problems, whereas the Mac had the power of a windows UI, and a great underlying infrastrucutre, but sometimes the complaint was that 'it wasn't easy to get to the guts of the thing', which is something that Linux allows. Linux should construct a brilliant O/S, and if so, then is even more well placed than the Mac's as a onestop technological masterpiece.
Consider that there is a general trend in computing that's what I will label as 'the feminisation of computing' as an abstract concept to mean that over the past 20 or 30 years, interfaces started out with a direct mapping to the underlying software as an analytical construct, but in the mean time, we have been creating an architectural gap to insert layers between the raw computing substrate, the domain applications, and the interface media to build an intuitive interface on the top.
So, we have been 'reaching' towards the users by providing interfaces that are easier to use, and we have been separating the interfaces ('MFC') from the underlying functionality. However, at the same time, users do become more literate, and learn, so therefore, build up a repertoire of skills to work with the interfaces. Catering entirely for everyone that doesn't know how to use 'the stuff' now, is potentially dangerous, there needs to be a good balance to bring the illerate up to speed, but not alienate them in the process (i.e. what MSDOS has done!), but allow for further growth and experimentation, and allow for use by the high technologists.
Over time, what users want perhaps is divine simplicity, but the ability to reach through the clouds and play god with all the details that sit underneath -- if you'll excuse me for drawing a religous metaphor.
matthew.gream@pobox.com
For data entry apps, that doesn't just mean putting fields in a logical order that you can press enter and jump to the next field, it implies that where appropriate the interface should be laid out the way the person works and then as much as possible anticipate what the user will do next. If that means skipping three normally unused data entry fields based on what was entered, then it should. A killer app in my opinion or a good UI will mean the difference between 4 hours of data entry and 1 hour of data entry accomplishing the exact same task.
Assuming that for a given app no two users would work the same, the application should *learn* from the user. The user painstakingly goes to the fields in the order that he enters them, and presto the next time (or shortly after the next time) the fields start to rearrange into the order of entry. Fields that never get touched naturally move off the bottom, as exceptions are noted rarely used fields move closer to the fields that require them occasionally, but get skipped by default. As time progresses the application should become very close to the way a person works and yet can be different for every person using the application. The same would hold true for GUI operations, if applicable, things that a user normally does together should move closer together (reducing mouse movement) Context menus should reflect what the user ends up doing not predetermined lists.
A good desktop won't lay things out nice and look pretty, but will track the context of what I'm trying to do and anticipate how I can do that better *and* in that context lay things out nice and look pretty. It should also speed the process of finding things, not everything nessasarily goes into a filing cabinet! And if it must the filing cabinet should rearange itself on demand to make my tasks easier. (and not just in one folder but accross the entire system) The interface should be aware of what I'm looking for and place the things and types of things I'm looking for closer too me. (What do you want now?)
The best news is that as UI become less static and more user driven, adoption of other mediums becomes easier and easier. For example speech input since function and interface are forced to be separate and tied dynamically.
This is what he is saying, but it makes me wonder why it is that coders care enough to work on better free software, but ui designers do not. Why don't UI designers join free software projects?
If they don't, and free software takes over anyway, they too will be stuck with bad interfaces.
Free and free is a powerfull combination and if such programs are 'good enough' from a UI standpoint, we know they are going to be better from a technical standpoint, don't we, then they will win out in the market.
BTW. I think a lot of this could happen in the free software world in university research settings. Will some professor and his team achieve world fame for this sort of work?
Bob Clip - friend of A Nony Mouse
Thanks for pointing this out for the fifth time! I can't believe I missed the first four...oh, wait...
You could be imitating a Mac. Good grief, one mouse button?
Snail
If something is truly intuitive, this means that there is no learning process required. The thing in question simply works the way people expect.
I agree with you here, the problem is that we are assuming that the same things 'make sense' to everyone. Actually, we don't quite assume that, we assume that there are two classes of people. Those who write programs and those who don't. We then assume that something that is intuitive for one person a one of the groups is going to be intuitive for every member of that group. Somehow, I really doubt that this is the case. Perhaps this means we can never design truly intuitive interfaces.
(I am free associating here and my mind is getting ahead of me.) If this is the case, perhaps we need to work on reality modelling. We design the underlying reality, and a way for the user to design his own interface as he describes to the system how he interacts with reality.
Now we come to a thorny issue, but one that must be addressed. Does the average person have a valid understanding of reality? Does the average person interact with reality in sensible ways? Do you? Do I? Do we understand reality well enough to model it?
Ramble ends...
Bob Clip - friend of A Nony Mouse
These posts have got to be kidding! Gnome and KDE are excellent desktops! I am hooked on Gnome myself! The Mike K. guy who wrote the article in the first place has no clue to what Linux and Open Source is all about. Open Source is our way of sharing code with the rest of Humanity. That's what the Linux movement is all about. And if we get to topple Goliath over in the process, great! I run 100% Open Source code.
...
It would be nice if the FUCKHEAD editors who run this trash heap actually read their own shit!
This was already posted here over a month ago.
Most linux users are technical people. To them, Linux IS easy to use. Why would a developer spend the time working for FREE on something that's probably just a hobby to dumb down an interface just so joe-user can click his way through some trivial task that most of us would just write a script to do? Most of the real developers don't give a crap whether or not Linux takes over the desktop.
Congratualtions, you have just given us a nutshell description of the OS/2 Workplace Shell. There already exist plenty of good UI concepts, including the WPS and MacOS, out there, but they have all been beaten down by the MS monopoly.
Scott Banwart (Who is posting anonymously from school)
Imitating Windows straight down the line has the benefit of requiring almost no retraining but I think there's some unecessary crap in there as well.
There's only one way to do useability studies and that's to do the F*ing useability studies. Hopefully redhat and va will sponsor some of these on behalf of GNOME.
to give someone just starting is windows
You designers are turning your personal
vices into universal truths
When you go to the toolbar & find 'about'
concept you get licensing data. 'windows'
is filled with mysterious toolbars, arcane
instructions & abusive language disquised as
information 'you have executed an 'illegal'
instruction etc
THE FIRST INSTRUCTION ON AN INTERACTIVE
COMPUTER OUGHT TO BE "IF THIS IS YOUR FIRST
TIME TYPE Y E S" the computer will then launch
a discussion on how to use a mouse etc
everbody else would hit a letter or mouse to
launch a configuiration for internet or
games or whayever. Get rid of scandisk
Get rid of toolbars the damm screen is small enuff
watch out the evil empire can control you
by juicing your various organs with
a MICROWAVE LASER (maser)
can put lumps in your breast too.
ok, I am not going to pretend to be informed. The first question that comes to my mind is, "why don't you get involved and do something about it then?" If this guy thinks that he has useful comments about UI problems with linux, he can help or shut up in my opinion. Unless he is helping to work on a solution for the linux community, then he is just wasting our time. "If you aren't going to be a part of the solution, you are part of the problem" Just my two bits. Nos "don't tell me what is wrong with anything in my life unless you plan on helping to fix it"
you HATE USERS!
that's the bottom line because
STONE COLD SAID SO!
I mean really, if you are to stupid to use the command line, you are not worthy of even OWNING a computer much less use one
So just don't worry about USER INTERFACES, we don't need users!
Go look up "Lokkit" on Freshmeat before spouting off about "the core community" and wizards and such, dumfuck. Debian is full of wizards BTW.
as long as they work.
Of course, Linux people want to overthink it by about 200%. Why? Maybe in order to put the links in the desktop you have to allow something really close to an official standard for parsing desktop link files....HORRORS!!! That's one step away from a standard desktop! That's a standard desktop in principle!
SWEET JESUS, NO! Don't let it happen! There must be one standard for HTML, one standard for DNS, one standard for Kerberos, one standard for SMB, But we cannot allow one standard for desktop Linux not even in principle. That would be too ..easy... for ISVs to code for, and we don't want to invite them to our club.
I also don't want a voice activated computer as I tend to talk to mine especially while writing/debugging code. Some pretty interesting things might occur. For instance:
Me: Come on compile. Oh good! Now let's see if it runs. Oh shit, core dump! F*CK ME!
Cybersex?
You can't get any more direct than:
- of-the-study-thinks-is-an-end-user but the real thing.
End User Coder
The whole idea of someone associated with sendmail criticising Linux/OSS for being isolated from end users is quite absurd.
It's not the fluff that matters but how the product actually works in the hands of REAL users, not artificial-only-exists-in-the-lab-what-the-author
Release early, release often and openly communicate with everyone is far more likely to achieve that than any self proclaimed moses coming down from the mountains.
>Natural evolution - it seems to me that the good ideas will naturally
>rise to the top. With no Borg-like organization steering the
>development to where it is the most profitable, it seems that on the
>whole, this will just naturally get better.
Exactly. And if you read the comments by most of the "Open Source can't produce a real UI crowd" they are scared to death over this issue. They for whatever reason, seem to want to control the direction that UI's for Open Source projects go in. The reason I suspect is that they want to unload their pet theories concering UI interface design where they can have somewhere to point at and say "See? Project blah-blah-blah has adopted so-so's ideas on whatever." Suffice to say I for one don't think that a vast majority of the authors of these artices really have the best interests of the Open Source Development Community in mind when they publish these kinds of articles.
This has already been posted here.
--
"A dessert without cheese is like a beautiful woman who has lost an eye." -- Jean Anthelme Brillat-Savarin
This essay has already been covered on slashdot. See: http://slashdot.org/articles/00/01/28/116240.shtml
It's not always safe to merely yank media out of a drive. Also, for certain performance oriented operations (caching) the OS needs to be aware that there is a different filesystem in the drive or that it needs to be flushed.
MacOS mounts disks. It's just better at making the process seem transparent. Thus, the system is aware of which disk is which and will yell at you if you try to write something out to "Disk FOO" and shove in "Disk BAR" by mistake.
BTW, this awareness of the disks unique identity is something that we would do well to emulate in Linux.
Now once you're completed, tally up your points to see how you've done.
I think I have seen this article before. But it bears repeating. Everyone involved with promoting Linux on the Desktop _needs_ to hear it over and over again until they get it. I would like to see Linux be a real competitor on the desktop because I believe that is the only way we'll break Microsoft's monopoly. But I don't see it happening, because as this article says, we just aren't getting feedback from home users, and what little criticism we hear we simply blow off because too many Linux users prefer a command line and are quite vocal about it.
Worse, we appear to be losing the race as a competetive server. It isn't just Microsoft's paid lackeys who are producing benchmarks showing inferior performance. IBM, who supports Linux, ran their own multithreading tests and Linux fared rather poorly. IBM also offered code which the Linux kernel developers could incorporate under open source, as I understood it, which would improve Linux's performance. I haven't heard what happened to that offer.
I am afraid we are listening too much to ourselves, there are plenty of us to listen to, after all, and not listening enough to the rest of the world on how to improve Linux and it's interface.
Perhaps Red Hat should not be paying programmers but instead should be paying users to use Linux and provide critical feedback.
If you have to read one book as an engineer then read :-
'The Inmates are Running the Asylum, Alan Cooper, ISBN 0-672-31649-8, SAMS Publishing'
In essence engineers are not computer users we're fascinated by minutia. We love learning arcane details, pulling things apart etc.
We code interfaces that closely follow program design. We love to see some sort of script langugae in our products.
Lastly, we do not design UI, they are what we want to see on the front of our code. The only design criteria we ask is that the design is implementable.
Asking other people what they want in a UI is next to pointless. If you ask 100 people are you going to include all their suggestions? Or only the ones you deem applicable.
In short UI design is a seperate and more important task than coding. Most coders, myself included, are too closely linked with the machine to design a usable interface.
If we agree that the M$ way of doing interfaces is wrong and we want to improve upon it then we must sepearte UI design from program design.
In essence it boils down to engineers giving up control. In the open source world I think it will be near impossible for this to happen.
If you use an OO program design system then the UI should already be seperate from the implementation. Pound to a penny it's not though.
The technical demands of seperating interface from implemntation are not difficult it's the cultural problems of getting engineers to take a back seat.
Emmett smacks himself very hard in the head with a brick over and over and over again.
Thanks.
--Emmett
Also, check out #slashdot on irc.openprojects.net
Yep
Teching the masses is not the point and will never be the point. The masses will not learn and any software that is predicated on a painful interface will be opverthrown by software that is pretty and easy to use.
I disagree. People learn new things every day, whether it's how to use a microwave, or drive their car, or operate a new remote.
Computers are comples beasts. There is *no* such thing as an intuitive interface. And the only programs that are easy to use are easy programs.
After a time, people want to move beyond easy. They want to do something different-- they need to split their memo into two columns, or they need to insert a spreadsheet (and there's no such thing as an easy spreadsheet).
Or they want to create a database to store their recipes. Creating a database can be relatively easy, but not many people know how. Why? Because the concept of a database is not easy. (Okay, so the concept is easy. But it's not easy for a layperson to grasp.)
It doesn't have to be hard, that's true, as your example of the XF86Config file demonstrates. But for people who understand how the monitor frequencies work, squeezing out higher resolutions at better refresh rates is easy.
Too many interfaces that are "easy" are also restrictive. That is how they get their ease-of-use. It's like the Holly-hop drive-- two buttons, labelled "Start" and "Stop." The "Start" button starts the Holly-hop drive. You can figure the rest out from there.
Sure, it's easy to use. But it's hard to do anything *interesting*.
More important than "ease-of-use" is consistency. An interface must be consistent, unless there is a damn good reason to break consistency.
Anyway, this whole "Linux is not easy to use" thread is getting old. I have a lot of friends who are non-geeks, who visit for long periods. Given a choice between Linux or MS-Windows, most of them choose Linux. Why? "It's easier."
Microsoft is to software what Budweiser is to beer.
How can the open source development model obtain the necessary user feedback to development interfaces that the user will intuitively able to use?"
The only intuitive interface is the nipple. Everything after that is learned.
The problem is not making an intuitive interface it's getting over the magical force that takes an otherwise intelligent human being and turns them into a total moron the second they see a computer.
Part of the solution to this problem is good documentation. You don't need a gui that your grandmother can use. You need a manual your grandmother can understand.
Unix is no harder to use than Windows, it does need a GUI, there is no question about that but even GUIs need to be learned. Has anyone seen the manuals that come with macintoshes? They explain in great detail how to achieve things, using pictures and examples that even a moron could understand but helpdesks still get calls from macintosh users wanting to know how to create a new folder.
Silver
Linux is, first and foremost, a hobbyists system. Then, it is a server-side OS. Then, and only then, it is an end-user workstation OS that looks pretty and holds your hand, and comes preconfigured out of the box. In fact, out of the box is exactly where this end-user convenience should come from. Let the people making money on Linux distros add that value. They're the ones who depend on a growing user base. "Hey you, in the red hat! Are you listening?"
Finally, someone in this thread who "gets it". As much as I enjoy the little niceties of easier installs, admin tools that (almost) don't suck, and other improvements to most distros out there, Linux has been (and always will be) about "software by the people, for the people". DIY is what it's all about. If Linux is supposed to "take on Microsoft for dominance in the desktop wars", then the people who have a vested interest in making Linux "point and drool" friendly need to cough up the dough to do so. Companies like Redhat, SuSE, and Corel are already putting a lot of money into the GNOME/KDE camps, and this is already paying off for them, but they need to go further. Either way, Linux developers will stay with "their baby" no matter what Infoworld or ZDNet say about it, even if it were to drop off of the "desktop playing field".
MS, ZDNet, and the others want appliances.
I want a power tool.
If only "common" sense was actually that common...
:) Yep, I knew all the keyboard commands by heart. Of course, this was 6.0 and later. (5.5 wasn't as cool, but I still knew the commands--the key bindings were awkward, but standard at the time, ^K^B, ^K^K, ^K^V, etc., etc.)
:)
Not only is it unintuitive, it's darn slow! Who wants to be moving that thing around? Regardless, they supported that just fine, too, and the mouse worked well with it, if it was used. But all the keyboard commands were helpfully in the menus, and the help system was great. (much better than the standard Emacs-like Info browser, IMO...)
Yeah, most Open Source projects are far more customizable than their Windows counterparts, (because you can change it, silly!) and there's usually an (undocumented) substitute for unavailable keys. (or you can bind a key to it--I'm tempted to bind my "Windows" keys someday, but I don't need anything extra, really.
---
pb Reply or e-mail; don't vaguely moderate.
pb Reply or e-mail; don't vaguely moderate.
I'll be happy when the first thing someone says about Linux *isn't* about
* Learning vi
* Learning Emacs
* "...arcane commmand line..."
* "...text files..."
My reply to this is as follows:
* GNOME Apps
* KDE Apps
* LinuxConf
* RTFM!!!!
I think if Linux came with a "Getting Started" guide that just *touched on* the wealth of tools that are available, it'd be much easier. In Windows, it doesn't come with much. Your average Linux distro comes with a WHOLE LOT of tools, and users need to realize that.
The interface design is fine. It's just as good/bad as anything else. The last known good interface I saw was the text IDE for "Turbo Pascal 7", which is why I like RHIDE. The last "innovative" one I saw was WinAmp, (and everyone has already copied that) but that doesn't necessarily mean it's easier to use.
Widgets on Linux, Windows, whatever haven't changed much for a while. (but I'd kill to be able to resize more windows in Windows, they can't design the control panel worth a damn!)
---
pb Reply or e-mail; don't vaguely moderate.
pb Reply or e-mail; don't vaguely moderate.
don't forget ctrl-shift-esc for the task list...
My threshold is set to 2 and I just did a quick and dirty word count of the postings on this subject so far.
The word intuitive was used 60 times.
The word consistent was used twice.
My sense is that most people here are saying that an intuitive interface is difficult or hard to create. (Who was it that said "The only intuitive interface is the nipple"?) But I think that most people agree that consistent interfaces are easier to learn -- that is, once you've learned your CLI, emacs, or GUI interface, something that's a lot like it is easier to use.
I don't think the first post was claiming that sendmail's config difficulty is necessarily a bad thing, merely that he finds it amusing to see an article on user-friendliness on the site that gave us that config system...
Regards,
Denny
# Using Linux in the UK? Check out Linux UK
Police State UK - news and
Aha, just thought of something.... the wastebasket should be replaced by a fire. Even the tribesman from New Guinea has probably seen fire and has the idea it can be used to destroy things you don't want anymore. Besides it's a good excuse for some cool animation.
:-)
Now I think I'm going to go patent the idea. Na-na-na-na-na-na-nah!
There should be no files, only objects, contained in a global memory space which spans transparently across all the computers in the world. In this model a hyperlink is the same thing as a capability (as defined for capability systems such as Eros). All objects in memory are automatically persistent unless they are marked transient; so power outages will not cause lost data, yet you never need to know there is such a thing as a hard disk, nor do you need to organize it. The structures of organization can be completely arbitrary and user-defined, not limited to trees or tables.
Heh. When I think of how to design interfaces, I'm always aming at something that I think my mom could use. Largely because while my mom uses a computer (at least, you can call it that ;) my dad doesn't at all. He did send me an email once. Complained that he couldn't figure out how to make capital letters. honest.
-- This and all my posts are in the public domain. I am a lawyer. I am not your lawyer, and this is not legal advice.
What I want Linux to do is have /dev/sda be the damn drive like everybody expects! I should be able to "ls /dev/sda" and see what is on that drive, immediately, no matter what file system is in there, and give me an acceptable "no disk in drive" error if there is no disk, or "unreadable disk" if the file system is not recognized. Now I know internally that Linux will have to install the correct file system module, and run a test to find out which one, etc, but this can obvioiusly all be done automatically.
"But how do I make that disk be /usr?" I'm sure people are asking. Simple: "ln -s /dev/sda1 /usr" That isn't too hard!
Also if the hardware has a button that gets the disk out, the driver should sync the disk enough that it will be ejected with a good file system. If it is noticably hard to get the disk out, the driver can lock it in until the disk is synched. If it is impossible without turning off the computer the driver does not have to do anything.
This really ought to be done. There is no reason for the internal housekeeping of Linux to be visible outside the kernel.
Devfs can be used, and in fact i hope it would be used! The link points from /usr to the file in /dev. Get it?
Your examples:
mount -t nfs foo.umbc.edu:/usr/local /usr/local
We need a network-transparent file naming scheme. Then this is done by:
ln -s /net/foo.umbc.edu/usr/local /usr/local
smbmount //bar.umbc.edu/public /samba -o username autechre
The username could be done with ioctl calls to /samba (after the link is made) or with a naming scheme like this:
ln -s /net/bar.umbc.edu/public/username.autechre /samba
My wife uses linux. I stay ssh-ed to the machine all day. When she has a problem, she starts a 'talk' session, tells me what she needs, and I fix and/or setup/install what she wants. One day she asked for a word processor... I downloaded, installed, and started AbiWord, all from my desk at work. Now THAT has got to be the easiest user interface EVER.
Linux needs something completely new, which NT simply cannot deliver - "Hire-a-remote-administrator"!
I visualise an installation which automagically calls the distributor, starts a chat session, and the user starts paying for support by the hour.
Once more unto the breach, dear friends, once more, Or close the wall up with our American dead!
I'm thinking of changing my sig to
Proud to be a member of a "self-selecting group of loudmouth power users".
I see even classic Slashdot is now pretty much unusable on dial up anymore.
The problem that is cropping up here (as I see it) is that people are expecting far too much in the way of assistance from the interface. What they are trying to achieve is considerably more complex than tying one's own shoes, and yet the user time and time again will complain that an interface is broken simply because they can't figure it out (although they have no problems tying their shoes, most of the time).
Zero-knowledge interfaces are a myth for tasks above a given level of complexity. No interface, no matter how complex and "intuitive", is going to let a common user do a stress and wear analysis on a four-lane split level suspension bridge that will keep cars from spilling over into the drink after a few months use.
The thing that really needs to be addressed, is the required minimum knowledge level of the user, because without it, they're going to fail even if the interface is another user who does know what they're doing.
A computer is a marvelously complex device, typically containing more tiny switches inside than there are stars visible in the night sky. How the HELL people expect to be able to use one when they refuse to even attempt to look for documentation is beyond me.
(Want another example? Ever watched someone desperately struggle to figure out how to configure their software to use a proxy server, when they don't even know what a CIDR block is, or better yet, that their own desktop machine has an IP address? I have a particular person in mind from this example--don't say it can never happen.)
If you had a fscking clue, you'd understand that having the scroll bar on the LEFT side would be a helluva lot more useable. After all, that's usually where you're looking as you scroll to find something. It's where paragraphs start. It's where the index is. It only makes sense to put the scrollbar near where you usually look. Look up "Fitt's Law."
And, again, if you had half a fscking clue, you'd realise that if your pinheaded physical-world analogy were to hold true, moving the scrollbar down would have to move the CONTENT down. "O look, you move paper down by pulling it toward yourself."
In fact, if you had *any* clue, you'd realize that if modeling computer GUIs on real-world interactions was a good thing, then Apple's horrendous "let's pretend it's really a Bang and Olufsen!" QuickTime interface wouldn't have been roundly denigrated by the HCI community.
Just because your lukewarm mind can make absurd connections between real-world and GUI-world interface elements DOES NOT mean that there is a connection, nor that there should be one.
Twit.
--
--
Don't like it? Respond with words, not karma.
There are some 21 ways to leave your application...
Just hit the ALT-F4, jack,
Try a ALT-F-X, stan,
and then try ALT-F-C (in Explorer), ALT-F-Q, clicking the boxed-X, clicking the application icon in the top left and pressing ALT-C, clicking the menubar>File>Exit, menubar>File>Close (in Explorer), menubar>File>Quit (in some real idiotic programs), and let's not forget CTRL-ALT-DEL, selecting the program and pressing (or clicking!)
ALT-E/End Task!
Well, you get the idea. Once you go through all the possibilities of mouseclicks and keypresses, there are over twenty ways to close an application.
It's horrendous.
--
--
Don't like it? Respond with words, not karma.
Oh, what bullshit. There's no paradigm linkage between the two. In fact, I'll wager most people flip a page by its *EDGE*, not the corner.
It's one of the few examples you pulled out of the air, based on silly thinking. Care to explain, next, why the maximize button is so close to the close button that it's a three-pixel slipup to hit the wrong button? Care to explain, then, why the menubar is not set to the top edge of the screen, where it's impossible to overshoot it when mousing?
Windows does not have a particularly good UI design. It is, at best, an attempt to do a GUI by doing the opposite to what Macintosh did, to avoid losing the look-and-feel lawsuits.
--
--
Don't like it? Respond with words, not karma.
Windows is not intuitive.
Amen. All I meant was that it has become intuitive, in the same way that, say, English is to English-speakers. But to somebody learning it whose native language is, say, Chinese, it is verging on nonsense. There are virtually no standards. But it's intuitive to you and I.
But I absolutely agree with you that Windows is a hodge-podge of questionable or barely-existent UIs. But, unfortunately, people think that Windows equals computers. And that will be hard to get beyond.
-Waldo
Perfection is an admirable goal, but let's reach the point where such mistakes are eliminated first.
"I love my job, but I hate talking to people like you" (Freddie Mercury)
You're confusing "familiar" with "consistent and understandable". Why are there more PalmOS PDAs sold than Windows CE machines? It's a good thing more people aren't able to learn new things, or there would be millions of Palms out there. *shrug*
Other people will likely present examples of Windows being decidedly non-intuitive ("To shutdown, click Start" and "To save, click on a backwards drawing of a floppy") -- but that's not the point.
The point (as I have argued before) is that copying an existing user interface somewhere it doesn't belong is a worse idea than designing a flawed interface from scratch. Putting a Start button in Linux is like putting a crank on the front of your car.
--
how to invest, a novice's guide
If I was a sendmail fan, I would tell you that "you're not supposed to edit the .cf file!" ... but I'm not; I prefer qmail.
Perhaps not necessary for floppy drives, but not something that's totally unnecessary. (and I've sure gotten plenty of milage out of being able to mount another machine's /usr on /usr when I accidentally deleted everything in /usr, and I can see doing the same with a CD-ROM.
The article in question really left a bad taste in my mouth. On the one hand Open Source is dinked because it doesn't meet the author's idea of "innovative" and simply copies other software. On the other hand OSS is dinked because it doesn't work the way the masses expect it to. On your alien third hand he doesn't seem to appreciate that in most cases the author *IS* the user.
When I think of an "innovative" user interface, I think of Kai's Power *. After hearing the reports of how innovative they were, I decided to try some of them to see what I could learn. I literally had to read the manual to figure them out, they were definitely not intuitive.
Imagine if every Linux app had an "innovative" GUI. Everyone trying to out innovate everyone else. I think it would be a mess. Users don't want innovation, they want consistancy.
Consistency means that a class of applications all work alike. However, it seems the author thinks that's unc00l. However, until Open Source software projects start getting the resources to run Human Factors testing on their interfaces, the best we can do is model them after other popular interfaces. What's wrong with doing that?
Unlike most large software companies that employ developers, most developers on free OSS projects are there because they *USE* the results. Why is so much importance placed on the users who don't contribute anything to the projects, instead of the ones who do?
Note that I'm not saying that Aunt Mildred needs to get in and do some coding. That's the last thing the package needs. These users *CAN* contribute by doing usability surveys of their peers and coming up with a user interface that will actually be useful and usable.
I find it hard to blame the developers that these users aren't willing to contribute to the project. I'm not talking about spending 10 minutes whining about the problems a particular user has with an interface the developer took hours to create. A real contribution that *NON-TECHNICAL* users can provide is to spend some hours contacting other users and getting a concensus for an interface design. We're still talking probably a fraction of the hours a developer has put into the project.
So, why aren't users doing this? From a purely selfish standpoint, it directly benefits them. Until users are willing to stop bitching about how developers don't recognise them, and take the future in their own hands by doing something worthy of recognition, they will be left out of this world. And rightly so.
If you're not going to pedal, either stop complaining or get off the bike.
Using the mouse is something this user would have to learn, even if he starts using Linux.
But you're right about the doubleclicking. Lucky us, win98 has that html look and feel for the desktop. A single click replaces a doubleclick (like in KDE).
What's so bad with using the w9* interface? 90 % of the desktops runs windows software. If you want users to switch to linux, better make that transition very easy, ie don't change the whole user interface. This way users have two very big pro's for using linux: a familiar user interface, and a very long uptime.
Maybe MS is developing a desktop, which can be used on linux, instead of kde, gnome etc.
Who the hell moderated this down? Jesus, this post is funny, insightful, and even contains a smattering of pathos.
It's also on-topic, relevant, and every other thing a good post should be. His point is, in fact, the very first thing that went through my mind when I saw the article. Sendmail is a very non-user (and admin) friendly program. Get it? See? It's FUNNY, idiot.
If you don't understand a post, for the love of god, don't moderate it. Just go turn your little 'I want to moderate' button _off_ and leave those of us who post with some depth of understanding alone.
Danke,
--
blue
i browse at -1 because they're funnier than you are.
Well, we were on the subject of keyboard navigation... Yes, just about every windows app lets you click the X, but how about standard shortcuts for regular tasks... for example, cut and paste is generally ctrl-x and ctrl-v, except for quicken, where it's ctrl-instert and i forgot the other one...
It wouldn't be so bad if all those different variations of quitting applications worked in each application, but instead each program uses a subset of those.
One nice thing about windows compared to the Mac is that you can ALMOST do everything from the keyboard if you desire... But there's way too much inconsistency across applications to make it completely feasible.
I do know what I'm talking about, I'm sorry to say... Most people are taught to look in the file menu to quit their programs... So when they go there, that's the keyboard command that they're most likely to learn.
Additionally, Pinball may be a unique example, but the issue with that is that it's a program that Microsoft developed and ships with the OS... It'd be more understandable and acceptable if it were a 3rd party app, but it's not... It's technically "part of the OS" as microsoft would say. Yet it introduces a non-standard way of doing things...
Also, unlike Apple, Microsoft does very little to enforce interface rigidity upon develoeprs... That means that in most instances, ctrl-x cuts and ctrl-v pastes, EXCEPT IN QUICKEN, where for some reason it's ctrl-insert to paste and i forgot what it is to cut... It's just lame.
I think that whatever the next step is in opensource UI's, a great idea would be that yes everyone can develop for it and make up whatever tricks they want. But there should be some guidelines in place, and maybe even a certification program so that people will know which applications actually abide by all the recomendations.
I think people need to get over the idea that a command line makes them more productive, enables them to do more, etc, take a step back and LOOk at the current offerings...
GUI's are cool because they give you a Graphical view of your file system. And they give you keyboard shortcuts that represent things that you never could with a CLI... Like "Select All" when you're in a folder view.
On a mac you type "command a, command delete" to rid a directory of files. In Linux, you type "rm *"... which is 4 distinct key strokes rather than 2.
In windows, to switch between tasks, you type command tab. That's all.
Again, before anyone goes out and tries to create the next "killer" interface, take the time to thoroughly learn the existing ones, otherwise, my bets that there'll be a pretty much USELESS interface created. Watch people using their computers. Windows users. Mac users. See how they do some things the same and others differently. See what baffles them. Try to think up better ways for them to do things... Or put their options where they look the first time... Again, otherwise it'll be an utter waste of everyone's time.
That was exactly my point AC. If the need for ease of use was felt by enough developers, either directly or through requests, then the features would be widely available.
If Debian has these features, and they're not widely publicised, then the Debian people are dropping the ball. Ease of use is a great way to get new/un-expert users. I'm surprised that Debian isn't blowing it's horn loudly on this.
Next time you call someone 'dumbfuck', have the spine to sign your name.
-- What you do today will cost you a day of your life.
In the majority of all Open Source projects, the intended end user is the developer who initiated the project.
Most Open Source projects are the result of a developer's need for a program that doesn't exist yet, or for which there aren't adequate implementations for the particular need.
Open Source projects are unlikely to gain developers who don't already understand the project and its goals. Once a project is well underway, making it intuitive can often mean re-designing it from the ground up.
Someone who is interested in how intuitive a project is for the baseline end user must be involved in it from the start.
People who sell software are quite interested in how well the average user can use the program, but the incentive for OSS developers just isn't there yet.
The focus has to shift from "what do I need," to "what does the average user need." To do that requires putting a LOT more effort into the project, possibly removing it from the "do it in your spare time" domain, or else applying a little of the Mongolian Hoard Principle.
Essentially we are now developing commercial software under the OSS paradigm, and it requires commercial development disciplines.
Everyone seems to want open source to take off right now. This is a nice thing to want but we end up with things like the Gnome 1.0 disaster. I love Gnome and use it as my primary desktop, but 1.0 was not stable. I hope Open Source learned a little something from this. "User friendly" will come with time, but if you try to rush it it will flop on its first try. As user base grows we will need good bug/feature request tracking software such as bugzilla to be used more often. I am not fully aware of all of bugzilla's capabilities but as I understand it it should be able to take bug reports and catagorize and prioritize them as needed. As projects like this get better people will be able to submit bug reports on user interface, program operation, feature problems, compile problems etc...
All in all, people need to have a little more patents and these things will come "When they are ready"(tm)
Paying taxes to buy civilization is like paying a hooker to buy love.
For the most there is no innovation in the look-feel (some outstanding exceptions such as E)
/.
This may be a good thing in part since that makes the learning curve of OSS use all but vanish.
We could however leverage our quick turn around by creating a forum for user feed back (and new feature / visuals) so that OSS will quicker adapt to what the user needs / wants.
This may be a new business venture for our friends at
It seems to me that the CLI is the closest thing to the "Star Trek" voice activated interface you describe. Where else could you tell your computer to 'Update all programs, mail me the results" or one I used recently "find /usr -name *.h | xargs grep foofunction". You could even X-ify it by putting a "| xmessage -file -" at the end, or do it by voice by cutting the first column (before the colon) and dumping to a text-to-speach package.
A CLI is a full-on object-oriented interface that would be difficult to impossible to represent graphically, but one is welcome to try
-- Remember: Wherever you go, there you are!
I've seen this a lot. Many systems have 20-50 icons on the desktop, that is how we ship them now. Utterly confusing and worthless, I can never find anything, because they are always sorted differently on every machine, very cluttered.
Also most users refuse to run at resolutions greater than 800x600, some 640x480. And the new machines have Matrox G400 video with like 32MB RAM (and some are shipping with dualhead !!). What a waste.
-- Remember: Wherever you go, there you are!
I think this is a very good idea. I would like to see service contracts, including remote admining of a system, included with system purchases.
Only major problem is security, how much do you trust this person or company to root around on your computer. Would the average user be able to find out if the admin copied the answers to their purity test, would you be able to tell?
-- Remember: Wherever you go, there you are!
Most non-technical end users can't give good feedback on precicely why a UI sucks...they just know that it sucks and it frustrates them.
One of the really smart things that Apple did was to create UI guidelines ("Preferences" are ALWAYS in the Edit menu) and lean on developers to use them. Then you do not have to relearn where to find common features in applications; as you use the apps, it becomes intuitive how to perform those actions since they are always done the same way.
Developers seem to have a great time creating new interfaces, but they should instead consider :
Is this so much better than everything else out there, that it is worth making the end users relearn everything?
In most cases the answer is NO.
The Linux developer community needs to realize that to make software usable, it is generally more important to be consistant that clever. As Nigel Tufnel once said "There is a fine line between clever and stupid"
I wish that all commercial software developers who develop end user software were required to read "The Design of Everyday Things" by Don Norman. The quality of UI would improve markedly, IMHO.
Thank you for bringing my comments back into perspective. I was emotional and late for a meeting when I wrote them (as evidenced by the great spelling errors you quoted).
People do learn new things everyday, and If you want to take advantage of the power of new computers and new programs people are going to have to continue to learn new things. The real issue is "How can we make it easier for users to learn?".
Intuitive vs. ease of use is very hard to clarify. Intuition is very real, it is based off of past experiences. The breadth of experiences that somebody is able to bridge is a measure of intuition.
Programs that do complex things do not have to be hard to use. what is important is that somebody that knows what they want to do and has some familiarity with the environment (UI, OS) is able to do so with out having to dig through incomplete readme files and certainly witout having to dig through the source.
I very much want something that has "easy" and "complex" interface modes. The next post down talks about "basic" and "power user" interfaces to their scanner UI. Excellent and very much what I am after.
consistence and intuitiveness are very much interwoven. If a program is consistent with the overall UI and consistent with similar programs (for better and sometimes worse) than it's intuitiveness is greatly increased. If a user can sit down and go "In power program for NT I do this to acomplish my task, now in super program for Linux I am am going to try this similar sounding/looking thing."
Maybe we are (or probably me) lacking the language to really communicate the core issue. That is why I feel that we have continued to see so many of these articles over the years. People keep trying to find the verbalization and perspective that make the answer more obvious and understandable. The problem is that we are still grapling with the underlying question and not even getting to the answer.
chris
p.s. I hope XF86Config never goes away, I am just glad that there are good tools out there that make it possible to not have to hand edit XF86Config just to get a functioning X. I still have bruises on my head from 6 or 7 years ago when I first set up XFree86.
-- I need more coffee. It's Monday. There is no such thing as enough coffee on a Monday.
Lusers will always be lusers.
Anyone who has had the misfortune of having to explain technical issues to the general public knows this. No matter how you make a system look, they still won't get it.In other words, wipe those frowns away! Don't worry your pretty little heads about it. It'll be okay, really. Now, toddle off now and leave me to my command line.
The range of Linux users is expanding, and many people would like to expand it further. I doubt your statement that Linux is well designed for its current user base (as opposed to the one two years ago) is even close to accurate lately.
Further, even if I'm an open source hacker working on package X, that doesn't imply that I have the faintest care or concern for the internals of package Y, which I'm just trying to use! Even hardcore open source hackers still use large parts of Linux as a tool, as consumers of the technology.
>Am I the only one tired by titles of 'It's [whatever], stupid'? This always sounds to me like a somewhat veiled attempt from the author at placing themselves above the reader from the get-go, by claiming they have such a clever bit of information that the reader should feel 'stupid' for not knowing it.
:-)
:-)
I think you're reading WAY too much into the titles of articles. I mean, it's almost like you're trying to be (gasp) politically correct...
Just kidding there, but generally the "It's something_here, stupid" type of articles are mainly trying to point out the obvious, not the insightful or clever. That is, if you didn't already see this, you must be stupid.
And if I can't look down at other, less-knowledgable people with disdain, then what's the point of being a computer guru?
---
- Give a man a fire and he's warm for a day, but set him on fire and he's warm for the rest of his life.
errrrr my bad. I didn't fix permissions when I updated moved from the development stuff over to the main one.
"Fighting the underpants gnomes since 1998!" "Bruce Schneier knows the state of schroedinger's cat"
i'm actually working on this right now. I'll be getting more stuff done this weekend and should be ready to go live sometime next week. Check out OpenInterface.
Basically what I'm doing is creating a repository where artists can submit examples of icons/interfaces for developers to see and contact them about thier project. I've decided to license everything under the Opencontent License (opencontent.org) and I will encourage artists who submit to follow suit. I'm hoping this will aleviate some of the burden on coders and the can focus on doing what they do best...hacking at code.
"Fighting the underpants gnomes since 1998!" "Bruce Schneier knows the state of schroedinger's cat"
uggggggg....i keep forgetting that extrans is borked now. Here is the clickable for the link impaired.
"Fighting the underpants gnomes since 1998!" "Bruce Schneier knows the state of schroedinger's cat"
The main problem, as stated before, is that what people think of as intuitive is generally biased by their experience or knowledge. And this reminds me of what someone said on OCLUG: "There is no such thing as an intuitive interface design, the only intuitive interface is the nipple, everthing afterwards is learned."
(I forget the source...some Ottawa Hacker)
cheers,
Frank
A truth that's told with bad intent, Beats all the lies you can invent. -- William Blake
Those looking for wisdom in the battle for better interfaces should take a look at Zope's management interface. I'm not saying that the interface is very good right now (some would say it's horrendous), but the approach taken by Digital Creations is very good.
Zope is a collection of objects and each object provides its own management user interface. Because everything is browser-based, interfaces to new classes of objects are relatively simple to create.
Zope's roles and permissions features are not easy to master. They are complex, and the user interface doesn't make them much easier to understand. But the user doesn't have to learn a set of keywords and read a HOWTO to set up the permissions. All of the instructions are onscreen.
Thus Zope does not try to present everything in such a way that the user instantly knows what to do in every situation, but it does reduce the need for a HOWTO or a help file. That is the goal we need to achieve: require no external references.
I know many of you are adept at configuring Apache, but I would bet there are very few of you who don't need to look at the reference manual each time you change httpd.conf with a text editor. (Yes, the embedded comments help, but they don't solve the problem.)
Both the Mac GUI and Zope (and to some extent Windows) have had a small advantage in the GUI department: developers are forced to provide a GUI interface for everything the program might do. Those who are working on improving the Linux GUI might try something drastic: rm -f /usr/X11R6/bin/xterm
Zope falls under the same category. Almost nothing in Zope is configured via text files. Every object has to provide its own management interface, or the user can't use it at all. It is a sound design philosophy.
Notably, the SWAT tool for Samba provides a very effective interface as well. Samba includes some extremely strange options. The interface hides the advanced options by default, but they are easy to turn on. Every option includes a link to its description in a help file. SWAT eliminates the need to break out a reference manual while providing access to every odd feature of Samba (AFAIK). So SWAT is another example of a tool that simplifies the interface without removing any advanced features.
It's not borked! That is what extrans is SUPPOSED to be. It was borked before.
Already posted. See this article.
"...Is this world not a call I can screen out" --
Actually, this was done about 10 years ago. It was written by IBM, and was the basis for the OS/2 Presentation Manager interface and the Windows 3.0/3.1 interface. It is called CUA '91, standing for Common User Access, 1991 (I think the name is correct). It defined a consistent interface, which unfortunately is largely ignored these days.
-- Error: Cannot find file REALITY.SYS - Universe halted, please reboot!
A good starting point for something like this is Apple's Macintosh Human Interface Guidelines. I'm talking about the general contepts, not the niggly specifics.
One thing that Bruce Tognazzini has mentioned many times is that volunteer user feedback is not enough or accurate. People don't actually know how they want to use an application, and frequently they are wrong about which parts of the interface are effecient. Actual user testing (put user in room with application, watch & time them) still needs to be done. It would be great if an organization could be set up to support that.
I have always purchased RH support (not the real unlimited thing, but the add on to the regular box set) and have struggled to find, over the years, *anything* that is covered in that $2-300 cost. The good news is, I rarely need it, the bad news is it's never helped me. I am sure that the $1-60K contracts would have provided complete service, but I couldn't afford it for my company. Having said that, all the commercial software support contracts I have bought have been COMPLETELY WORTHLESS. That's why I gave up on commercial software!
This argument is getting old. Microsoft has done a great marketing job of convincing people that they should be able to press a few buttons and wham! an instant running mail server, web server, db server etc. Anyone who's configured exchange, IIS, or SQL server will tell you this is just nonsense.... so why is the open source community still full of folks who think this is the goose that lays the golden eggs? I don't care that it's tricky to set up Apache, or sendmail. I don't give a flying fart that it's tricky. As long as it's easy to run a browser and some office apps the UI for Linux is just fine. Wait a minute, it *is* easy to run a browser and some office apps. Doh!
When I was doing a large scale rollout of Win95, we played M$ videos to our users. M$ cited the months of usability studies and 1000's of man hours that went into the "revolutionary" new interface. This was re-assuring. What wasn't reassuring was to see all those interface methods drop like flies with office 97, to ie4, to now ie5, office 2k. If the research was legit, the interface would have stayed. My users now all complain about the stupid new level of complexity and non-intuitiveness they have to deal with. What is Apple doing now with it's interface? The concept of a one size fits all GUI is just pure nonsense and needs to be resigned to the garbage can (icon).
Being aware of the uniqueness of disks may not be optimal. We'll have a whole new crop of problems if some silly person ejects a disk while it's still mounted, walks over to another computer and writes to the disk, and then brings it back to the first computer to read it. Macs are notorious at refusing disks that have been treated this way.
P.S. Yes, I know. Hit Command-. several times to make the dialogue box go away.
== I am not Me.
"The only intuitive user interface is the nipple"
I'm a little disappointed that we are still using the word "intuitive" for user interfaces. There's no such thing. Making a computer act like a desk (or a file cabinet, or whatever) doesn't give me more power for the amount of effort expended--it just makes it more familiar. And if the cost of that familiarity is decreased functionality, then the user actually loses out.
People need to be aware that computers are POWERFUL. You can do a LOT of work very quickly and accurately when you have access to a computer AND you know how to use it. What other powerful devices does the "everyday Joe" use? Cars, power tools, phone, etc. How many of these devices have "intuitive" user interfaces? None. How many of them have "metaphor-based" UIs? None. In fact, cars are so hard to use we don't let anyone under 16 operate one and we require a license to do so.
I'm not blaming the user. I'm saying that, instead of investing thousands of man-hours (not to mention dollars) in shortening the learning curve, why not spend the same amount of time just helping people over it?
,br> A learning curve is a one-time event (per person). Once you know a UI (however arcance) you know it. It doesn't make sense (from an engineering or accounting perspective) to optimize a device for the first use, especially if that optimization degrades performance for the rest of the life of the product. And yet, Windows and Macs (and a lot of Linux "UI of the Future" projects) do that very thing. It's easy to jump in and learn, but once you master the little morsels of functionality you've been given you hit a dead end. I'd much rather flounder for a few hours/days/weeks but end up being able to exploit the full power of that marvelous invention: the computer.
Granted, Linux could be made easier and still retain functionality. But the efforts I see in place (KDE and GNOME in particular) aren't doing that. Both of those projects are constructing GUI versions of all the CLI tools. Great, except that the GUI tools can't be piped together, etc. "gless" is fundamentally less powerful than "less". And the GUI and CLI tools are so different that if the user wanted to go to the CLI (assuming they knew it existed) they'd have to relearn everything anyway.
No, the right way to do this is to first layout the level of power/flexibility the current UI has. Then examine the weaknesses (hidden options, incompatible parameter usage, etc). Then propose solutions to the weaknesses keeping in mind that the power/flexibility cannot decrease for even the most advanced user.
I could rant a little more about how some peoples idea of a good interface is one that has transparent xterms, but I'm getting all worked up and I need to stop.
--
Here is the result of your Slashdot Purity Test.
Linux MAPI Server!
http://www.openone.com/software/MailOne/
(Exchange Migration HOWTO coming soon)
Intuition is acquired. That's why people feel that M$ Windows is intuitive -- it's been the primary OS for PCs for quite some time now -- everyone's familiar with it.
AAARRRHGHGHGHGH!!! Intuition != Familiarity. A baby has never seen a nipple, yet knows what to do: intuition. I know how to change my video driver in windows because I've done it millions of times: familiarity. These two things are not the same.
And my objection is more than the pedantic "You're using that word wrong" (although that's fun too). You are saying that Windows is easy to use because people are used to it. So then logically, the best UI for Linux is not GNOME or KDE but time. That's exactly what I'M saying to: training training training is the key key key.
Some of the things that computer do are fundamentally different than everything else humans have ever dealt with. So why would asking computers to do these things be any different? (most) Humans have no intuition for dealing with computers--what is needed is assistance.
--
Here is the result of your Slashdot Purity Test.
Linux MAPI Server!
http://www.openone.com/software/MailOne/
(Exchange Migration HOWTO coming soon)
Although they have been getting a lot of press in the last couple of decades, for most people (e.g. none-Slashdot readers) computers are a tool and not an end in themselves. Anyone who has ever helped a coworker knows this: "Why won't it show me the summary?" "How do I get it to send my email" etc. Part of any so-called software engineering project should include a user needs analysis e.g. who are you designing for, why will they use it, what do they use it for, and so on. As I see it, most Open Source projects are designed for programmers by programmers. This is fine if that's the target audience, but as previous posts have said, Microsoft got to number one (among other ways!) by understanding their user. For my money, Office is an extremely useful program that allows me to complete my tasks quickly and efficiently without a lot of worry about how it works. This is what people are looking for, and frankly, I don't see it in most Linux apps.
Mounting as a process may well make sense, but I don't think you can really say that the UNIX way of still mounting disks manually is particularly clever.
OK, I'm a throwback who still likes Amigas, but they've got stuff they can teach you. In this case, floppy disks that don't have weird, multi-fork formats like the Mac but have a proper icon and launching system (well, pretty good - not ideal though) and where the system notices when they go in and out of the drive.
The Windows approach is pretty silly - it REALLY should be able to notice a disk change by now - but the UNIX approach could do with some automation for user-friendliness.
Greg
Greg
(Inside a nuclear plant)
Aaaarrrggh! Run! The canary has mutated!
Like a wise person (whose name I have forgotten) once said somewhere, "The only intuitive interface is the nipple. Everything else is learned."
Seriously, the problem that wind'ohs users who decide to install linux (or any other "alternative" OS, for that matter) have is that they expect it to be exactly like the (piece of shit, IMO) OS that they used before. No, they don't want to learn yet another user interface or read any documents without shiny 32-bit images or fluffy suitspeak, since they already know this computer stuff, roight? They simply don't want to wrap their tiny little brains around the fact that linux is not windows. And never will be, if I can help it.
I'm here after work tooling around burning backups of my MP3's -- er, data, and I happened to stumble on this article. My Win2000 box's desktop is actually pretty bare. I've dumped everything that gets left on the desktop into My Documents. I've got sub-folders and shortcuts from there to the D: drive. I keep everything in one spot to back it up.
The only function of Active Desktop I use is to create small toolbars that point to my other machines' share points. The quick launch bar has shortcuts to my most used Apps, stuff I use to help people troubleshoot and Winamp. My desktop Icons were stolen from NeXT. At one point I had Active desktop completely mask my desktop so that all I saw was a blank canvas.
I've completely reworked the Start Menu. Someone thought it important enough to put it there, so I use it, and contrary to wisdom, it actually works. I've grouped the Apps in logical ways, IE and Opera in Internet, my personal tools in My Tools, etc. It was great feeling getting Opera to see the 'net. I placed a shortcut folder above the Program folder with things most people would leave strewn about the desktop.
I've taken time to learn how to work Windows. How to organize and install programs, how to get it to do what I want and not what it wants. Getting things to install sometimes involves editing inf files by hand, or digging deeper into the CD to find out where something really gets installed from. Did you know that NT4, Win98, and 95 can fit on the same CD? Copy the i386, Win98, and Win95 directories to the same CD. The rest is useless fluff. I used 98Lite on my Game installation of Win98SE and used Opera instead.
In all my experience here the only time a system needs to be reinstalled or restarted is when someone has done something irretrieveably stupid. These people you want to rescue from Windows will be making the same mistakes in Linux. You will be exchanging one problem for another. I know the pitfalls and moods of Windows, you know the pitfalls and moods of Linux. They're going to whine to you or whine to me.
I'm not wearing rose-colored glasses when it comes to Windows, you are just not cynical enough.
The party's over
The ability to mount, unmount, rename and remount a given filesystem on the fly is what is really useful to me. Combined with a disk partitioning utility of some sort, an admin has alot of control over their system. The system can grow organically over time and can be expanded without interruption.
WHen I'm done, I close the window (how did a document become a window? Is that intuitive?) and it goes right back where it belongs, I don't have to root around for the right place to put it back. That's not something somebody unfamiliar with computers would expect.
This reminds me of when I was in grade school we did our word processing on Apple II's. The funny thing is that if somebody just loaded up the program to print the document most people thought they had to save it again after printing or the document would be gone.
END
A more common misconception is that Linux's so-called GUI's are good enough for most users.
If you have done any reasonable research into human-computer interfaces, and have used a whole smattering of OS GUI's, from Windows, to Mac to BeOS, OS/2 and the various *NIX's, Linux GUI's (of which there are only KDE and GNOME, window managers don't count) are insufficient.
Often times it _is_ easier to head to the command prompt because the GUI doesn't facilitate a certain task easily. Tasks which have already been solved on other GUI's...
The install is not as important, because it's only done once. And if Linux heads to the desktop, not even done by the end-user as it will be pre-loaded.
Yeah, i guess OS/2 had a system fairly close to what i was describing. The rexx scripting language that OS/2 used was actually used by the Amiga as well. I'm not sure who had it first. It hardly matters.
I agree. Below are more thoughts on these issues as they have come to me:
Yeah, I also have to do battle with my desire to overgeneralize. In a way i want to make a framework of basic system calls that will be available to applications/components/etc... written for this scheme, and then implement that on top of X and posix, on top of WIN32, etc...
The other part of me says ARRGH! INDIRECTION, BLOAT! Now i need to think about how to make this system portable but not bloated. My target application is purely desktop client users. Servers shouldn't even have local consoles. I'd be more interrested in allowing a slightly narrow and simplified API that could be easily created everywhere than providing everybody's favorite non-essential system call. Among things that i would consider a must to provide:
A single, sane time interface.
Threading, messageing, IPC.
Tools for seperation of application and UI.
The last is the only one that i see as particularly novel. The trick is that although programming models like microsoft's pay lip service to that idea (seperate classes for object and view), they also make sure that they tie you to one specific UI: THEIR UI.
What i really want to do is make it so that every GUI and CLI command translate into some primitive in a basic communication model. This will scale from the basic primitives all the way to the largest scale apps, and there would be no hidden internals. A user (or more likely script) could pass a message into a program and it would be valid, normal behavior...
Hmm. Much thought required.
---
Play Six Pack Man. I
This topic reminds me of the middle ages, when it was thought commoners just couldn't be expected to learn how to read. So elaborate art was commissioned to teach the Bible via pictures, and the rest was handled through the oral tradition.
Oh, and not teaching the commoners to read had nice side effects for the upper classes too. Then came Gutenberg, and the next thing you know you Tony Blair is tossing out the inherited members of the House of Lords :-).
Fast forward to the 21 century, we see history replaying itself -- now commoners are considered too stupid to program instead of too stupid to read. Having watched secretaries with high-school educations become TeX Macro wizards, I can say the commoners are pretty damned smart, and are starting to notice they have nothing to lose but their chains.
"Intuitive" in interface terms means only that you can use information you've already gathered to guess what will happen when you're faced with something new. In the case of the first GUIs, such as the MAc circa 1984, that meant using a desktop metaphor to allow a user to extrapolate from a real world desktop to the Mac. Unfortunately, today we haev many users who understand (to some level) and work with Windows and the Mac. Therefore in order for a GUI to be intuitive, it must go some way towards meeting the expectations of those users, who *will* expect it to work the way those other systems do , because in their terms "That's the way computers work". It's inescapable.
It doesn't mean that you can't do something new, it just means that you need to think about the way in which a user experienced on existing systems will perceive it.
ben_ the technologist and platform agnostic
When I first came to university there was a VT101 terminal for use accessing the Library Catalogue - this this was great, menus at the top that you accessed with curson keys, context sensitive help was always at the bottom the screen, searches were quick, you could use the think without your fingers leaving the keyboard and advanced searches weren't too dificult to do.
When I left university they came up with another way, the web way. Now instead of a dumb terminal you need a computer with a browser (more than 4x the cost) you need to know how to use a mouse and a browswer (those cursor keys are real easy) you have to figure out the weird interface to the search, the searches take longer, the data usually has more screens of crap between you and it.
Usability testing, do it.
I think most people here understand this, but the tone of this "story" seems to imply that there are so-called intuitive interfaces out there.
Having done several years of desktop and internet support, I can tell you that NEITHER Windows or the MacOS is "intuitave" for most of whom we might call "joe/jane random user". Learning to use any UI is like learning anything. You have to learn, granted the learning curve for dragging and dropping icons into folders *might* be a little less steep than "cp file1 file2" for some.
I spent most of my time doing phone support giving "windows" or finder "101" lessons.
People do NOT "intuitively" grasp "windows" any more than they intiutively grasp carburator repair. They have to learn. The questions to ask when you're building a UI is: "Is this UI going to be easy to teach?" - "Can useful documentation be written to explain the basics of operation?" Maybe even "Is this system going to be efficient to use once learned?"
The whole idea of a "perfect" UI is a crock. Until UIs are completely transparent (i.e. you get to explain to the "computer" what you want in your terms) -- which isn't going to happen RSN -- there will be people who don't like computers.
Having said that, the only potential difference between KDE or GNOME and Windows or MacOS is market share and (for the individual) "Do I know anyone who can help me work my new machine that runs ?" - which is a reflection of the first difference...
Just my 2 sense.
Ahh yes, I'm sure they did no research and just did the opposite of Macintosh for that very reason. In fact I bet they made no coorelation between users of applications and the applications themselves. I bet they didn't even ask them where they wanted the standard components of their applications, which they then laid down as a standard guideline for every application thereafter to avoid lawsuits.
Why would anyone make an assumption about the left(begin) to right(end) way of thinking when designing the UI. Then use that procedurally throughout the entire design, from menus to even the way you click on a button... O look you left click it starts the app, you right click close is right there! O look the scroll bars are on the right side so you can manipulate the fate of the window as the mouse pointer is tilted leftward as similarly would your right-hand if you were reaching towards a desk, (perhaps because people feel so use to turning pages with the "*EDGE*"). Shall I go on about how the default setup of the desktop is obviously ergonomically designed for a right handed (the majority) user as well?
I'm just focusing on this because you seem to think it's so ludicrous. Look at it dude, it's consistant down to the last minute click and point.
Yet this was never considered?... brilliant. Forgive me for noticing an *OBVIOUS* pattern, in this shoddy user interface.
The best thing about windows is it's interface. Haven't you ever seen those commercials on TV with the woman from microsoft testing childrens reactions to the ui of windows? Microsoft has invested large amounts of time and money to uncover the psychology of what people want in a user interface. I used windows for a very short time before I got into linux/unix as my OS of choice and I still have the propenisity to design my UI's with the same consideration MS has taken with windows, because that what feels natural.
For instance the x close button on the top right side. I can see how that relates in western readers of books, when you are done with the page you reach to the top right corner and turn the page (close the window). When you start reading again you start at the top left side, "File->open->new file" (similar to the mental processes that occur when you start the new page).
That's just one of the few examples I came up with off the top of my head.
Lovely article
Posted twenty times before
Burn the heretic!
Thank you.
One problem is of course that multi-discipline cooperation means hard work. Suppose Psychologist A suggests that the interface has to be changed in some way. This involves work on the side of the programmer.
The question is: Why should the programmer do that extra work despite the fact that he himself operates very nicely with the existing interface? He does not get paid for this, he does not benefit personally from this, it was not his idea, the work is not intellectually challenging. Why should he care?
So the factor of motivation comes into play. Concerning commercial software, there is money and perhaps the PHB to obey. But where is the motivation coming from in OSS projects? And does it suffice to do boring work for the sake of success among users who normally don't interest you?
I think this is a problem that has to be solved.
-- PsyBorg
AfterStep and the package management stuff is a good example for form and function.
The author writes in the opening paragraph:
What you don't hear is how innovative [open source] interfaces are. Why? Because they're not.
As if that's somehow bad.
One of the most widespread fallacies of user interface design is that innovation is good. It ain't. Innovation in a user interface is bad, unless the benefits of the new interface justify the cost of learning yet another interface.
Consider the car's user interface: steering wheel, pedals, shifter. Hasn't changed in half a century. Now, what happens if some whiz-kid designer comes up with a new innovative user interface for your Ford Escort? Maybe he replaces the boring and not-very-user-centric steering wheel and pedals with a simple, intuitive "omnidirectional pointing device" (joystick). Let's further stipulate that this new UI is actually easier to use than the current version, once you learn it.
Now, believe it or not, you've been driving cars for so long that you have become one with the old user interface. Despite its gross imperfections, your years of experience with its ways and nuances make it a more effective interface for you than the new one. Guess what happens the first time you use the new, "user-centric" UI. You drive like half-blind, drunken idiot, simply because you're unfamiliar with the new interface.
Has the new, user-centric interface helped you? Nope. The reason why is because it's not the interface that counts, it's how well you and the interface work together that is the true test of an interface. In other words, for most "innovative" user interfaces:
You + old interface > You + new, innovative interfaceFor any new interface where the above inequality holds, the innovation it contains is unjustified, and the interface ought not to be foisted on the world.
Yet, we are greeted with multitudes of crappy, but oh-my-yes innovative new user interfaces daily. Why, why, oh please tell me, why is this so? Why can't designers stop innovating for long enough to realize that most of their innovations are crap, or at very best not justifiably better than the present interfaces? Why must everything be innovative?
I think that I know the answer. It lies in the very culture that surrounds design firms, cutting-edge web shops, and "new media" consultancies. The answer is this:
The design culture treasures innovation more than goodness.
The design community treasures innovation to the point where, if an interface isn't innovative, it's considered bad. This is immensely dumb. Nevertheless, that's the way the design community thinks. (I've seen it in action, folks, and it ain't pretty.) At design contests, for example, "innovation" ranks at the top of the judging criteria. Your esteem in the eyes of your peers is determined by how innovative you are. Innovation pervades the designer's every waking thought.
It's madness, folks, and it must be stopped.
Attention user-interface designers of all walks:
If you design user interfaces, you owe it to your users not to innovate unless you can prove that your innovations not only make the user experience better (for the users, not your ego) but also that the experience is sufficiently better to justify foisting yet another friggin' user interface on the world.
You have been warned.
When the new revolution comes, and the dogs are loosed upon the criminals of the old world order, among the hunted groups will not only be lawyers and politicians but also the designers who innovated gratuitously.
Think about it.
Easy, automatic testing for Perl.
its NOT part of the OS...
its part of the windowmanager... You dont have to use it with linux... pick a different window manager...
... hi bingo
The real elite of Linux understand the power of the system is its flexibilty.
I see your point in arguing that Linux is used by many and should therefore reach out to accomodate them, but I feel that the point you're posting in response to is the one with more support for it because Linux was created as an alternative by a person/small group who wanted it for themselves and figured it wouldn't get far. They never really dreamed (I think, but heck, i'm not sure) that they would end up making an OS that would achieve such wide usage or great acclaim.
Still, hearken back to why it was created; and the fact that the creator/s never asked for such acceptance, nor did they shun it either. Delving into the world of Linux was primarily been a "if you wanna do it, go ahead. no one's going to do it for you." Thus the "if you think it's broke, fix it yourself" kind of attitude, etc. (sorry if my generalizations are wrong.) Therefore, if you think that developers of linux/oss should make "intuitive" interfaces as a service to their numerous users, I think you're wrong. I don't think they have to just to forward their cause. I think that yes, it could be a very good idea, as the article mentioned, if they want the popularity to continue. But if popularity is what's important now, then I think the spirit of Linux has been lost and we're now into the commercialization/winning the popularity contest over windows realm, which I think linux never intended to veer into. Furthermore, Linux is a "by the people, for the people" thing; yep, it was pretty much created by those who'll use it. However, I do realize that bunches and bunches of people use linux who didn't create it and couldn't have if they wanted to (me included.) I know i'm not the only one, but I'm not complaining. Sure, somebody probably is (otherwise, this debate wouldn't exist), but to begin to cater to those who can't pick it up just by glancing at it, you're compromising the spirit & operations of Linux.
Insert mind here.
...is of critical importance. My background is game design and before that QA (i.e., playtesting). I am a firm believer in the power of the tester -> coder feedback loop, as I have seen it work magic any number of times.
:) People with an interest in migrating their core activities to an obviously superior platform, with the inclination to be actively involved in enhancing available applications, and with enough technical knowledge not to make utterly inappropriate/unfeasible suggestions ("why isn't there a paper clip that pops up to help me?"). Communications skills also help (as in any QA role).
What I think Linux needs, and what I am trying to bring to it, is less coders. Note that I don't say less technical folk, because that's not what I mean. It is possible to be tech-aware without actually being a coder. I consider myself to fall into that category. To be specific, I run Slackware on my laptop (been through Redhat and Mandrake). I have all sorts of trouble configuring it, but persevere, ask friends, etc., and generally get it sorted. I am technical enough to configure most things, compile source distributions, compile my own kernels. Generally I cannot make meaningful modifications to the source before I compile it, however.
What I hope to contribute as I come up to speed on linux (been using -- hmm sounds dodgy -- for ~ 6 months now) is feedback. My personal interest is audio -- I am a heavy user of Sonic Foundry's Windows products. For the most part they are very well designed. I am constantly on the look out for linux apps that offer the same functionality (slim pickings so far =), and when I find them plan to use them and offer feedback from an end user perspective. My QA background makes me very fussy about UI, so the programmers of such a project can expect detailed nitpicking, suggestions and (as a side effect) bug reports.
So I suppose I'm saying that OSS could use more people like me
Ideally, developers would slowly form relationships with trusted end-users, and learn to trust their judgment on interface and other end-user issues.
I don't think there's much to be gained from trying to make linux a "lowest common denominator" OS like Windows. Linux/Unix already does an excellent job of being a LCD OS in a specific context -- I have a friend who works with the police, and didn't even know he ran linux! His sysadmin has done a great job of insulating him and other non-technical users from stuff they simply don't need to know about (as an aside, linux is clearly superior in this regard. Any idiot can browse to C:\Windows and delete it, but that same idiot is limited to deleting their own home directory under linux). Also, if you're going to try to make "idiot friendly" config tools you have to go the whole nine yards. Half doing it is worse than not doing it at all.
In conclusion I'd say I think things are on the right track. I and people like me need to basically start whingeing louder about UI issues. The coders who listen will be rewarded with much larger user bases. In the meantime, intelligently written scripts are a much better option that half-assed GUI tools. I found OSS (as in Open Sound System) a dream to install, lack of GUI notwithstanding.
Message ends!
gribbly.
maybe
To start with, let's not use the MSFT approach. I've been involved on alpha Test Releases, as well as internal betas, and it mostly boiled down to "If it has an error message, it's ok.
That is not the way to go.
What we need is a:
cheap (in terms of programmer resources)
fast (in terms of time taken)
better (in terms of ROI)
method for improving the User Interface.
My suggestion is that an Open Source project should have a Test Group that is:
A. unfamiliar with the application
B. has not seen the code
C. just uses and tests the releases.
These would work best as two-person groups - one to use the app, one to record the problem areas. Their job is only to document User Interface problem areas and pass these on to wish list and bug list queues.
We also need document testing groups - which try to use the docs - some to proofread, some to suggest problem areas to be rewritten (remember, it's better not just to complain but to offer an improvement), some to write docs for areas that weren't intuitive.
But don't try to solve everything. Let people work on what they do best.
Will in Seattle
I could swear this is a repeat...this must've been posted before...yawn...
It's 10 PM. Do you know if you're un-American?
Or like clicking on the "Start" button to shutdown your computer.
huh??? EVERY single person i know that uses windows uses the start button, its really great, has most of the stuff i ever use right there and i don't have to shrink anything down to get to it, i don't even have the task back visible, i just hit the start button on my keyboard or ctrl-esc does the same thing if you have an old keyboard. really the only thing in windows that needs to be copied is the task bar and the start button. Well and the control panel, and thats already there in alot of distro's and also the desktop but that was really coppied from like a mac or something, so copy it from there instead! and make sure its at the top of the file hirearchy in save/open dialogs! if its not it useless!
Sure, lack of feedback from more naive users is an issue. But that could be overcome if (in my experience, anyway) coding a UI weren't so deadly dull.
In a commercial environment, a project manager can assign the UI coding to someone. In open source, someone typically has to volunteer to do it. And who is going to volunteer to do UI programming when there is always internals to work on somewhere?
-y
150 Opening BINARY mode data connection for slashdot.sig (129323052 bytes).
> What about a 'driving' interface?
:-)
A while back, I was working for a small programming shop. We used Clipper (anyone remember) to program some application. By convention, all keyword where in UPPERCASE. We had the idea of creating a shift pedal so we did not had to reach for the side of the keyboard to press shift, then keep the little finger there while we were typing the keys. With the pedal we could use one foot to press the shift key and type all along.
It did work great! But after a while we were not typing a lot faster with that device.
Anyway it was a fun hack
Copying windows is not all bad. Microsoft did the market research to find out what people like there apps to look like, but couldnt make them work. Linux will just take there nice looking start button, and make it do something other then give a blue screen.
In the last few weeks I've come across two opinion pieces which suggested that the pursuit of "intuitive" interfaces is one of the worst goose chases in the history of computing. One of the authors was Tog, the other was one of the original Xerox luminaries (the mouse guy I think).
Intuitive for a "golden surfer" who sits down to a computer for the first time in their life and intuitive for someone who has used computers, consoles, Palm pilots etc. from a young age are "effectively" irreconcilable.
A quick sanity check on "intuitive design" is Merlin Donald's "Origin of the Modern Mind". He talks about the evolution of human language as beginning with mimetic (gestural) culture and evolving toward a more abstract (and expressive) semantic culture.
Why is it that "interface" specialists insist on boiling down interface elements to the lowest common denominator of human linguistic evolution? Why is it that we are still aflicted, a full generation after the personal computer was invented, with the "Quest for Fire" one-button mouse with an "expressive" vocabulary limited to point and grunt?
We should get rid of the entire concept of "intuitive" and focus instead on accessible, expressive, and effective design choices.
It's also being about easy to learn. Some things just shouldn't be made intuitive for the sake of it, eg. would you want a user to have to use a photocopier icon to copy files? This would obviously be very annoying. And some people mentioned the dragging files to a bin. Well, that's probably intuitive. But is it good? In the Mac, you have to drag disks to the trash to eject them. But users say they don't want to delete the disk! And in windows, how many of you actually do the drag and drop? The delete key is much more convenient.
Sometimes, it's ok as long as it is easy to learn, and once you have been shown once, it is easy and painless to do it again, even hundreds of times. Like Ctrl-Alt-Del. It's not intuitive. But once someone is shown once, they will be using it every day afterwards without giving it much thought.
Here's a fun UI site http://www.iarchitect.com/mshame.htm
I think that part of the reason that us linux users feel ify about our guis, is because thats all they are, they're frontends, whereas in windows, the gui is often the only ui. When we think of frontend, we think limitation because we know what can be happening behind the scenes and it makes us insecure about using guis -we dont trust them. hehe. we'd rather write our own conf files sometimes. Its just like id rather write my own html in vi than use some editor. Thats one thing windows does well - hide EVERYTHING that is behind the scenes from the user even knowing about it, let alone thinkin about it. It works as far as comfort goes for users who dont care... but, I'd rather have linux. What I'd like to see, are more standardized system configuration. Like, I want a standard package format, like debs, but used everywhere. And, I'd like to see all services run in a similar way... like most things are run inetd so far, but i want everything done by inetd if possible. I want a place where I can say, in all servers, "Use login info from /etc/passwd info" or "keep my own login info database" (I want both options) Simple stuff like that. Just a thought :)
Having read the original linked article the issue at hand is application user interfaces, not installation and/or setup issues. It's a question of "Does that clipboard button copy or paste?" not "Can I configure this program without a text editor?".
There are a lot of users that will never have the motivation, desire, inclination or whatever to even think of trying to install a new software package themselves, let alone try to configure it. But these self-same people can often (as another post pointed out) become macro wizards and formatting gurus. I have quite a few friends that use windows and often phone me up to ask question about simple configuration options. It's not that they are dumb, but that computers are not their field. (I actually know someone who brought their Win98 system back to the shop to get them to install Real Player).
The real issue here is the question of user interfaces for application software. There is a great need for good GUI based software for the users that just want to get their work done and not solve visual puzzles at the same time. This issue will start to go away with the arrival of large commercial companies such as Corel.
Also I would like to point out that when there is more layman friendly software under Linux this should not mean that you couldn't use your good old text based apps. Unlike Windows, Linux should be about flexibility.
Finally, on the topic of installation/configuration, this is an ideal area for a service/support industry. For those users that don't want to be bothered administering/setting up their systems, local companies can do it for them. They could alert customers to new software versions, security patches and the like, and Linux can be administered remotely (telneting in as root). This could actually be a benefit to people that just want to e-mail, surf & write letters.
"I'll take the red pill. No! Blue! AAAaaaahhhhhhhhh"
- Monty Python meets the Matrix
Actually, Alt-F4 or alt-spacebar-c will work in all applications which pop up in a window. Plus, I can navigate all but the worst written apps EXCLUSIVELY through my keyboard in windows, and I can't say that for the mac or X.
keeping the world safe for prematurely grumpy old men for oh, about 7 years now
From reading the comments most people are interpreting this somewhat differently, but as I see it, what we need is a standard application which developers can interface into that asks the user for feedback about the particular product they are using and then transmits the information to the developer. Also with user consent core dumps could be analyzed as well as interface feedback. i.e. It could simply mean there is a feedback button in the menu system of the windows app which brings the user to a comforting familiar interface which always looks the same but plugs in the relavent information from the program which it was called from. The only thing the user sees is a quick survey and a submit button. If it was included standard in even just one of the distros it would be a very valuable tool for most developers.
I'm not sure I agree that there has been no "feedback loop" for OSS development. I think there has always been some kind of feedback from users to programmers. Unfortunately that feedback has been from other coders and not from "average" users.
How many Linux users could even be classified as "average" computer users? That is and will continue to change. Now that OSS software has started to become a bit more popular it will begin to broaden it's user base.
As long as the community welcomes the new users instead of telling them how stupid they are for not being able to use a CLI. etc...that feedback loop will grow also.
The whole idea of someone associated with sendmail criticising Linux/OSS for being isolated from end users is quite absurd.
Sendmail is IMHO a good example of a very stable and configuratble, but very hard to learn app.
--
Graphical user interfaces, they get all the glory, but in this world the UNIX apps do all the work... -- Rico Tudor (BeOS developer)
Is walking "intuitive"? After you know how, yes. Beforehand, no.
The number one misperception about the computer industry at large is that computers should make people's lives easier.
They won't.
Did the mechanization of farming make farmer's lives easier? No. There just aren't as many farmers anymore.
Computers might make the job you _used_ to do easier. But they simultaneously change the nature of your work. If you can't adapt, you're out on your ear.
As far as I'm concerned, all the hubbub about "intuitive" is just a lame excuse for being out of touch and irrevelant. "It's not intuitive enough", they whine. You're not relevant enough, I say. Get with the program or get out of the way.
--Lawrence Lessig for Congress!
The "average" user is not going to want to be anywhere near the source code. If it doesn't work like they want it, straight out of the box, they're going to toss it and buy something else. If OSS developers intend to mass-market their products, they must "dumb them down" to the masses.
It's sad, but true...
http://billy.j.mabray/
human://billy.j.mabray/
"Every good system has a backup." -- Dale Hanchey
But these two things are not mutually exclusive, the ideal UI for *nix would be one that initially is limiting in configuarability (to enhance ease of use) but can be modified to be highly configurable (so power users can change the interface to make it work efficiently for them)
The problem is that neither *nix nor Windows solves both these problems, Windows is skewed toward the initial user making it a pain to deal with to get it to do what you want quickly when you've learned it, and *nix gives you all the options too quickly for it to have a gentle learning curve for new users.
-------- This space intentionally left blank --------
Actually, I rarely touch the 3-5 standard icons (Internet Exploder, My Confuser, My Dorkuments, Notwork, Recycle Bin).
I find the desktop is a great place to drop URLs that I want to come back to later. If I'm reading an email, I'll drop important addresses onto the desktop.
The icons that stay around and get used a lot are the ones down on the taskbar thingy. They hide out of the way until I want them. They are handy.
Again you're mistaken.
Ctrl-C and Ctrl-Insert are both copy in windows (both are supported by default windows common controls) and Ctrl-V and Shift-Insert are past in windows.
You seem to mistaken by the fact that there are many ways to do things, you seem to think that means that each app only allows one (and a different one) way of doing things. You are wrong.
And Microsoft doesn't say they're 'technically part of the OS', they're part of the Windows OS.
UIs are already Open source, look where that got us....look at X, KDE, Gnome etc, copies of Windows (they are!).
There is a windows standard recommendation that apps have to pass to get ceritication.
:)
And I NEVER hav eany troubles using keyboard nav in any apps i use (or prolly will use), Alt-F4 and Alt-Space-Close are all provded by windows. Alt-F-X are based on menu commands, so it's pretty obvious what they are on the app (look for the underlined neumonic in the menu).
That copy and past thing I've told you before, Ctrl-C & Ctrl-Insert are standard, Ctrl-V and Shift-Insert are standard. They're both supported by every application I'ver used (never used quicken).
And I use keyboard shortcuts all the time, it's completely feasible, if you don't know what the nav keys are for close, look at the menu (or use the two standard Alt-F4 and Alt-S-C). If you want to close an MDI window, use Ctrl-F4, if you want to cycle thru mdi windows, use Ctrl-Tab, cycle thru top level windows, use Alt-Tab, look up help, F1, search for something, Ctrl-F or F3.
It's easy. And if it's an app like (quickken) which you use often, it shouldn't be hard to learn, knowing both ways of doing it is an advantage since other apps support both, and you can use which ever is closest to your fingers
Personally, I don't think that Windows is all that great of an interface. It employs terribly mixed metaphors and is really not suited for many types of displays (read: NOT PDA's).
Okay, I'll bite; what sort of voice interfaces are available for Linux? This could be a way cool way to drive your computer...
I still prefer a CLI for many things and deal with the GUI because it seems unavoidable. But a long time ago (when I bought my Atari ST, actually) I started wondering why the different interfaces were assumed to be mutually exclusive. My Atari (and the Mac and the Amiga) seemed to based on the idea that, once you have a GUI interface, you couldn't possibly need to use a command line. (Hmm, all were closed architecture too...) Now we're speculating on a voice activated interface (VAI?) and I'm still wondering if that would end up being the only way to manipulate your machine...
So here's an idea for an "intuitive" interface. First, it would feature a "whatever you're comfortable with" interface. I.e. a command is a command is a command, whether you type it in, click on it, or state it out loud (with some properly thought out command objects for the first two, subsequent advancements in voice recognition could easily be added for the latter later). Second, it would "learn" the skill level and preferences of the user(s); either starting at a kindergarden level and building up as a particular user "progresses", or set by the user (though not directly or there would be all sorts of people thinking, "well, I'm and advanced user..."; perhaps as a first-time use pop quiz).
On an application level, AutoCAD (v14 and earlier at least) does the first one; one can point and click and drag and drop, or key in everything from the command box. I haven't found much of anything that deliberately does the second. (I.e. an "advanced" user will quickly toggle all the "don't show this annoying pop-up message again" flags as they come up, making an application kinda recognize one's skill level...)
Brutal honest time here; I've used little on Linux aside from X and the command line. Is there anything vaguely like this already?
"I'm a scientist! I don't think, I observe!" - Dr. Clayton Forrester
No.
Creating a model that fits the user is just half the story, and I take exception to drawing conclusions on how well UIs can get by just stopping there. The final goal is not to just make a program fit the actual tasks that need to be done, but to expose the mental model and functions of the program in a way that is coherent, acknowledges human cognitive limitations, and fits the user's overall goals. When the program exposes a coherent model this way, (inter)actions becomes predictable for the user, and the user can explore, learn, and get the task done. Educate the user, yes, but you can try to do that by havng the UI show what it does and can do, with just a minimal need for refrences and help-systems. Maybe only a 'Getting Started', after that it should just be clear.
Yes, this is hard. Yes, it requires multiple testing iterations with actual users for complex programs. And yes, those cycles really are mega-frustrating on the ego of the developers. (No, Bob wasn't the answer, nor that Paperclip.)
Even the programmers must have had a clear idea of the things and tasks that their program was suitable for. If they didn't, and thus made a muddled mess of a program of which they do not know how to expose the functionality in the UI, they should go back to the drawing board. Telling the user that they're stupid and should just read and remember more is just plain rude and elitist and, actually, an admission of deafeat.
This is tough cookies in the UNIX OS world because actual execution units that need to be configured are usually the result of a mashing of small programs, each with their own model of reality and syntax to manipulate; witness the mess of having to configure an actual system to do something else besides what the installer enabled. Apple bit the bullet and is at least trying to create a consumer-level solution, albeit for advanced consumers; we shall see how their XML based unified solution to sysadmining actually plays out
The real question is how intent LINUX really is on taking over the world. But making better programs, programs with UIs in text or graphics that get out of the way because they are easy to "get", can't be so bad, even for the UNIX elite, can it? There is nothing efficient about having to look up man pages for switches and syntax. Nothing. It may be an historical accident that things evolved this way, but that doesn't make it inherently superior.
FJ!!Nonsense.
Sun's talks about "network computers" (read: X-Terminals) have nothing to do with dictatorship over the users. They are about manageability in large corporate networks.
I, as a network administrator, want the users I support to be able to go to *any* station in the building, log in and get all their apps configured with their personal settings so that they can just start. I want to be able to just plug a new station into the net and it immediately get configured just like the others around. Try to do this with a standard Windows or Linux setup.
(I admit I haven't actually implemented this with X-Terminals or the Sun stuff, but it sounds more promising than anything else)
If you are afraid of losing your "desktop powerhouse" think about how much you can do with a common PC setup found in any larger company with the network unplugged. You won't be able to even log in.
Sun won't be taking away your home desktop PC. I you are believing this, you are believing marketing rubbish. It is about the corporate desktop and about how to make the admin's life easier while the user doesn't care or even doesn't notice.
Yep. The author already said that.
The question is how to adapt the Open Source model to getting feedback from the average nontechnical user.
Mounting seems just like an extra action.
Well, mounting is always done, even on Windows or MacOS systems -- it's just done behind the scenes. And some Linux distro's do the same thing. I use Corel Linux, and in their file manager, you click on the CD-ROM and it automatically mounts the drive and opens it up just like in other OS's. So you're right -- it's a good idea and I'm sure that other distros will catch on.
-rt-
** Evil Canadians are taking over the world. Learn about the conspiracy
Here's my DeCSS mirror. Where's yours?
Here's my DeCSS mirror, where's yours?
Linux installation and configuratioon could be easier, but it is quite manageable already.
KDE and GNOME as application launchers and file system browsers are of limited functionality. How much time does anyone on ANY computer spend browsing files? Minutes a day at most!
Window managing is a little more important, bacause it is needed all the time. But we have learned to abstract window management away from applications, and users can pick their managers. Windowing is used so much, that it makes sense to learn whatever you feel is best. So no problem here, only a plentitude of choice.
But then comes the real problem. The applications. They are not polished enough. Linux is an intellectually open platform which has gathered ideas from all of the history of computing. So applications have very different interfaces, and being a perfect genius with installation of the OS and tweaking of KDE/GNOME does not help a bit, when it comes to switching off automatic spell checking in Wordperfect.
On MS platforms the market is so huge, and the money involved so big, that raw competition has evolved us mostly easy to use applications. Linux applications are not numerous enough, and have not enough critical users. So the real user feedback problem is with appplications, not with basic OS utilities or UI guidelines. More Linux applications will bring more users, and more feedback. Start with more stable, flexible and efficient business/entertainment/non-programmer applications, and they will be made easier by feedback when they have first been made at all.
Anssi Porttikivi / app@iki.fi
intuitive doesn't mean windows, mac, or linux. intuitive doesn't even mean typewriters, ledger books, filing cabinets, etc. intuitive means push what's in your way, pull things towards you to use them, things a child would do. i know there is a medium towards which we need to gravitate, but that doesn't mean dumbing down, or ms'bob'ing the system. my grandma will never use computers, so lets just get over it and train who needs to be trained...and stop saying awful things.
- real hackers don't have sigs -
2) There is no such thing as a one-size-fits-all interface. Build an interface that novices can use, and experienced users will find it annoying. I personally find Window's habit of doing things without my permission and without informing me very worrisome. Skins are a good thing.
3) Usability testing and focus groups cost money. There is nothing in the Open Source model that says you can't start a commercial software company that takes your favorite OSS and wraps around it a better designed interface. In fact, if you're so smart, why haven't you started your own company? Seriously, though, just this year we have gotten to the point where there are now Linux mega-corporations like Red Hat and VALinux. I expect them to spend the money to research new user interface ideas, and then build them on top of Gnome or KDE as appropriate.
4) Open source is better at building infrastructure: true. Commercial vendors are better at meeting user's needs: you must be kidding. I've got news for you: Microsoft doesn't listen to my feedback, and I have doubts if they listen to anyone else's either. If they did, then why does MSOffice still have a talking paper clip?
5) You must have working infrastructure BEFORE you need innovative user interfaces. The open source movement is still building the framework for software that will be used for the next hundred years. I'd say that their priorities are correct - concentrate on functionality and reliability first and foremost. Leave it up to commerical vendors to tailor it to specific user's needs. GUIs will come and go; BIND and SENDMAIL are forever.
6) There's a old saying in open source: "Show us the code." Rather than bitch, anybody that wants to can develop a better user interface to any program they want. If you think a specific program can be improved, go ahead, improve it! Make it do everything your little heart desires. If others agree with you, it will go into the next release. If not, well then, maybe you've just created a code fork. But hey, you've got the interface you wanted!
"Freedom means freedom for everybody" -- Dick Cheney
authors of open source software for non-hackers should from time to time ask their family members and friends (the clueless ones) to perform simple tasks with their piece of software and see what problems they typically run into: open-source software guis are often so horrible that the worst problems would become apparent in a simple usability test with 2-5 users from The Real World.
open-source clis, on the other hand, mostly rock. unfortunately, there is a gap between clis and guis. why can't i double-click on a filename in xterm to open a file? drag it to a text-processor to create an embedded object?
the perfect user interface for linux would be a stupid, simple interface, not just cloned from windows but with some actual innovations and more adapted to the underlying system, and more transparent to the cli. somthing like that would have the potential to be accepted by both mainstream users and those among us who eat sendmail.cf files for breakfast.
and now for something completely different
UI Guidelines probably wouldn't work for linux - what incentive is there to comply? - but something like an "interaction quality" guide might, I would think. Documenting user expectations (trained or semi-intuitive) on certain types of control or arrangements of control, documenting standard patterns of UI design, documenting approaches to solving interaction problems.
Of course I don't really believe the linux comminity as it is can do it - I'm spending a lot of my time now doing interaction design and I do find it's totally incompatible with software engineering - apart from needing contradictory instincts, with my interaction designer hat I need to understand software engineering but not be frightened by the work involved in doing something that's just entirely what anyone coming to the system would expect, but just happens to be very hard to implement.
BTW, Alan Cooper's "The Inmates Are Running The Asylum" is the current bible for designing user interaction - it's methods are the best I know of.
The author apparently only knows Open Source as a buzz word. The whole idea behind Open Source is its amazing ability at contacting the end user. You open up the source code, and right there at the top of the source is usually the guy's email address. How often do you see the email address for one of the developers at Microsoft? If the end user does not like the gui, it is his right, under the OS way of thinking, to change it to something everyone will like better.
The answer is relatively simple: The Open Source movement has no feedback loop to end-users, and no imperative to create one.
That is almost but not quite entirely not the reason the UI's in most OS programs suck. The reason is that most of us like that friendly # symbol from which to do our work, and don't put the time and effort into making pretty buttons and psychologically adapt windowing interfaces. 'nuff said.
"May the Code bless you and keep you until the day of your Compiling." ~Requiem
It also has a well-defined notion of what is a configuration file and how to deal with these when (un)installing or upgrading a package. There is never any question about where a file goes. It hides lots of messiness from the user, and an inexperienced user can easily set up the configuration program to only ask high-priority question. This allows the inexperienced user to be hidden from much of the complexity that is going on, but if needed it is trivial to access this complexity. Now that is what I call a good user interface.
As you can probably tell, I like Debian. Even upgrading from RH6 to Debian 2.2 frozen.
Imagine what it would be like if every time you went to the doctor when you got sick, the doctor ridiculed you for your lack of health awareness. Or when you went to get your taxes done, the accountant laughed at you when you ended up having to pay the government.
Very few people in the Open Source movement are truly visionary - most are masters of a certain skill set, much like a doctor, lawyer, accountant or professional athlete. Yet many feel that this skill set endows them with the right to look down their nose at others less skilled than themselves.
If the goal of the movement is to truly crush the competition, we will have to play the game.
bun-fhuinneog agam!
The author here has it dead right.
Most users are not programmers. That doesn't mean they're stupid or clueless. They just aren't programmers.
And they don't want to be.
I saw a post on Slashdot along the lines of "Educate the user. Make the user a programmer and everyone will understand."
Users don't want to be programmers. They want to write letters. Or publish a magazine. Or mix some sound.
The only people who care what's going on behind the scenes are other programmers.
JohnnyB - johnbowman.net
Jakob Neisen would say something along the lines of users spend most of their time using other programs. Do what the other guy is doing.
The reality is most users are using Windows. Monopoly. Unfair competition. Whatever. That's the fact.
JohnnyB - johnbowman.net
I can't agree with you more !
Is a said in an earlier post today on this topic, multi-disciplin is IMHO the key of success if you want to have rock-solid apps with nice interfaces on top of them.
That implies psychologists and professionnal designers (in other words, artists) too.
The mistake of much of the open source movement is to believe that application design is the work of hardcore programmers alone.
I've no proof about it, but I'm convinced that more non-CS people have worked on the design of Windows or MacOS UI than CS people have worked on the code for those UI.
Stéphane
Instant Karma's gonna get you, Gonna knock you right on the head (John Lennon, 1970)
He told me that he doesn't touch the (f'en) .cf file and if he ever sets up a new box he takes an existing file and then messes with things he knows are safe. When people are amazed by his accomplishment, he smiles knowingly, and nods....
Let's say your disk was getting full. All the admin would have to do is copy, for example the contents of /usr onto the new disk
. Then just mount the new disk to /usr, and everything looks the same to the end user, and all the programs that look for things in /usr too.
"Reality is less than television."-Brian Oblivion
I like this idea! Most people, when they talk about UI standards, seem to be saying "everyone will be forced to do things this way!" One of the things I like best about X is that there are so many different ways of doing things. If you don't like the way this graphical toolkit does things (or whatever), use another one! So any standard that works against this choice is useless to me.
-RickHunter
--"We are gray. We stand between the candle and the star."
--Gray council, Babylon 5.
Exactly. I personally like the "Linux" way of doing things a lot better than the "Windows" way. Especially since the Linux way usually doesn't hide necessary details from me. People just have to spend a bit more time learning it. Printed, well-written documentation might help a lot with this. O'Reilly's been doing well with some of that, but they're only one company.
-RickHunter
--"We are gray. We stand between the candle and the star."
--Gray council, Babylon 5.
There is no need to split hairs about the definition of intuition.
Find two people who have never seen a computer. Sit each down to learn how to use a computer on their own, without outside instruction (not an unusual situation). Give one of them a Mac. Give the other one a VT-100 terminal with a telnet connection to a VMS environment.
Now, who will be the first to print a letter? *Clearly* the Mac user. Who will be the first to launch a web browser? Again, the Mac user. The first to send e-mail? An extremely quick (and lucky) VMS user *might* win this one, but the odds are low.
The Mac interface is *more intuitive* than the VMS interface. It requires *less learning* than the VMS system. It is also more commercially successful. VMS is great for many things, but it requires more learning.
-- Stamp out entropy. ->dryguy@bellsloth.net
How about a monolithic install? The fact is that most people buy a Wintel box with Office pre-installed and never buy another application. The only exception are the gamers. What folks want is a CD that they can pop in their computer that will install an OS and an office suite. They would absolutely LOVE an upgrade CD that would update both the OS and the apps while leaving their documents in place. In fact, there could be specialized CDs for people who want to edit home movies or publish a newsletter. J. Random User would be in heaven. No decisions to make after he leaves the store. I don't mean this in an arrogant sense either. Some folks want to mess with their cars and keep them on at the peak of performance. Others just want a Honda Accord that will get them to their destination reliablyand that's fine.
Probably somebody is going to flame me for this, but it is something I have been wondering about. What is the use of mounting? Now I am just a lowly Windows 98 user, but I absolutely can't see a good reason for mounting nowadays.. a drive is there or it isn't. If it isn't I should get a message that it isn't there. (Something more intelligent then A R F.) One of the key things in designing any system, in my opinion, is that you minimize the amount of actions nescessary to complete a task. Mounting seems just like an extra action.
Use Adsense for Charity
This article has already been posted on Jan 28th by 'Taco. Unless it's been updated since then it is redundant. Perhaps when Taco he gets back we'll stop getting the duplicate stories. If you're considering posting to this thread you should probably read this one to make sure your not repeating yourself, or someone else (unless of course someone has already pointed out that this is a duplicate article (I couldn't see anyone), in which case I'M THE SIMPLETON).
The one thing that I notice while looking at all of the user interfaces that have been thrown in front of me is this: does everything *have* to be so Windows-like? KDE is practically a Win 95/98 clone, GNOME seems pretty similar, and a lot of 'wm' customizations seem to take everything directly from M$.
I would think that Open Source and Linux would be the perfect place to try new front ends and methods of viewing data. (Like those mentioned just about every one of those PC mags...)
Besides, I'm thinking of filing suit against KDE. I mean how *dare* they make a browser part of the operating system! (in jest, of course...)
Case in point: the only reason I use MS Windows at work is because that is what a majority of the people at work use so my hand is forced; however, because MS has such underhanded tactics as not making their office products backwards compatible, there is some discussion about moving away from them given that we, (we being the tech group), have found that open source software is simply easier to maintain. Ergo, it's better.
- tokengeekgrrl
"The spirit of resistance to government is so valuable on certain occasions
Is not some clone of an existing UI. We need to create the next user interface. We've seen the popup menu system replace the command line, and the GUI replace the popup interface. It's time to replace the GUI with something more usable. Moving a mouse to point and click requires too much time. We need a keyboard-driven shell with graphical capabilities. For once, the Open Source community should be the ones doing the innovating. Face it, copying the Windows interface is merely admitting the superiority of the design.
The society for a thought-free internet welcomes you.
I absolutely agree with you on all points. There are some interesting disperities (spelled wrong) between all of the programs available for Linux/X Windows. For example, it is incredibly annoying to be able to scroll through menus and lists with ease using my scroll-wheel in GNOME and then switch to Netscape and it doesn't work, or Emacs and it doesn't work. The copy-->paste functions certainly do not work between programs at all. It is as if every programmer thinks he is in a "I know best" bubble when they are programming and thinks that OSS means that nobody should have to follow any sort of UI guidelines. Setting up a sort of "standards" body for UI design would help out Linux immensely. Nobody wants to learn the command line anymore. Only people who were brought up on it, like myself, think it has any value at all. Should Linux UI be like Windows? Why not? My mother came to visit and wanted to check her email from my computer and do some other stuff. The minute she ran a program without a "File" in the menu she just gave up. Only when I created a user for her in Netscape did she understand how to check her mail. People today think about their computers as more of an "appliance" and really don't give half a shit about how it works. And don't tell me people do. That is why Windows works for people. They install something by clicking on "setup," they don't touch anything else, they click on the new icon, and *blammo* their program is running. This is "consumer level" UI design and it works. Linux does not have this and it needs it badly. Sure, "geeks" and "power-users" and the such can use Linux how they see fit, but people like my mother could give a shit less how "wonderful" Linux is if she can't just sit down and use it without any "hassles." And people don't want a Linux "guru" to sit down and setup a system for them when they buy a computer or invest hundreds of hours learning to use Linux when they could be doing something else. Windows actually offers this to people and that is what they expect. So, I wish Linux the best of luck. Right now I am using BeOS for Intel for the simple reason that the GUI is incredible and so is it's simplicity. When I wanted to use my Brooktree 848 capture card in Linux I had to recompile the kernel, load the drivers in conf.modules and then fuck with configuration files for hours. That's bullshit. In BeOS I simply dragged a file named Brooktree to a "devices" folder on the desktop and *poof* my video card came right up. No hassles. See what I mean. Linux needs this and it needs it badly.
Your VMS user will see a keyboard with letters on it and might think to type the first word that comes into his head, "help" for which he gets a full page of help and maybe some instruction. Of course, this assumes on the learned use of typewriters and that the user knows to press return afterwards.
The true answer of course is that any new user should have some documentation and preferably a little tuition before using a computer for the first time (At uni, we got issued with a "getting started" pamphlet which expressed what was needed to use unix in a few pages without requiring page after page of explanatory graphics as you would for Windows.)
About the only really intuitive interface is a touch screen with a very simplified interface behind it and even then, experience shows that users will often need a little encouragement to start prodding the screen. But the "point with finger" is very low level and even instinctual. The mouse tries to emulate this but needs to be learned and often the interface it interacts with is certainly not intuitive.
Rich
Rich
In these cases, it's more often that the user asks in an obnoxious way or demands that someone give them the answer and really, they should probably consider that someone more than likely has had the same problem as them and they should be asking where to find the answers rather than asking the question directly. In many ways, it is a good learning experience because, firstly, the user learns how to ask useful questions rather than clueless ones, secondly, by finding the documentation, the user is more likely to get many more potential questions answered and thirdly, the forum in question doesn't get cluttered to uselessness by users asking clueless questions.
Don't forget that the answerers are not being paid for support and can respond any way they see fit. And if someone can't make it through this "trial by fire", they probably don't have the right mentality to cope with Linux anyway.
A "clueless" user who is not unixified yet asks: "Why is X configured so many places?", "Why is there an stty, a termcap, a keyb.conf, an xkeymap, a resource file floating all over?" or another finger-on-sore-spot question and immidiately is flamed as Microserf of Gates lover or other verbal abuse.
No excuse for rudeness here but there is often a reason for things which is not easy to explain or too long winded. By using things the way they are, a user will come to understand. Understanding the way Unix is is a mindset that takes a time to develop.
A "clueless" user has a need for a program to do this or that. He is not a programmer but quite knowledgable in his field of profession, and poses an idea for an OSS project. Some replies that "If he wants this $&?! program he can do it him self!"
There is more to a project than programming. Often, users will take the attitude "I want this, you guys do it for me, let me know when it's done". Well, that doesn't cut it in the OS community. Manage the project, draw some graphics, make some design specs whatever but if you want something done, you're going to have to put some in yourself.
My girlfriend is a fashion modeler and uses programs under another OS to make models for sewers. She don't know a thing about programming but all about clothes and asked me, how such a modeling program would come to be in an OSS world. She is the "clueless" user of computers, but the skilled modeler, who desires to contribute to the OSS with her knowledge if she can convince a programmer to transform this knowledge into programs.
A fashion sewer designer? Well, there are two options. Pay someone else to do it (OSS does not preclude this) or find some interested people to form a base team (others may join in if interest grows). If neither of these are acceptable, maybe it isn't a viable open source project.
If feedback an input from users are weighted solemnly on the user's computer skills, the OSS community will never go beyond making programming tools.
So what, who's to say it has to? OSS software is driven internally, not by others prodding from outside. It has it's own agenda and it's own rules. The only reason you're seeing it encroaching into the desktop arena at all is that some of the user-programmers have an interest in seeing it there. Always remember that most of us gain no financial benefit from this, we do it because united we stand, devided we have to buy Microsoft products.
Rich
He explained why the programmer, with his intimate knowledge of the capabilities of the software, the data formats, the interfaces, the controls, the options, is the last person in the world who should be defining the interface that a non-programmer user will use.
Exactly. This to me is the most important point of all.
I'm afraid I will have to disagree. Installation isn't the only thing that Linux GUIs need. Linux GUIs to me are just not "slick enough". That's a hard sentence to clarify. Yes Windows has many, many design flaws, some serious and some not so. But the interface is slick. Gnome and KDE to me are just clumsy, even though they are partly based on Windows. Don't ask me to quantify that any further, but that's my opinion - I prefer to use the Windows GUI with all its flaws than fight through KDE and Gnome.
For me anyhow, a lot of the problem with Windows installations is not [running the thing, rebooting, remembering to log back in as admin, installing the other stuff that was required but the documentation didn't mention it, banging my head on the table, rebooting again....] but rather the fact that when it's all done, I have no idea where the **** the installer put anything. I use the start menu because I have no choice - the files installed are in no useful place at all, don't have names that would help me recognize them in the least, and half of them wouldn't work if I double-clicked the icon anyway.
In a Mac installation, by contrast, it asks me where stuff should go, I tell it where to put things [possibly restart, bang head...] and when it's all done, things are where they oughta be, they have names that make sense, and if I want to put them somewhere else that would make them more easily accessible, it won't break anything.
What seems to have been missed here is that UI includes lots of things, including the filesystem actually being self-descriptive - Windows has a start menu, and that's the one place where all the good UI in the entire OS is concentrated. Most Windows users only use the desktop and start menu because if you go any deeper, it's a tangled jungle.
Strange as this may seem to some *nix people, I love Macs because I love to know how the OS works, and how things are arranged - because the UI extends to low-level stuff.
I can't talk about Linux installation processes, except for actually installing the OS itself, and my experience there has been that not only does it fail to do so, but it also ****s with the other OS it's to cohabit with (Mac OS, in case anyone hadn't guessed yet). As for the filesystem (this gathered from working on machines at school), it looks to me like it's kept neat the same way a ten-year-old keeps his room neat - shove everything under the bed in no order at all, out of sight. By this I mean a bunch of unusefully named dotfiles stashed at the root level of my home directory, neat only because you don't see them, and everything else crammed into a folder named, accurately enough, /etc.
What is the robbing of a bank, compared to the founding of a bank? -- Bertolt Brecht
I've neer had a problem with prefs files corrupted. Mostly my problems have been too many ResEdit hacks getting into conflicts, or just plain buggy software.
Well, maybe the Linux filesystem is hot stuff, I'd love to find out...
What is the robbing of a bank, compared to the founding of a bank? -- Bertolt Brecht
I seem to remember that rpms do sometimes install shortcuts into my GNOME menu. So we have that allready.
And RPM's are actually easier than a windows install; You don't even have to tell the program where in the filesystem it should put itself. (Actually I don't even know where they go). It isn't neccesery.
This is how a userfriendly computer should be. And it is only on linux, not on windows:)
Natural evolution - it seems to me that the good ideas will naturally rise to the top. With no Borg-like organization steering the development to where it is the most profitable, it seems that on the whole, this will just naturally get better. Sure, some good ideas won't make it, and some bad ones will, but overall, I think Darwinism will prevail here. My mother - I used to be convinced that it would a looooong time before I could ever get my mother (fill in *your* typical home user here) to abandon the dreaded Winblows (tm)(r)(c) - after watching her for the past few months (she's been "connected" for only a few months now) I realize I was wrong. And believe me am I glad to be coming to this realization. To say she's fed up with Winblows flakiness is an understatement. If someone ever finally makes a Linux distro easy enough for my mother to install, she'll be there. I cannot count how many times; now that she's competent and brave enough to venture out into uncharted territories; she has told me "I clicked here, did this, and WHAM it does (add goofy unstable behaviour of your choice here)!!!"
mas cerveza, por favor politically incorrect stu
Well, I question your grandmother reference.
Designing a windows-like interface on the assumption that even your grandmother will be able to use it (presumably because it's "intuitive") flatly ignores the fact that your grandmother--er, my grandfather--doesn't actually find windows very intuitive at all!
IMHO, Microsoft attempted to kill two birds w/ one stone when they released their windows GUI:
1. Compete with the innovative and relatively non-counterintuitive Mac UI.
2. Provide some sort of intuitive interface to their own OS, filesystem, and utilities.
I've listed them in order of (presumed) importance :)
I think (1) was accomplished more through MS' now-legendary strongarm business style, and (2) was never accomplished at all, really. Once MS dominated the UI market, any need to improve the UI to the point where any end-user could actually use it more or less vanished--and windows has failed to become significantly more intuitive since then. It's interesting to note that neither has the MacOS, and that all attempts since then have been simple variations on the same Xerox PARC theme.
I know that the above statements are probably tasty flamebait for all you OS zealots out there, but keep in mind that most of the "rilly rilly cool sh1+" that has been added to your fave OS/UI is not "improvements that make the UI more intuitive to the new user" but "bitchin' shortcuts and utilities for the experienced user/geek".
My question is this: since the issue is clear, where is the Open Source UI development initiative? Why is everyone sitting around complaining that "intuitive means copying windows, even though windows sucks" while continuing to code for the cognoscenti?
My answer is this: I'll be glad to use any OS/UI that /. gives me, and provide end-user feedback. I'm not a coder--but that's okay, since I'm more interested in an OS that works, and a UI that is easy to use. I do do work in a computer-technical field, so I do have some knowledge of the power-user's needs, though :)
If you're interested in forming a developer/user relationship for the purpose of researching and designing a truly new and useful interface, drop me a line at otteratkmfmsdotcom.
Any sufficiently well-organized community is indistinguishable from Government.
I do not think that this is a lasting problem. The reason is that there is a big number new non-hacker linux users. These new users will prefer office solutions like staroffice and Koffice to older editors like vi and emacs (which I use myself) Everybody that makes open source programs wants to have users, and when it becomes visible that the somewhat friendlier office solutions steals away the users, I think that open source developers will start to think differently.
Whilst the interface is important, the interface alone will not make or break an operating system. The main reason why people use Windoze is because they can get Microsoft Word for Windows but not for Linux! Like it or not 80+% of the computer population use Word...you email a colleague a document in Word and he or she can open it. You send someone an Excel spreadsheet they too can opening it. Until Jot becomes fully Word compatible (in which case Micro$oft will sue the pants off you), there is no incentive for anyone to move to Linux.... they just cannot use it as a day to day work tool. True Linux has some great applications, it is robust, reliable, open, customisable...you know what I mean. But there really is nothing in it to attract the masses (who only use computer for work, games, eduction or downloading Porn off the net). I believe that the lack of applications (and the fact that it was tied to proprietary hardware) is what caused the death of OS/2. Word Perfect's initial lack of support for Windows cause the death of Word Perfect. If we are not careful, Linux's lack of compatible productivity applications will kill it.... it will become the Betacam of the future! If Micro$oft were to own Linux, put the pictures on it and then release Microsoft Office for Linux, Outlook for Linux and (please try not to puke) put paperclips and Einstiens on it [well maybe a little annoying animated penguin that asks if you want help]... you will find that Linux will take off. But would we want to use it??? My $0.02 ReportDave... a real life report...but who do I write for??
Well, I have been using linux intensively on my system for the past several weeks now, and there have been several things that bug me. Number one is the plethora of different appearances of programs, I suppose dependent on what toolkit was used for widgets. Secondly is the not-necessarily logical setup of directories (like the several different etc directories). None of these things are easily explainable, and it took me forever to even find out what to call the filesystem structure just in order to find the FHS documents to explain it to myself (and I'm still not totally in the clear :) ). I think that these standards should be explained in every distribution, and that there should be a generalized toolkit used for compiling X apps that will then draw on the widgets and functions of a particular UI. Dan
Not in all cases(); People like gnome, KDE, enlightenment, ICE, ....., Look at Corel Linux, that is quite windows looking, people like this too. That's the point though, give everyone the option to select what they want. I guess if we introduced them to a windows like interface for a start and later on they might look at "options" and switch on another more advanced window manager. They will love it and may find the one they like, you could call them themes almost. Just my opinion though, gnome/enlightenment are the most advanced window managers in the world. Yes they are not always cohesive but nonetheless the best. Not to say that KDE has any problems either. Diversity is the UI I want.
AAARRRHGHGHGHGH!!! Intuition != Familiarity. A baby has never seen a nipple, yet knows what to do: intuition. I know how to change my video driver in windows because I've done it millions of times: familiarity. These two things are not the same.
You're right, they're not the same, but I wasn't trying to say they were.
This is a bit difficult to explain. It's not that you're used to doing a certain thing, it's that when new things pop up you 'intuitively' know what to do -- it's how you behave. You're faced with an unprecedented situation, yet you still know what to do, just like the nipple. It happens all the time, with everyday life.
Even if you disagree with the word 'intuition' in this scenario, it's still not simply familiarity -- there's still rabbits coming out of hats, and if you look at why a child knows to take a nipple, you'll note that it's also an instinctual drive. Instinct being distinct from intuition as it is a pre-programmed assumption, intuition being more flexible.
Mandelbrot went on for the longest time about how he needed to 'create intuituition', and I think he had it right.
---
script-fu: hash bang slash bin bash
[ approaching AI ]
If the Open Source world actually cares enough about this to act, which I'm not at all convinced of, there are a number of things I think might be viablly done.
.look file is totally incomprehensible. Now image if a user could right click on his wharf, chose "configure" from the context menu, and be presented with choices about background tiles, text, and items, all in a graphical manner. With similar options for the entire interface, Afterstep could be a very viable end-user GUI.
By the priciples of open source, it would be best to provide a large bank of end user input to a large community of programmers. One way this could be accomplished is something along the lines of a user interface poll. This ideally would not be simply a list of checkboxes, but something more creative, such as a simple UI designer which allows the user to place elements wherever and however he likes. Common design choices could be integrated into applications. The only real possible problem is that people would probably tend to redesign Windows, MacOS, etc.
Another possibility is making common linux interfaces more easy to configure. For example, take my favorite window manager, Afterstep. With some time and a text editor, you can configure Afterstep to look, feel, and behave in nearly any way you like. Unfortunately, to people accustomed to graphical configuration, a
And 'mid this tumult Kubla heard from far
Ancestral voices prophesying war!
Here's a thought:
We all know the OSS movement got the initial "nod of approval" for doing server programs - which blow the socks off the corresponding commercial software. This front has (and is) been done well. Now the community focus is turning to desktop applications. It seems, to me, quite logical to assume that once we have these desktop applications, and they are good enough that people actually use them - the community focus will turn to the user interface aspect.
Think about this for a moment... The OSS movement is basically a herd of restless programmers pounding on code, improving it by iteration. Every so often somebody notices an area that just wasn't taken care of. In the beginning there is plenty of skepticism, on the grounds of "this area is not sexy enough - no geek will do it" and "geeks just don't have the right attitude - they can't do it". It might take a while, but these voices are then lost in dust left behind the crowd of hackers who have turned to this new frontier - busilly making it "sexy", and learning about it and , well - "getting it".
So let's wait a little longer. Once we get the applications, someone is going to start reworking the interface, then another... Remember, being open-source means it is possible to rewrite the UI for an existing application.
In short - stay optimistic, the OSS movement just hasn't gotten to it yet - but it will, sooner than you think. And just like the other things which have, and are being done - prepare to be pleasantly surprised.
The two great secrets of success are: don't tell anyone everything that you know.
Well I dont know how intuitive Windows is. But in my experience I have worked with people that use Windows, MacOS, and Linux in their daily lives and I got to say for a regular end user MacOS wins. Whereas with someone that knows a thing or two Linux wins. Windows is the go between, GUI enough for most but tweakable enough for regular users to trash the OS daily. My opinion something that offers Mac type GUI (protection for tweaking critical components) and switchable to something console or command related (total tweakability) would work. It would allow both camps to be happy.
I have to agree with you. When I started using linux at home a few years back, I was a "newbie", and it was difficult to get used to. Now, I am still using Linux, and I still find some things about it difficult. Of course, much of the same could be said for the Nova that I used to have...
During that time in between, I sold the Nova, and I migrated back to Windows and Mac. After some thought and dollars spent, I de-installed both Windows and MacOS in favor of migrating fully to Linux and OpenBSD. Unfortunately, I couldn't get the Nova back. Why would someone want to do something that is difficult and requires wrenching? Why would someone want a Nova with a kickass powertrain and not much for looks?
Because it kicks ass, thats why! And you don't kick ass by having it easy. And, shit, if I want it to look kick ass, I can fix the body/desktop myself and paint it any FUCKING color I want. I can put little hula skirt dolls on the dash. I can have a funky blue flame as my perfmeter.
The funniest thing about the article is that it very clearly and succinctly described all of my reasons for sticking with OSS, while the author completely ignored the fact that I was the supposed target audience- which made me just laugh at the guy who wrote the article in the first place. I mean, think about his credibility here- how much usability did Wired REALLY need? *sarcasm on*hmmm, the "users" have to "scroll" down the "article"...ok, I'm confused, lets hire a "consultant" who can make this "work" for "people"!*sarcasm off*
'Hail Eris, baby, hail Eris...pfffffffttt.' *cough* 'Yeah.'
I think that all Operating Systems need to take a more scaled-down approach to their interfaces. Even Windows is too much for some people. Perhaps to keep the Web-appliance company: the work-appliance? Look at what HP is doing.(really go look-> http://www.hp.com) How much more does an average office-worker need besides a word processor, database, a browser, a file manager, an email client and a cup of coffee? Because face it: most users don't use but 25% of their workstation's total potential. And most are so un-computer-skilled, they tell me things like, "Well, if the shotcut is not on the desktop, then it's not installed, right?" Inexpensive computers whose full potential users utilize every day. Basic and user-friendly(which used to mean "makes the sysadmin think he's running a day-care, not a network) operating systems. Maybe it's time for an OS that doesn't display so many error messages, but conencts the user to the help desk for training on what they did wrong. Okay I think I rambling now... ****45*****
*45*
I'm not saying any person can become a programmer, but the average person can. Computers could feasibly be easy enough to use that Joe User could write a program simply by defining the problem. That's what 4th-generation programming is about. For that matter, that's what the computer interface in Star Trek is about (though I always wonder, if the computer understood everything else, why did you need Scotty to beam you up?). There will be fifth-generation programming, too.
You shouldn't have to be a genuis to figure out how to make and maintain a classified advertisement Web site and customize it for your needs. We all know that defining the problem and the need is half of the solution. So if the user is smart enough to know what he needs, there is no reason that computers cannot advance to the point where they could meet the user at that point.
The kind of attitude expressed by this post is the typical attitude shown by lawyers, doctors and other elitist professionals. If we techies can continue to keep the languages we use obscure, we can continue to maintain our elite position in society. And, you know what, really, it's better that way, isn't it? If we actually let the masses understand (the law, medicine, computers, government...), they would just abuse this power. Look what they did when we just gave them their choice of fonts! No, it's best just to keep everything geeky completely inaccessible to the riff-raff, and we really brilliant people who know what's best for them will take care of them, whether they understand it or not.
Yeah, right.
I did it for you and to demonstrate that it is not too hard:
1. I moved my hand to the oblong device placed right to my device with letters on it (which we will call a keyboard latter on for obvious simplicity) and composed of two buttons (although I also have seen some similar devices with three buttons and have heard of rare beasts using only one button), device that we shall call a mouse from now on.
2. I accidentely moved the mouse and to my great surprise it didn't catch fire, although a strange thing happened: there was an arrow moving on the screen as if by some magical mean, this surely is a mark of the devil's work but despite that I experimented with this mouse and understood that the arrow was following the orders silently given to her by the mouse, herself following the orders that my hand gave her.
3. I moved the mouse to the name of the author's article.
4. In an unfortunate move I clicked to the up-leftmost rectangle available on the mouse, luckily for me my computer didn't explose and instead the text that was displayed under the magic arrow, the author's name, changed its color.
5. Being a curious being I asked myself what would be the consequence of clicking the right...let's call it a "button", of the mouse.
6. To my great horror another rectangle appeared on the screen, covering the text that was beneath, in a panic move I clicked on the rectangle on a part that was filed with the word Copy.
7. Being traumatised by the preceding experience and sure to be the guinea pig of some satanic sect I hit the "Page Down" key, which caused the whole thing to disappear and be replaced by a similar looking thing with other text, I did it again to experiment with more method, and again,... until I couldn't do it anymore.
8. That is at this time that I saw something that looked different than the rest of the text, it was a white filed box with no text inside with another box on its side filled with the search text.
9. I tried clicking on it, like I did with the author's name; to my stupor and to my delight...nothing happened, except that a vertical line was appearing and disappearing all the time, which I found kinda strange.
10. Seeing that clicking on it with the left button didn't work, I tried to do it with the right button, to my lesser stupor than the first time another rectangle apeared and hid the absence of text in the box.
11. Knowing what to do this time to make this box disappear I clicked on it. To my grand demise nothing happened, so I did it again.
12. I finally found that I had to click on it on a part containing text that was in black, the only one available were "Undo" and "Paste", given that the first one was available in the first box I tried the second one, "Paste". To my great delight the name of the author, which I had highlighted before with the left button, appeared in the box.
1x. I did nothing in this number, being afraid of the consequences this could have.
14. Gathering all the courage I could find around (which isn't that much, although I am no Anonymous Coward) I tried to click on the box which was placed beside, box containing, as I mentioned earlier, the text "Search". Doing so provided me with great fear when I saw the image completely disappear to be a black page (another sign of the presence of unholy things in this heretic machine), which was, luckily for my mental sanity, shortly replaced by another page with a lot of green text.
15. The lines of green text were following a line of dark text, which was containing the text "User Feedback and Open Source Development", this looked like the name of the page on which I was before.
16. Most interesting was the second line, which was conatining the text "Open Source's Achilles Heel", not knowing what this Open Source thing was, but just in case it was a monster I had one day to destroy in the name of somebody that didn't ask me to destroy it in it's name, I clicked on this text, which made the same phenomena occur again.
To my great disappointement I saw that this new page wasn't nything new but merely an older version to the story I was reading earlier.
My guess is that if even if the people working at the company doing the interface to this site are not able to use one of the most basic parts of it then how do you want total stranger to use it???
"The obvious mathematical breakthrough would be development of an easy way to factor large prime numbers." Bill Gates,
"I'm using X right now, and it's hell, but there's apps for it that I need to use, ones that would otherwise be developed for the CLI that I love." This is an awesome quote that is as intuitive as it is sad. Waht he has said is the same thing I heard my high school computer lab teacher say when windows 3.1 was getting popular. It was the same thing I was saying when my DOS games wouldn't run. The thruth that I see in this is that if more people wanted the CLI version of the program, they would have developed it before the X version. This clearly is not the case. If we look into this paralell history furter, we remember that Windows 3.1 apps weren't nessisarily standard. Some programs worked in completely different ways. The clipboard (OLE) was just a memory block that each and every program had to choose to handle or reject. It was also back then that my grandmother still wouldn't touch it. COM was developed a while later to solve this. This is still a concept that the Open Source community has to grasp before Linux can. There are so many ways of designing an application still, with Motif, Lestif, Sawmill, etc... that every person, no matter how much of a "poweruser" they are or aren't, has to learn it virtually from scratch. This may not take a long time to do, but these frustrating moments are enough to drive away a good chunk of prospective users. Now: It is impossible with the Open Source model to *force* a user to write a program within a common interface structure, this much we know. Perhaps then it can be encouraged? We have done it with some things, like the GPL itself for instance! Think about that, because it determines the direction of Open Source software; developers conform to fit the common user and succeed with the common user, or developers unbound, unleashed and uguided... to be doomed to the underground beneath smug commercial products who will. -Effendi
-Effendi
Some people think it's intuitive because they are used to it, but it isn't.
It is like saying that Ctrl-Alt-Del is the logical key combination to reboot a machine. This isn't intuitive, it's just that you know it now that you've used it so many times... (at least if you use windows ofthen)
give me all your garmonbozia
That's right, depending on what you want to do there may be 4 or more different ways to do one thing under Windoze. These include right or left clicking on the desktop or menu bar; shortcut icon on desktop, Start Menu, or taskbar; dropdown menus; function keys and keyboard key combinations with 'ctrl,' 'alt,' and 'shift' keys.
On Windows there is no one agreed upon one right way to do a thing, some are not even documented. M$ wanted to cover all the bases, this has helped contribute to bloat and bugs. Some of the duplication may be reasonable, a typist may not want to stop typing to use a mouse. A internet surfer doesn't want to use multiple keystrokes when a mouse click will do.
Designers of Linux GUIs should try to allow users what is best for them, perhaps through the use of user programable function and keyboard 'hot' keys, shortcut icons, etc; in other words, instead of providing everything imaginable for the user, have changable defalts and allow the user to select and customize how best to interface with the computer. Just my $0.02 worth (M$ even replaced the cent sign with the $ on the keyboard!).
"Open code, in other words, can be a check on state power." -Lawrence Lessig
Here is one example. As a user, I absoulutely don't want my application X to crash for no apparent reason. "No apparent reason," BTW, is the only thing that I, as an end-user, know what to call it. What I, the end-user, do not know, is that the new application Y, which is unrelated to X as far as I, the end-user, am concerned, replaced a shared library (.dll) file with an upgraded version of the same file. But the upgraded version of the library causes X to crash. That, friends, is a very poor user-interaction model. Either share libraries like we learned to share in kindergarten, or do not share at all.
There are tons of other examples of why Windows is not the epitome of intuitive user interaction. Open Source has the means to create a solution, but it needs an end-user feedback loop, or it will never be more than a small subset of cool toys for geeks.
There are 10 types of people in the world. Those who know binary, and those who do not.
Anyway, for OS to work on designing UIs, it has to harness end-users as easily as it does expert programmers. There should be a way to do this.
For example, I can configure the pretty pictures on my Sonique mp3 player by editing a plain text file. Scary at first, but not that hard. You want 1000 swirling stars in the background? 'Swirling_stars=1000'. Click save. That's it.
Now if there was something like that for gnome or KDE, people (average non-techie people like me) could make their improvements and the developers could select the best for the new release.
OS shouldn't just be a computer techie thing. In theory it could work for pretty much any design problem.
The trick is to get the innovations that will be genuinely helpful for everyone, not just power-users. I want to configure my Word toolbars and write macros to do particular jobs. But that doesn't help the temp secretary who sits down and wants something easy to use and understand.
Cheers,
Dave
The first thing I do when I sit down at a Microsoft operating system is to move the taskbar to the top of the screen. I simply cannot deal with pushing up the start menu. I think menus should be pulled down. Please never use a menu that pushes up.
I have always assumed that this was a shallow effort on Microsofts part to be different from the Mac OS. Is this documented.
> If you think, for instance, that a mouse is an >intuitive device, you should see 80
> year-olds who never touched a computer before >try to figure them out.
I once was helping my grandmother with her win95 over the phone. I wanted her to move the pointer to the top of the screen and kept saying 'move the mouse up'. She eventually replied. The cord won't let me lift it any higher.
> Rather than considering so-called 'intuitive' >controls, the goal should be to develop methods >which are built upon existing and well-known >ones.
I would agree that most of the time you should choose what is familiar. But you should allow for thinking outside the box in interface design a little bit so that we might possibly stumble upon something better.
I was teaching my grandmother that buttons with x's on them mean close. This was not intuitive to her. I was habit for me. ( she had to learn how to close all over again when she got new software because she did not pick up the pattern. ) Maybe someone will think of a method of closing things that grandmothers will learn more quickly. But unless your really sure that it is easier to learn you should stick with the familiar.
I think any OS faces an uphill battle. Sure most people know windoze, but at my office we have to show people how to use the mouse on a regular basis. That doesn't necessarily make them stupid (although....) it just means they don't know a thing about computers. At this point I think M$'s only advantage is marketing it's product, rather than the intuitiveness of the product
BTW, haven't we seen this bit o' news before? I know I've read this artical somewhere...
-Earthman
Earthman
Say it to me face w/ out wasting space...
The whole point about usability actually has very little to do with one's technical skills or intelligence. It's a question of functionality. Even an uber-geek likes easy to use apps from time-to-time, unless they are just plain masochistic.
A prime example of a good open source app that runs on several patforms is Blender. It's astonishingly powerful, free and delivers professional quality results, but it is a pig to learn, and is far from intuitive, even for power users.
Interface design is an art in itself, just like good web-design. Poor interfaces can be the death of good programs. At the same time of course, no amount of window-dressing can make a bad program good.
The main point of the article for me was the argument that in order to create a good GUI you need quite a centralised structure, and this is one of the biggest difficulties that the open source movement faces in trying to reach a wider audience. What we really need to discuss is how this problem can be addressed, rather than how stupid most users are or how wonderful vi can be. It's not a question either of 'educating the users'.
Users need to be encouraged to educate themselves, and the task needs to be simplified as much as possible. That does not mean dumbing down in any way. It means intelligent design.
First, users aren't stupid just because they want to get some work done rather than marvel at the edifice that is "the computer." Nothing wrong with marvelling mind you, I do it myself from time to time. However, stupid implies an inability to think, far from my experiences with the large majority of users. You might say that it's the very fact that they can think that makes them so difficult. Rarely are they sitting down at the computer without some purpose (read work) in mind. Computers for their own sake have little appeal. Means to an end, nothing more. To believe this and then come up against obtuse software, well, that's a problem. Because they don't think like programmers and can't "figure it out," the "real programmers" want to dismiss it as diminished mental capacity. Sorry, faulting someone for not thinking like you smacks of immaturity at best, bigotry at worst (and no, this is not a politically correct diatribe). Coding for machines is great. There's a need for hardware programmers and the work is enjoyable (so I hear). If you want to create exquisite chunks of code that other programmers will marvel at, go for it. All that said, why are you talking about users?
Second, users are a reality. Deal with it. There may even be, in your application programming experience, a clinically certifiable intellectually sub-normal user here and there. Users don't irk me anywhere near as much as the attitude of "anyone incapable of using a command line shouldn't have a computer" (or something to the like posted by Anon Cow. elsewhere). Makes me want to say, "Gimme a call when you return to reality." Few things make a user go from confused to angry quicker than the attitude (even if unexpressed) "you are an idiot" on the part of the person(s) supposedly responsible for either creating or supporting a piece of software they may want/need to use. You think "clueless" users are a pain? Try "clueless and pissed off." You either develop a bit of foresight and sensitivity or you end up getting angry yourself, wondering when these goons who don't appreciate the work you do are going to get a clue.
Truthfully, I don't say all this with the idea that minds will be changed and again, I don't completely know all your attitude, so chances are you're a good guy. I enjoy programming and shock of shocks, I actually like interacting with users. I don't like being hollered at any more than anyone, and sometimes there are people with a problem with something I've done. Rarely, however, can I chalk it up to stupidity.
Later all.
For anyone interested in this issue there is a web site at www.seul.org that is endeavouring to make Linux easier to use. I get a periodic newsletter from them.
It's geared toward a wide range of Linux users, which makes sense given the goal of bringing Linux to the broader public. The site is fairly big with stuff for newbies and nerds. Not surprisingly there's always need for volunteers.
Most users I've seen refuse to use the start menu. It is just not "convienent" so what do we admin have to do? we have to give them a gd icon on the desktop . . .
and another thing, since when do users to ignorant to learn windows get to demand that we give them something else, ie:
at work, the previous admin before me and the current fella happened to like norton navigator a lot. dont get me wrong, it is nice, but he trained all those who are here to never touch my computer. . . huh? so now if i tell them to goto my computer and then . . . they just stop, and are completely lost
2^3 * 31 * 647
Hello,
I found this discussion very interesting and felt inclined to respond. I, as a monkey, think its very important for us non-humans be able to understand the Linux GUI also. As you all may already know, the Windows GUI and occasional blue-screen has been known to put several of my fellow monkeys into shock and dis-array, and sometimes to death. Therefore, it is VERY important, to life itself that the Linux GUI be properly designed.
As a side note, are you aware of the fact there are severe monkey beatings going on RIGHT NOW?! Many have assumed they are by tired, crazy MS employees, but noone really knows. Just remember, dont mess with us monkeys! We will fight back with a vengance! Monkey POWER!
(c)Mastah_Monkey
The problem with standardizing is that it kills competition and creativity, flexibility for the user, and improvement. The idea behind multiple OS's/install progs/ways-to-do-things is that there is someone who has a different idea and likes something different and is going to try to beat out the rest because they are, in someone's opinion, worse. So how do we decide how to standardize anything? Answer: democracy. It's what most people like or are willing to deal with and will therefore work most of the time for most of the people. This always leaves someone who dislikes what the majority has chosen for him and strikes out to make a difference for at least himself.
However, where does this leave us? The only thing to do is to let things take their course. What is happening is what people are making happen. So you don't like how a product is turning out? Sure, complaining here is great, but how does that help you in the long run? If you don't like attribute foo about product bar then go out and do something about it. That why there is Linux and GNOME and all the different options for controlling these popular masses of silicon and plastic.
Anyway, I do practice what I preach. I don't like the Windoze interface, so I use WindowBlinds to skin and customize all the buttons, and use the "Quick Start" toolbar instead of the Start menu and the Desktop, and I'm working on getting Linux to run. So I'm doing something. Go out and do something yourself, and then and only then will you be satisfied with what you end up with, and if you aren't, then blame yourself, because you made it that way. That goes for all of you.
P.S. BTW, all your comments are great, so lets do something with it, ok?
-------------
really / nothing really ever matters / so the only things that matter are what you make matter / i am in control
I guess it's just whatever you have experience with. I always say 'Grandpa' when trying to exemplify the "average user" but I guess most people have more female relations that aren't computer literate.
I don't think it's exclusive to OSS people either, just about anybody who wants to explain complicated things uses a "Relative So-and-so" name. I'd imagine it also comes from the old mentality of thinking women couldn't do anything except cooking and cleaning. Now, we say women can't understand computers instead of saying they can't understand cars or something like that.
I can tell you something (from Tog, actually) about 'intuitive'. The early Apple usability testing to determine 'intuitive' was not theoretical- it was experiential. They would literally sit people who'd never seen a computer before, down in front of (for instance) checkboxes, and say "Do stuff- and then tell us what these things do!"
Interface elements which could _not_ be figured out without help by intelligent but totally inexperienced people were changed until they _could_ be figured out, just from fiddling with them and seeing what happened. That's the bedrock upon which pretty much all GUI was built. (The Xerox GUI that inspired all this was actually a _lot_ more like Linux- it was basically Smalltalk! Contextual menus, no direct manipulation)
For example, Tog has a case study- the 'one or more' control. It's not a checkbox, because it forces one choice to be on always. It's not a radio button because it can have all or several choices on. What does this control look like, and how does it act- in such a way that a totally inexperienced person will (after playing with it) say, "Gee, looks like I can turn on one or more of these options"? It ended up being _square_ radio buttons. Tog had fought this because the turned off state looked like a turned off checkbox. Upon mentioning this to his peers, they replied with one voice, "In its turned off state it IS a turned off checkbox!"
- Turned off one-or-more: clicking it turns it on without affecting anything else.
- Turned off checkbox: clicking it turns it on without affecting anything else.
At that point it was merely a question of setting it up so that if you tried to turn off the last option, another one would turn on kind of like squeezing a blob of mercury that would squirt out to the top or bottom- something that would make people go 'Hey, it is making one option stay on!' without causing them to make up weird rules like you had to use one of the top two options, or something.Interface is _interesting_, and Kaa has summed up 'intuitive' wonderfully with this simple guideline. It's not really necessary for the first try to work if the person can guess what the feature is for, and how it functions.
For a wild eye-opener on this try reading Tog on an old Apple ][ interface issue- how do you get the user to tell you if they are using a color or monochrome monitor? This was the toughest problem they hit! Yes, the answer was a variation on 'display, in different colors, the question "Are these words on the screen being shown in color?"' Hint: monochrome monitors were often green or amber screens- and green is a color ;)
Wheee, interface!
Competent programmers _know_ that they sometimes have to throw out a design or code that doesn't really work, and excellent programmers do so on a regular basis. Many people complained when Raster and Mandrake threw out the code and design for Enlightenment and then redesigned and rewrote it, but it is now a kick-ass WM that has the highest "eye-candy" factorout there. But, on the other hand, throwing out everything about Linux that one guy didn't like, and reworking things even down to the very internals of the file system seems a tad drastic! For one thing, the reason that a lot of the designs that went into Unix 30 years ago are still around, is that they are well-thought out designs that have stood the test of time when other designs have failed. This is a point that I try to drive home with people who bash Linux by saying it is based on "old ideas from the seventies" or something similar. Unix works, and the design of all the pieces is rock-solid, having been fine-tuned, extended and even replaced here and there as new things come along. Is it the best design for every circumstance known to computing? A wholehearted "NO". But the Unix design philosophy and implementation is what works in most situations, so it is what so many people have come to know and love.
It is obvious that you don't know what you are talking about when you confuse the internals of the system with the UI. Journaling refers not to how the files are presented to the user, but how the system itself maintains the files so that they don't get lost or corrupted if the power is cut off.
The way the system works "under the hood" so to speak, is not how it has to be presented to the user. For example, the way that the system displays the layout of files as a "tree" of directories (or folders, for Mac and Windows people) has absolutely nothing to do with the way that a Unix/Mac/Windows computer actually stores files on the disk. If you wanted to have a UI that categorized files according to there type/use and kept them in containers that were operated "assigned" to individual programs, that's fine. Here we separate the UI from the actual operation of the system. That's what Unix people like, by the way, separation of data from presentation/manipulation of that data.
Also, just because you do not understand Unix and it's history, don't criticize it. The text based UI's have evolved tremendously over the years as we went from batch-oriented cardpunch machines to terminals to workstations, and the shells and text tools of today are very rich, powerful things. You denigrate them because you do not understand them. A GUI is good for many things, but there are some things that simply can not be done in a GUI environment. I like the ability to chain together half a dozen command line tools in an interactive shell session to perform complex tasks that I simply would not be able to do in a GUI program. I also like web browsing and graphic editing and other programs that are best suited for a GUI. I use several terminal windows alongside GNOME/KDE/X programs because I like the best of both worlds.
As far as competeing for the home desktop, well I'm sorry, you must have confused us for a tightly knit group of people. Where did you (or anyone else for that matter) get the idea that _every_ Linux user wants to "compete for the home desktop". That would be nice (IMHO) but not necessary, and groups like KDE and GNOME are trying to do help out, with the help of money from VA, Redhat, Corel, and SuSE. I use my Linux box to _WORK_ on, it is a tool. A lot of times that work is fun, because I am a geek. I also have a windows box for games, so I guess it is a $2000 playstation or something.
As far as your last cheap shot about DOS/Windows and Linux/X, please remember that Win3.x _attempted_ to provide a GUI and simulate a multitasking system on top of a weak single-tasking DOS kernel (mostly by working around it) and failed miserably.
Windows 95/98/ME attempts to provide a more robust multitasking system, but ties the GUI to that multitasking system. Windows NT/2000 also tries this, and succeeds more often that 98. Unfortunately, the GUI _IS_ the system, so you can't have one without the other. I like how parts of NT are designed, but I can't drop in another UI to replace NT GUI, nor can I leave out the GUI altogether and get any meaningful work done. It would be nice if NT allowed one to go directly into a CLI and did not waste any CPU cycles on the GUI. Linux provides the system, X provides the GUI, and both are very well-designed for their targeted goals, and I can interchange them as necessary. Your "uber-GUI" that you describe above is doable on X without replacing the underlying OS layer. If it aint broke, don't fix it!
If only "common" sense was actually that common...
The first question I had was "what isn't user friendly about Linux's interface?" I mean, you have your classic WIMP interface- anyone capable of using Windows will come up to speed on how to use Linux reasonably quickly.
Unless you'd like to claim that Windows isn't ready for the desktop...
And then it occurred to me that I have a slightly different definition of "using" than everyone else. I mean doing things like word processing, or browsing the web, etc. Actually getting work done.
I do not mean installing the OS, or getting drivers to work, or installing software, etc. Those I call "administration". The implicit assumption is that the user is also the administrator- and that the box must not only be easy to use, but also easy to administrate. Professional admins need not apply.
In a home environment, there is some legitimacy to this thought (although I'd debate it- how often to YOU have to go over to various relatives or friends houses to fix their computers?). In the corporate world there is NONE WHATSOEVER.
Consider for a moment how many positions are available for Windows administrators (certified or not). With a network of even as few as 100 desktops, you need a professional, trained admin. Very rapidly you need a team of them.
Or should grandma be able to administrate a network of 2000 desktops, half a dozen file/print servers, and a million dollar database server?
Once you accept that even small corporations have professional admins in residence, then it isn't a problem if Linux _requires_ professional admins. In the market it's likely to penetrate, the admins are already there. The budget is already there. And Linux is ready for at least that desktop (the home market will be a longer time in comming).
I don't really think that user-friendliness is as hard to achieve in an open-sourced project as many people think. But there is one thing the developers must do more than anything else for it to work.
Many of us have non-technical friends. They can be an excellent resource in helping to make a program more user-friendly. The trick? Listen to them! Ask for their feedback on what things they find difficult to do, what things they would like to be able to do. The fact that they are non-technical is an advantage here, because it means that the only real way for them to be able to effectively use the program is if it is easy to learn.
The real trick is, of course, making the program easy to use as well as easy to learn, and that means optimizing the program for its primary mission. It's important to avoid creating a "swiss army knife" since that will not only make the program more difficult to learn but will force the developer to concentrate his effort on something other than the main purpose of the program.
Programs that are simple but powerful are rare, but are among the best finds out there. I think that simplicity and power in a program aren't mutually exclusive when the program is done right. One cheap (and perhaps relatively easy) way to make a program powerful and simple at the same time is to create the user interface so that only the most common functions are readily available in "normal" mode, and arrange for the rest of the functions to be available perhaps via a command line (the program will be a lot more powerful if it has a built-in scripting language) or perhaps via an "advanced mode".
Just remember that large arrays of buttons and tons of menu items will make the program more difficult to learn, so it's important to concentrate on making available (at least in the default UI mode) only the most important functions of the program.
--
Use 'slashdot stuff' in the subject line in any email you send me if you want to get past the spam filter.
They cannot solve the halting problem because it is impossible!
But they make a decent attempt at figuring things out (even though doing it perfectly is impossible) and as a result can do quite a bit of optimization.
Likewise many more options can be graphically indicated than most people think. Ask an OS/2 fan some day about the Workplace Shell.
Cheers,
Ben
My usual seat in the cluetrain is at A HREF="http://pub4.ezboard.com/biwethey.ht
"It cannot be done" is too often an excuse for "Don't even try."
They are not equivalent.
One program cannot figure out what another one is doing. This is called the halting problem. However we have an entire class of program known as a compiler that typically attempts to do it anyways and figure out how to make the program faster. Sure, it cannot work in general, but it can work well enough to be very useful.
Similarly it is easy to say, "Oh, this is hard. If you cannot do X, Y, and Z then you shouldn't touch this." Sometimes that can even be true. (I don't want to trust any network whose security was set up by an automated program and then forgotten.) But in many cases it is quite possible to produce a good user interface that can do a lot more than one would expect. The value of this cannot be underestimated.
An interesting admission from a friend who is an old Unix hand (in fact he was a kernel hacker before the days of Linux). A couple of years ago he saw a demo of Photoshop in the store running on a Mac. He doesn't use Macs. But he likes photography, so he tried using the demo and found the program friendly and easy to use. Since he had a copy of Windows at home he bought a copy for Windows, took it home, and couldn't figure the darned thing out! Same program. Two operating systems. One has a good user interface, one a so-so one, and the difference showed.
Consider that story the next time you want to feel all superior about user interfaces.
Cheers,
Ben
My usual seat in the cluetrain is at A HREF="http://pub4.ezboard.com/biwethey.ht
Apple's HUI guidelines are here. Well worth the read.
--
--
Don't like it? Respond with words, not karma.
I'm afraid my twisted mind always sees the "____ For Dummies" titles and reads it as "_____ for F@ckwits"...
As in,
Breathing for F@ckwits.
SendMail Configuration for F@ckwits.
etc.
I do hope no one is overly traumatized by my language. I wouldn't use the @ except that then I'd just traumatize those who are offended by 'Fuck'...
Opps. Just can't win...
--
--
Don't like it? Respond with words, not karma.
Additionally, while it can be difficult, it is possible to incorporate feedback from "ordinary users" (or at least slightly adventurous non-programmers) into the open-source development process. Particularly since the new beta release, the GnuCash mailing list receives quite a lot of mail from non-programmers asking questions and commenting on the project.
Any sufficiently advanced technology is indistinguishable from a rigged demo
--Andy Finkel (J. Klass?)
now you know what shape to give to your OK buttons.
The only intuitive interface is the nipple. Everything else is learned.
You make some valid observations, but I think you missed the author's point. There is much more to the GUI problem than the desktop and it's icons. What's at issue are the sloppy user interfaces in the apps themselves. Disorganized menus and menus that lack important functions, inconsistent "look and feel" across different apps, and options that are not grouped properly by category. The consistency problem is getting solved by GNOME and KDE, but some of the other issues are still painfully apparent.
----
"Oh, bother," said Pooh, as he hid Piglet's mangled corpse.
It's not the most sensible or well-informed post I've ever seen, but at least it's a break from the constant din of propaganda around here.
"Christianity neither is, nor ever was a part of the common law." --
When people talk about Open Source products, you hear
about their speed, their efficiency, and their features. What
Only when people who aren't actually part of any real projects talk about
it, do you hear nonsense like that.
you don't hear is how innovative their interfaces are.
Why? Because they're not. At best, products created by
the Open Source movement offer workable imitations of
popular commercial interfaces; originality is rare, and
In the cases where the commercial interface is good and interface is
part of the design goal, (gimp vs. photoshop, for instance), this might be
partially true. However, even with a perfectly intuitive interface, very
few commercial applications will give you the benefit of ALSO being able
to configure your app up the wazoo.
In the case of something like apache vs. IIS, the apache 'interface,'
while being a flat text file, is _extremely_ easy to configure. It is also
completely well commented, documented, and easy to get free support for.
There is _also_ a gui-style configuration utility, if your definition of
'intuitive' is 'gui.'
routinely found in the underlying code. This presents an
interesting dichotomy: why is the best software writing
organization on earth unable to produce innovative
interfaces, when small commercial software companies do
so with regularity (if not always with commercial success)?
This is not an interesting dichotomy, this is bullshit. OSS is not, by any
means, 'the best software writing
people, for solving problems and scratching an itch.
The answer is relatively simple: The Open Source
movement has no feedback loop to end-users, and no
imperative to create one.
Bullshit, and bullshit. I read the entirety of the gimp-devel
and new-httpd lists every day. Those lists spend entire days hashing
over tiny changes to the interface. Your entire premise is that these
projects are coded without attention to user interface, and that shows
that either you've never been involved in an open-source project, or
you've been invovled in some that work in a totally different way than the
ones I've seen.
I vote for the latter - I think this article is a fine example of
'me-too,' 'hey-mom-look-at-me-i'm-writing-about-open-source
Guess what? Like all the equivalent tripe, it makes me ill.
The majority of Open Source software is still written for
programmer-users: the systems are made with flexibility -
not usability - in mind. Enough hooks and parameters are
Bullshit. Are you going to tell me GNOME and KDE are written for the
programmer?
But that's only a small slice of the world's software users.
Most people want a product that works right out of the box.
Gimme a percentage of applications that you think work 'right out of the
box.' The fact is that _most_ software, OSS or not, still sucks.
Commercial software actually sucks MORE, because it is more
cash-efficient to add new features than it is to fix old bugs.
I'll go ahead and
before this stupid article interrupted me, but I do wanna say one more
thing:
Don't confuse an interface that seems obscure to you with lack of
attention from the programmer. Just because YOU don't get it, doesn't mean
it wasn't very carefully designed to work exactly as it does.
i browse at -1 because they're funnier than you are.
having used Mac's and Windows machines for a long time, I'd definetly like to pipe in and say forget windows, copy the Mac.... Windows is horribly unitutive and relies on so many just plain stupid techniques.
Quitting programs - there's at least 4 different key combinations which will accomplish this, but each program only accepts one. (Alt-F-x, Alt-F-q, Alt-g-x (in Microsoft pinball where the file menu turns into a game menu instead), and alt-f4 occassionally.
Microsofts tabbed dialog boxes used to be nice, when there wer just a few screens worth of options. But now, when there's 10 or 20 tabs all crammed into one small window, it's utterly confusing.
If you want to create an interface that good enough, then by all means, copy windows to the T. But if you're looking to create a "better" UI, you'd be best to look past windows towards BeOS, Openstep, Mac OS, and Mac OS X.
I agree wholeheartedly. Unworkable, well maybe, but it's worth I shot. The big problem is getting OSS-active folks to think like Lusers. I run into the problem constantly at work. The UI people and design folks tell show me something which i think is completely unintuitive. I suggest something, and they tell me my idea is going to confuse everyone. Usually, they're right.
So the real trick is to recruit some Lusers. With the hype OSS has recieved, you'd think it might be possible to get some "regular joes" to participate.
Look perhaps to the graphic designer community. Here is a non-technical community that is generally sitting in front of a computer all day. They understand 2-d design and layout, and also computer-human interaction. But they're not geeks.
I honsestly don't know how to organize such a venture, but maybe somebody else does. just my nickel.
Users are stupid. By definition. And that is what makes it hard. You can not, as you do with another programmer or a poweruser, ask him/her to help you so that you can improove their program. No, you have to pay them for being test-dummies. And no free-software project will ever be able to afford that. :). And design the user interface minimalistic. It is easier to misplace or misname buttons if there are hundreds of them than if you just have two of them. Also, just the huge number of widgets may disturb or confuse a user.
That said, I'l like to point out some errors in the article. The first is that commercial programs often comes up with innovative user interfaces. This is some rare times true, but most often, commercial applications are must-get-it-working-before-deade-line-programs. Those are never innovative. Most commercial programs have very clouded user interfaces, filled with toolbars and buttons, just confusing the user.
There are two simple guidelines to help free software projects on this point; ask your parents about the interface. They'l probably be users, _and_ like you enought not to ask you for money
--The knowledge that you are an idiot, is what distinguishes you from one.
--The knowledge that you are an idiot, is what distinguishes you from one.
I think you are arguing a different point than I am.
Nowhere did I say that people shouldn't have access to the full power of their computers.
However, I did say, and continue to say, that you cannot abstract away all the difficulties without losing all of the power, too.
Policy-based routing, along with special cases for proxies for certain services, is powerful. You cannot make it "intuitive" in the sense of "accessible to a naive end user".
There is certainly a lot of excess complexity in modern UI's, and it should be reduced. However, never make the mistake of believing that you can somehow magically make something complicated so easy that a naive user can understand it without effort.
If you look at modern UI's, they're often much worse than older ones, because we've started getting used to them, so now there's all sorts of underlying things you need to know to even *begin* to make sense of a Windows system. "Move your mouse over the grayish areas to find pictures you can click on" is not an "intuitive" concept.
Should non-techies have access to the full power of their computers? Should non-mathematicians have access to the full power of analytic geometry?
I have no problem with people having access to anything they want, but it is *logically impossible* to make the interface simpler than the task it performs.
"Average" people, right now, have access to a great deal of computing power, and a great deal of information, that they didn't ten years ago. We're making progress. However, just as people who want to drive need to spend a fair amount of time learning about cars, people who want to use computers will still, always, need to learn how.
Natural language? Still takes time to learn how to interact with a system, what its limits are... It's always there. *People* have a learning curve.
My blog: http://www.seebs.net/log/ --- My iPhone/iPad app: http://www.seebs.net/seebsfrac/
You make a very good point; there's more than one kind of complexity, and some of them are easier to avoid than others.
I think the problem is trying to eliminate the wrong complexity. We try to make it so that an "end user" can easily do something, like routing or network admin, which is simply *not actually an easy task*.
Screw that. Let's try to get them something they can *USE*, maybe some word processing.
My blog: http://www.seebs.net/log/ --- My iPhone/iPad app: http://www.seebs.net/seebsfrac/
Well, since the article seems to be a rerun, and we're on the subject of better user interfaces (or lack thereof) . . .
I've been thinking of some alternate user-interface ideas, and one possibility I picked up from reading Tog (or was is Nielson?) was that of using "official icons"-- namely, ISO standard international symbols-- as a good, consistent visual supplement to the interface. (These symbols, I believe, are seen a lot in European countries. I don't think they're used a whole lot in the U.S.)
Problem is, I've looked everywhere, and I can't find any references for such symbols. The closest I've come to is this site detailing IEC 417 (which is why e.g. the "Play" and "Pause" buttons on media players are all marked the same), but that's about it. The ISO site helped, but not by much (their search engine sucks rocks).
So, would anyone here know of any links detailing said standard international symbols, or at least some relevant ISO numbers?
(P.S.: I already have ISO 7000 - "Graphical symbols for use on equipment")
iSKUNK!
The site sounds like a great idea! It'd be especially cool if it could provide a means of creating direct feedback loops between good UI designers and coders, and also a forum on next-gen interface ideas/possibilities. I'll sure check it out when it opens.
:->
There's just one problem, however . . . when I submit my name/e-mail to the announce list, Zope sends me a login/password dialog. Authorization fails. There's a bug rattling somewhere in there
iSKUNK!
Yes, sendmail config is difficult. But is that a problem? CLI's have a steep learning curve too... And that's what makes them so powerful. Idem for sendmail.cf. And just like with CLI's there's an easy option (called the GUI), there's the m4 file for sendmail. What's the big problem?
To quote /usr/games/fortune (and that quotes somebody I can't remember): "the only intuitive interface is the nipple, after that it's all learned."
> I've used a prototype ancient GUI where scrollbars worked the other way around; the thumb moved down to scroll up for some reason the designer considered compelling, and it was pure hell to use. Needless to say it didn't survive.
I generally like your post, but I'm not sure I agree with your logic on this part. It's trivially easy to find mechanical devices where you move something down to move something else up. If the widget you describe didn't survive, I suspect it was because it bet on the wrong horse, not that it violated any natural principle.
> Most Linux programmers seem naturally disdainful of graphic interfaces and therefore are slapdash in implementing one.
I think this may have been true in the past, but I also think we're seeing what we might as well call a "new generation" of Linux programmers who aren't at all disdainful of GIs. Whether that results in good GIs will be interesting to see.
--
Sheesh, evil *and* a jerk. -- Jade
> What's intuitive?
As they're fond of pointing out over at c.o.l.a., the only truly intuitive interface is the nipple; everything else is learned.
The obvious conclusion is that we should discard the outdated point-n-click metaphor and replace the GUI with a truly intuitive LUI (Labial User Interface) using a grab-n-suck metaphor.
However, although it is easy to imagine replacing that 19" monitor with a 36" set, and although the LUI would surely be fun to use (or at least comforting when the inherent complexity of the machine upsets you), I have not figured out the best way to use the LUI for such mundane tasks as data entry.
p.s. - Imagine the rash of "one suck shopping" patents and "look and feel" trademarks this technology is going to spawn!
--
Sheesh, evil *and* a jerk. -- Jade
This story was posted and commented upon. I guess, at least, people should read the comments already posted about this.
Emmett, man, do a search on Slashdot before you post a new story that contains such an old URL. It says right at the bottom of the page on sendmail.net that this article is from late January. I don't mean to flame you, but if you do that search, it'll probably save you a lot of hateful email in your INBOX.
Now, the hundred other people who are going to flame emmett, just save it.
---------The early bird gets the worm, but the second mouse gets the cheese.
I wholeheartedly agree. I'm a very technical user, but installing things in Linux, even when using RPMs is still harder than installing DOS or Windows programs.
DOS was easy because all you had to do is put the files on your hard drive in some random folder and they worked. Windows generally had some kind of install program that did the work for you. Unfortunately windows made removing software you installed a real pain sometimes, but at least you could get it there.
With its logical hierarchical filesystem, and a standard way (actually a few different "standard" ways) of organizing things, Linux should be an easy platform to install/remove software for.
But it's not. Certain packages require other packages. They require certain libraries. They are distributed in different ways: tarballs, RPM, .deb, whatever. Then, once they're installed, there's the problem of configuration.
When a typical Windows app is installed, if it needs to be configured it will force the user to configure it either as part of the installation process or as something run the first time the program is started. With Linux software, often you have to get out your fave text editor and mess around with a config file, often written in some obscure way, until the program is configured however you like.
While this method does often ensure much more customizability than you get in a Windows app, the fact things don't work "out of the box" gets pretty annoying.
Ugh. I think I speak for all of us. ;-)
As to the question of "how do we make the GUI more intuitive?": we just need to extend the Open Source model. "All bugs are shallow", right? Think of a user as being one of those "solitary" developers, and set up a way for them to feedback into the system. Perhaps Linus can require that all GUI code come with a mechanism -- say an HTML page that includes a mail form -- to give feedback to the developers of that GUI. Or if the developers are loath to embed their email address, perhaps a forwarding box could be set up somewhere *expressly* for the purpose of receiving and routing feedback. The mail form that comes with the GUI could come with a Subject line already set up (or with some hidden info that encodes what the subject is), so that once the message arrives at the forwarding box, it could easily be sent on to the proper people. Thus the feedback loop is closed and it's a win-win situation for both developers and users.
DNA is a Turing machine. You, however, being dynamic and emergent, are not.
I hate the way Windows handles floppies. It grinds away at random times and pops up error dialogs, instead of just not considering the (diskless) drive as an option. I don't really like the Linux way of having to manually mount them, either, but I prefer it. It represents the system state without a floppy more accurately, as an absence rather than an error state.
The hardware ought to be fixed up, so it tells the OS when someone pops a floppy in, and prevents the user from ejecting the disk when it is in use (with the main button, though there should be an auxiliary mechanical button for emergency use, just like a CD-ROM).
Linux systems can be configured to automount CD-ROMs, because CD-ROM drives don't have a completely braindead design like PC floppy drives.
One thing i've been turning around in my head for a while is a system that is UI neutral. A system where there is a command line from which EVERYTHING can be done with NO exceptions, but there is also one (and optionally many) GUI's from which you can also do EVERTHING. I'm thinking of everything from file conversions, viewing, and management to redreding, compiling, and editing.
The think i'm thinking is going to be needed is a very lightweight and pervasive object model (NOTE: This does not mean C++ necesarily, but does not exclude it...) By lightweight i mean that there aren't many levels of unnescesary and painful inheritance and bulling-around. By pervasive i mean that in at least it's most minimal definition, it is a common element to all parts of the system, from widgets to programs to whatever.
Also a simple but powerful embedded language (NOT for performing major tasks. Infact it should not have loops or other structures that would encourage such EVIL!) but for acting as the glue for connecting parts together.
This would have a lot of the advantages that M$ claims that OLE/VBSCRIPT have, but we can learn form their mistakes, avoid their pitfalls, and not have to work around ajenda item 1 being "produce code that is incompatable with the rest of the world to flex our monopoly".
One of these days i'm going to start this project. I'm one of those people who is ever-annoyed by slow GUI's (infact i've been called a Text-Nazi by my friends for my continued use of JOE, PINE, LYNX, and because i categorically reject all HTML e-mails...) so my first priority is to provide and efficient and practical interface with equal access to both text and GUI users.
I figure i'll think about this more over the next while (seeing as i don't have much time for non-work programming these days anyway...) So in any case if you have any input (i'm not ready to start this yet, but i would like more input for my brain to stew in) feel free to email. Just to get a guideline, of all the GUI models that i've seen, the best (conceptually, the implementation could have used a little help) was on the Amiga.
---
Play Six Pack Man. I
That is entirely up to the developers. As someone on their way to releasing a GPL'ed music program for Linux, the last thing I want is to make the interface suck. There will be copious reminders to the effect that if you want something changed or have a feature/interface suggestion TO PLEASE MAIL THE AUTHOR. Just because the users in my case won't be programmers doesn't mean they can't contribute meaningfully to a project. I think many people are scared that if they can't program, they can't help an OSS project. This is not true: there's nothing more motivating to a developer than receiving lots of e-mails that say "great program but it would be nice if I could use the function keys to switch between windows."
For applications there are plenty of strategies to improve UIs:
--- Hot Shot City is particularly good.
It seems a lot of you have missed the point about the question posed by the poster...the question isn't "What should the Linux GUI be like?" but rather "What should we change about the Open-Source model to allow user feedback on user interfaces?"
;) [/flamebait]
:)
Actually, I think the article should've been posted as an "Ask Slashdot" instead of being a "news item" (which has been discussed before, as someone has already pointed out).
It is a good question. From what I have seen, some open-source projects openly discourage regular users from submitting suggestions on user interface improvements [flamebait] (you GNOME/KDE/Linux zealots have only yourselves to blame...
Other OS projects are structured such that only developers/programmers have any say as to the design of the system. Obscure mailing lists with awkward majordomo commands certainly make it difficult for the average joe to participate, and it doesn't help when the rest of the participants are developers who have no sympathy for the poor guy because he can't "hack it".
I'm hoping great projects such as SourceForge(is that right?) stuff will open the communication lines, and blast the doors held closed by these "open source is for everybody, but to heck with the user" hypocrites.
Open source has to be truly open-source for everyone...not just the programmers, who, by necessity, have a very narrow few of the system they are designing. In fact, when you allow the people who will actually use the program in, we may actually get around this "the source is everything" mentality and gravitate to an "Open Design Project", where everyone works together ont he specification, design, AND coding of the system.
Too allow the people without the programming expertise to participate, you'll need to add an extra cycle or two to the phases of an Open Source project...requirements gathering (you must ask what users want) and requirements verification (is this what the user wants?).
Excuse my ramblings...I just think that the term Open Source is just too limiting.
If you can make sense of this, paraphrase it so that my befuddled brain will know what I am talking about.
I think you're right on. "Intuitive" is one of those catch phrases like "Seamlessly Integrated" that should be taken out and shot.
In some rare cases, these phrases may have utility, but by in large you're just asking for trouble if you let them pass your lips.
Take the typical desktop metaphor. It is not "intuitive". It is better than intuitive. First I haul a docuent out of storage (which can be made fairly intuitive). It floats around on the screen, but i can "minimize" or "hide" it and bring it back; this is not intuitive, but it's very useful. In most X based GUIs, I can "pin" a document window to the screen (not intuitive, but useful). WHen I'm done, I close the window (how did a document become a window? Is that intuitive?) and it goes right back where it belongs, I don't have to root around for the right place to put it back. That's not something somebody unfamiliar with computers would expect.
That doesn't mean that programs that are unnecessarily hard to use get off the hook though. Probably every program ever written could be better, but the problem is that making life bearable (much less "intuitive") for Aunt Millie is just not an itch most programmers seem to have.
Post may contain irony: discontinue use if experiencing mood swings, nausea or elevated blood pressure.
- Know the audience.
- Know what the audience will want to do in the software.
- Know how the audience is going to want to do it in the software.
- Lastly, decide how to code it.
People decide whether or not to use something based on it's interface. So design the interface first. That will make them use the software. Then, if you've done that job properly, the way you code the software will determine whether or not people LIKE it, and continue to use it.In commercial environments, it is often the task of Psychologists to evaluate man-machine interfaces and user interaction with software. The topic of software usability is indeed not only a part of CS education but of Psychology education, as well, because many important psychological factors are involved, e.g. perception, motivation, learning etc.
As CS students / graduates seem to contribute much to OSS, why not try to cooperate with Psychology students in order to establish GUI / CLI evaluation projects. The students could work on real world problems, the user-friendliness of the software might benefit.
As mentioned above, this seems to work in departments of CS and Physics, so it might be worth trying.
-- PsyBorg
A "clueless" user asks some questions in a forum (Usenet/slashdot/irc/...) and get RTFM'ed before he has a chance of figuring out, where to get The Fine Manual. When he asks how to make apache do this or that people reply "vi access.conf" and expect him to in a flash of light be totally sure what to do.
A "clueless" user who is not unixified yet asks: "Why is X configured so many places?", "Why is there an stty, a termcap, a keyb.conf, an xkeymap, a resource file floating all over?" or another finger-on-sore-spot question and immidiately is flamed as Microserf of Gates lover or other verbal abuse.
A "clueless" user has a need for a program to do this or that. He is not a programmer but quite knowledgable in his field of profession, and poses an idea for an OSS project. Some replies that "If he wants this $&?! program he can do it him self!", and some gives some suggestions under the parole "if you don't program, you don't know what you are talking about."
. . .
My girlfriend is a fashion modeler and uses programs under another OS to make models for sewers. She don't know a thing about programming but all about clothes and asked me, how such a modeling program would come to be in an OSS world. She is the "clueless" user of computers, but the skilled modeler, who desires to contribute to the OSS with her knowledge if she can convince a programmer to transform this knowledge into programs.
If feedback an input from users are weighted solemnly on the user's computer skills, the OSS community will never go beyond making programming tools.
:-) = I am happy
:^) = I am happy with my big nose
C:\> = I am happy with my OS
A good example is anything which contains a lot of text. You have to say so much, that typing is much much faster.
you raise some good points:
Actually, this is exactly what Microsoft has done as well. In order to earn the "Windows 95 Compatible" logo for your software, it must meat specific interface requriements. Things such as what happens when you hit the tab button, and how custom dialog boxes should be laid out. One this about Windows programs is that, by and large, they are extremely consistent, even between vendors.
I don't feel that this is necessarily the way that the open source community needs to go. I don't want my KDE programs to act like my GNOME programs, and neither will act like emacs, no matter what we try to do (although many emacs movement keybindings work inside Netscape dialog boxes...).
Yes! Exactly. But now that you and I have recognized that, what do we do...?
darren
Cthulhu for President!
(darren)
This is _not_ a troll, but nor am I a karma whore either. In all seriousness, we should look to MS-Windows, specifically 95/NT for an easy simple, intuitive user interface. My son learnt it at age 5 without much instruction and he's now better than me.
Microsoft deserves credit where it is due. No-one doubts their marketing prowess, or ruthlessness either. I believe they have put considerable effort into UI design via expensive user panels and focus groups. They certainly haven't put effective effort into OS design, which has become recursive layers of kludge frequently going random. But there is no reason that we should not incorporate User Interface findings to the extent possible. Furthermore, it would ease transition.
To the specific question about OS incorporating user feedback, as a modest OS developer (cpuburn) I feel particularly qualified to comment. I get some user feedback, and if it seems like a good idea to me, I will incorporate it.
But if I don't like it, forget it, and I may be capricious. One MS-Windows user wanted start & stop buttons added. I firmly declined, even though I conceed it will improve ease-of-use. I prefer the user learn about TaskMan. If he wants buttons, let him add them. The code _is_ GPL.
Many posters are making a fundamental mistake in thinking that ease of use is something that's needed for "average folks" (i.e. grandma and grandpa). Thing is, tech-heads want ease of use too. Maybe 3r33t MP3 collektorz think that arcane is better, but most intelligent computer users and coders would rather have something that's easier to use as well. We can only fool ourselves for so long into thinking that 1970s sysadmin tech is where it's at.
This is slightly tangential, but something I think bears pointing out. Have you noticed that whenever OSS enthusiasts talk about bringing Linux/GNOME/Pick your favorite project to the "average user" they always discuss it in terms of making the product mentally accessible to some elderly female relation? I won't stress this so far as to suggest it's sexist that we never talk about "software your dad could use" but, to co-opt a phrase from a Dilbert cartoon: "My mom earned a PhD in Molecular Botany. What does your mom do?" (to which the PHB responded "leave my mom out of this" :-)
:)
The person you're trying to reach isn't necessarily female, and not necessarily a fool. Everyone can benefit from good UI design, otherwise this would all be wasted effort. Remember the theory from college "make the common case fast"? Well, this is fairly similar: "make the common case obvious". Advanced bells, hooks, and whistles are wonderful, but their prominence in your design should befit the frequency of their use. That's one thing I think we generally do pretty well at Microsoft, although it sometimes requires a lot of bitching and moaning before the natural instincts of our devs for obfuscation and tech talk can be overcome
-konstant
Yes! We are all individuals! I'm not!
-konstant
Yes! We are all individuals! I'm not!
Ah, I had to go back and reread that. I thought at first that you were talking about X (the window system) crashing for no apparent reason. Which, unfortunately, happens. Then you mentioned .dlls being responsible, and I did a double-take.
--
Win dain a lotica, en vai tu ri silota
Perhaps it would make the metaphor better if this hypothetical Star Trek computer only understood the Klingon language.
--
Win dain a lotica, en vai tu ri silota
I have no problem with people having access to anything they want, but it is *logically impossible* to make the interface simpler than the task it performs.
<br><br>
THIS IS THE PROBLEM<br>
<br>
There are two different kinds of complexity here. There is the complexity of how a given task is performed, and the complexity of what the end result is. I don't believe that these are in any way inherently correlated to each other. Complicated processes can produce simple results, and vice versa. <br>
<br>
In the first case, the user needs to be able to express what they want the final state of the machine to be, rather than tell the machine how to get there. In the second case, it's going to be simpler just to tell the machine what to do, and let it trust you that these steps are towards some desirable end. <br>
<br>
Given the correct tools, many difficult tasks are easy. Given the incorrect tools, many easy tasks are hard.
Trees can't go dancing
So do them a big favor
Pretend dancing stinks!
I took a user interface design class recently, which I thought was rather good. Some of us walked away from the class with an understanding of one simple principle that the instructor never stated explicitly as far as I can remember: No interface is perfect for every user.
He approached it from several different angles. He explained why the programmer, with his intimate knowledge of the capabilities of the software, the data formats, the interfaces, the controls, the options, is the last person in the world who should be defining the interface that a non-programmer user will use. He explained the importance of seeking out representative users and observing what they do rather than just asking them. And he said something that stuck, "The user is always right."
I used his own material to tear apart the all-GUI-all-the-time interface that has become the norm these days. I pointed out that the same things that make user interfaces approachable for the novice, and diminish the memory burden for the occasional user are often impediments for someone who executes the same task hundreds of times a day. If that person is also a programmer, with the skills to take a scripting language or a powerful set of configuration options and work some minor miracles, then the lack of those capabilities is keenly felt. For me, the joy and power of the Unix model of the world is that the interface I use daily can turn my repeated tasks into shell scripts that capture my knowledge of how to do it and relieve me of that burden.
A friend of mine commented in the late 80's that Unix had a lousy user interface. I replied that the shell lacks many of the things that people look for in a user interface, but as a programmer interface it is one of the best ever designed. I likened it to a readable JCL (he was a mainframe systems programmer). There is power, and an assumption that you can figure out what to do with it.
But there is no need for one interface, either a ubiquitous shell or a full-time GUI to be the single interface for an OS. Companies selling Linux to the desktop market can build powerful GUI installers and desktop environments on top of great open source tools. If the scriptability and the configurability is abandoned, Linux will lose the mindshare that moves it forward. It will lose the programmer-users who have come to it because it meets our needs. We can create the environment we want. We have shown the world that we don't have to settle for any software but our own. But giving Mom the GUI desktop that helps her find her e-mail twice a week, visit a couple of e-commerce sites, and do her taxes does not need to change the underlying engine.
Open source is already many things to many people. Everyone who has ever contributed a significant enhancement to an open source project has made his environment unique for a while developing it. The average user doesn't have the time to do that level of tweaking. But it has already been demonstrated that with the right set of usability features, you can sell it to many average users. Desktop open source can simply be open source with the GUI tools that casual users want.
The net will not be what we demand, but what we make it. Build it well.
A totally intuitive interface is impossible. Think about it.. Nearly everything you have done today was a learned behavior. Every human learns a different set of behaviors. Some people can grasp X and fvwm easily; it meshes with some of the behaviors they have learned. Others still have trouble managing to right-click in Windows; The interface is at odds with the behaviors they have learned, and it doesn't integrate well. Additionally, different cultures learn different things.
What about a 'driving' interface? Driving a vehicle, be it a bike, or a car, or a ox-cart is a behavior learned by pretty much standard across all cultures. What kind of interface could we build that would confirm to the actions and behaviors learned whilst driving?
.sig: Now legally binding!
Anyway, a compassionate user interface can be built, but it doesn't have much to do with the great pioneering work on Graphical User Interfaces that was done at Stanford Research Institute which was embellished by Xerox PARC and then commercialized in a degraded form by Apple. It has more to do with the way a teaching program figures out where the "student" is coming from, and then adapts its communications to the level at which the "student" is operating. As the "student" becomes more expert, the compassionate system, like a good teacher, raises the level of the conversation to the point that it challenges the "student" -- keeping the game interesting without being frustrating.
It's probably important to have some general sense of the user's frustration level as a primary metric in any compassionate user interface. Since people are fond of the word "annoy" these days, maybe there needs to be a global GUI attribute called "current_user_annoyance" which would be a primary contributor to a cost function in an adaptive algorithm for a compassionate user interface.
Seastead this.
You don't know what you're talking about.
I've never seen a windows app that only allows one way to quit.
Alt-space-close, Atl-f4 are STANDARD on ALL windows applications. It is very rare to find any application that overrides these.
Then there's Alt-F-X which is extremely standard across almost all applications too.
And Alt-G-X is very rare, sure pinball has it, name another though.
Your ignorance (on purpose or not) makes me disregard everything you say.
More like 99.9% of the time for top level applications (apps that are on the taskbar).
The X(close) button on the top right of a window is also there 99.9% of the time.
I find it funny how so many unix people here actually complain that windows gives you so many ways to do things, so what? The average user only needs to know one (most use the close button on the top right), power users can use whatever suits their needs at the time.
Giving heaps of options (ones that don't clutter up the screen) is neither counterintuitive or stupid. I doesn't confuse any users, since it never shows up, and it's not 'hidden' as such since the most obvious ways to close a window are available.
You are completely wrong about my level of understanding. I know exactly what journaling is and I know that you simpyl can't conceive of my idas. I'm telling you that referring to a file by its inode, or its filename, or its date, or its permission bits, is a very stupid idea. This is embedded in the filesystem. It needs changed. This is not one of those tried and true ideas that has stood the test of time. All the code to multitask and prevent deadlock, keep that, it works. A lot of the code can be kept, but the filesystem absolutely needs changed. Why? because there won't BE files any more. Like I said in my post, these ideas are *NOT* meant to be an application that runs on the system, it is meant to be THE system. If you implemented something like this on any filesystem that exists today, it would be so messy, hacked, and slow that people wouldn't even begin to conceive of the benefits. And its not just "one guy" that disliked it, its millions and millions of people. You know, the 99% of computer users that aren't using Linux. Those people. They don't want to have to learn a new language where "cp" means copy. In my system, there are no copies. There are no files. There is no hierarchy. How you could even consider keeping any of Linux when doing this (BTW I recommend doing this to Linux because its easy to mess with and much more flexible since there is no megacorp backing it, if Linux doesn't want to even try to explore something new because they love their dear, holy, obtuse and flat out BAD user interface, fine, but don't be surprised when Microsoft scoops you again and you find yourself copying to catch up and never once in your entire existence innovating anything)? (wow, big parenthetical comment ;) What're you going to do with md when there are no directories? What are you going to do with fsck when there are no blocks or nodes to order? This is a different system, it is not an iterative change. It is similar to the guy who prosed getting rid of all the command line stuff and going to a graphical interface. No one was smart enough then to do it right so no one won overwhelmingly (Xerox dropped the ball completely, Apple priced it too high, Microsoft had already won with DOS). Want Linux to ever win overwhelingly? Then stop doing the Linux that isn't winning and do something different. If the product that gets developed is no good, so what? its just a hobby, right? Your time is worthless, right? Those are the tenets of Open Source, you try it and a lot of people get to pick it apart. In the end, the user is the one that matters, because the progammers its just his worthless hobby time getting eaten up.
Esperandi
But its okay, put your blinders on, keep walking towards the cliff.
My point here was that in a situation where, to move something up, you can move a "handle" either down or up, moving the handle in the same direction will probably seem more "intuitive" to most people. In mechanical devices (think of a bow's rudder) often you see a direct linkage which helps you understand the relative movements. On-screen you don't have that sort of aid.
Also, of course, people can get used to anything - even to a command line :-). We're trying to find out how people who've never seen a particular widget before are likely to try to use it the first time. I agree there are many cases where there might not be any huge advantage one way or the other, but anything helps.
Disclaimer: I am a Linux user. I like Linux. I have issues with Linux. Deal with it. Here goes.
Maybe this isn't obvious to me, but why is EaseOfUse != Powerful when it comes to software interfaces? I just don't see the problem here, except that people are too lazy or too time-pressured to develop a decent interface.
For my (only half-decent) example I'm going to use the software that was written for my scanner, a Umax 610P. I love my scanner interface. (Kudos to whoever wrote it if you are reading this.) The interface has two modes, easy and expert. This is a compromise, but one that works well in this case. The easy mode is basically two buttons, Preview and Scan, with a simple, easy interface that allows you to crop what is actually scanned. The Expert Mode is a well-laid out, very simple interface that presents all of the options that the scanner can handle. It is clear, simple, and obvious, even to people who don't know anything about imaging. The only problem is that this software isn't Open Source (AFAIK), which is the only thing lacking in an otherwise excellent piece of software.
The reason I like this software so much is that it takes care of two of the three levels of user, one of which I happen to belong to. The three levels are:
NovicePower-User
Professional/Developer
In this case, I fall into the Power-User category, someone who needs to use most of the features of the scanner but who doesn't have the time to program his own interface, etc. Now the deal with this is that EVERYONE , at one time or another, falls into all three categories. NO ONE can be a professional and developer on every piece of software they use. Therefore, one has to design most software with at least the Power-User in mind.
And that is the crux of the matter. I think that a lot of programmers hold the view that if someone is too lazy, doesn't have time, has no desire, or is too dumb to understand a complicated interface shouldn't be using their product in the first place. Now, before I get flamed, there is a place for this sentiment, and it is a valid view. But don't let it blind you to the benefits of making an easy-to-learn, powerful interface that more people can grasp. After all, we're trying to benefit everyone, that is the thrust behind Open Source. And if you develop for the Power-User, i.e. other developers who must use the product but don't have time to deal with learning the ins and outs of the product, they will bless you, and Linux can continue in its inexorable march to world domination
Linux is not widely accepted on the desktop because it is a PAIN IN THE ASS to learn. The help files suck, particularly for people who are just learning the interface. Microsoft's help files suck for the opposite reason -- they don't assist the knowledgeable user with difficult tasks. The Linux and Microsoft command lines makes perfect sense -- if you know the commands, which have to be memorized. Editing a text file is easy -- if you know the format and how to lay out the file. If you don't, you have to figure it out from the pathetic documentation. This is possible for a smart person (my first experience with Linux was setting up the XF86Config file, ooh fun), but is time-intensive and frustrating. Editing the Registry is just hopeless, due entirely to the crappy interface (can you say...weak?). The GNOME desktop interface is counter-intuitive to a Microsoft-trained user (which is most people), but on top of that it does things that make no logical sense. I have yet to figure out the Alt-Tab sequence of window-switching, as it appears to be random (Just switch 1-2-3-4 guys!). It probably makes sense from the viewpoint of the code, but it doesn't make sense to my eyes and hands, which is what matters when I'm using it.
You know, it seems to me that the rules of designing a good interface are simple:
1)Don't mess with what the user is doing. Microsoft is particularly bad about this (ever had your start bar disappear when another app refreshes the desktop?).
2)Keep things consistent and logical, at least within your program. If your program interfaces with the desktop (Explorer, KDE, GNOME) then stick to the conventions. You guys who write the conventions, make them make sense!
3)Don't hide functionality from your user. Everything should be simply laid out, in a logical format, accessible by both mouse and keyboard. Everything configurable in your software should be here, not in some random registry setting or text file -- that kind of editing should be reserved for software modification or emergencies.
4)Don't present everything to your user at once all smashed together (Exchange 5.5 config menus, anyone?). Logical design, dammit! You're a programmer -- if you don't understand logic, just give it up now. It isn't easy, I know, and programmers are naturally lazy -- so design the interface to be as fast and as easy as possible, which means paying attention to learnability, layout, and keyboard access.
5)Talk to your users -- all of them, from fellow developers to the Joe off the street. They may have crackpot ideas, they may have good ideas, but they are the reason you are there in the first place. Get out there and talk to them. One user's comments are waaaaaayyyy better than none, and the more you get, the better off you are.
Lack of useability is YOUR fault, not the user. Don't go down the Microsoft path of denying users power to make things easier. Fast, easy, and powerful IS possible in my opinion, you just have to work at it. Learn! And remember, you may be using someone else's software tomorrow, and you'll be wishing that they had spent the time to give you a good experience...
AetiusWhy the heck does not more developers mix GUI and text?
*Store* everything in editable format (preferably XML style). Easily accessible for the admin. Throw a GUI shell over it. Make some labeled drop downs and checkboxen to explain the effect of the settings and to provide a non-expert way to fiddle with the file. Make damn sure that the connection between GUI and text is documented.
My favorite windows text editor (Kedit) is a good example of how to mix GUI and text. Every setting, keypress and menu command has a command line equivalent. My hands seldom leave the keyboard, since it is much faster for me to do a command line "c/foo/bar/all *" then to fire up a searh and replace dialog and type "foo" in one edit space, "bar" in another and check the appropriate options (every line, every instance on each line). However, a newbie would certainly use the menu.
Hey, has anyone defined a DTD for .ini-files? After that it would be rather easy to make a generalized .ini-file GUI.
(or has that been done allready?)
All opinions are my own - until criticized
>You mean like dragging a floppy disk to the
>trashcan to eject it?
No, no, no, no... this can be easily explained.
When Apple designed MacOS, they thought of the desktop as a temporary repository for files and folders. In that vein, they created a "Put Away" command...
You can drag a file or folder to the desktop, work with it as long as you want, and then select "Put Away" (command-Y) It will go back where it started.
Now consider a floppy disk... a floppy disk is a physical item, so putting it away would be putting it back in the drawer, or the box, or whatever. If you select a floppy disk, and then choose "Put Away", it will eject the disk.
The Eject command also ejects the disk... *but* it retains the contents in RAM, "keeps it mounted", so to speak. This is because early Macs had one floppy drive, making copying from one disk to another a complicated procedure.
Anyway... point is, when harddrives appeared, no one remembered/knew the function of the Put Away (which always seemed unclear anyway...), and Apple added the trash can 'shortcut.' Now the problem is that the trash can represents "delete" in all cases *except* a disk.
But, hey, the rest of the GUI kicks ass, huh?
(BTW, Apple also changed the Eject function now to "Put Away" by default, unless you choose "Eject and Stay Behind.", the keyboard shortcut for which I forget at the moment. Contextual menus also solve this in a nifty, neat fashion.)
Comment of the year
I agree that's an important reason, but not the only reason.
IMHO, the problem comes from the facts that
very good programmers and experts in ergonomics are two very different jobs
creating an interface is more an artistic work than a programming work : here again, artists (designers, graphists, colorists) are more competent than pure programmers.
The non-open source firms pay experts in ergonomics and artists to make their products more attractive for the end-user. There's nothing like this in the open source circuit. Usually, the same guy works on the code, the interface, icons design, etc. The exceptions are rare.
Now, I don't see it as a flaw in the open source model : artists and experts in ergonomics can work together with the programmer and open source their work too. But for a mysterious reason, it just doesn't happen... Hey ! It's not too late !
Stéphane
Instant Karma's gonna get you, Gonna knock you right on the head (John Lennon, 1970)
First, I don't think that "everybody expects" what you think they do...I certainly don't.
/dev/sda1 and a /dev/sda2 and so on into infinity, "just in case" I have something as that device with n partitions? Plus, then we must throw away this wonderful new invention, devfs, since if we are going to use symlinks and have every single possible device magically mounted for us, we certainly can't have /dev only listing the devices that are in use at the time.
/usr/local //bar.umbc.edu/public /samba -o username autechre
Also, what would you have it do if there are multiple partitions on the drive? And they are different filesystems? Symbolic links...I suppose then we must have a
If you want the behaviour of the OS syncing the disk before it is ejected, you must:
A: use a syncronous filesystem, sacrificing speed.
B: replace PC floppies with something like Mac floppies. PC floppy hardware can NOT do that, unless you want to do A (I don't).
Lastly, please explain how I can do the following using your symlink theory; I'm very interested:
mount -t nfs foo.umbc.edu:/usr/local
smbmount
WMBC freeform/independent online radio.
Both are in continued steep declines. This is the usual pattern after a speculative bubble. The real question is where they'll bottom out. Both are unprofitable, so none of the usual financial metrics apply. The theoretical bottom for a money-losing company is zero. Live by the hype, die by the hype.
The insiders can cash out after six months now. The restricted period used to be two years, so the company actually had to become a success for the insiders to profit. No longer. Watch the insider trading reports. So far, nobody can sell, but wait.
Note: I have no holdings or short interests in either company.
Right.
This can't be resolved by making interfaces intuitive, any more than we can make graduate-level math accessible to children by using "intuitive" words and pictures to describe it. Eventually, we have to accept that part of what we want to do is educate the users a bit more, so they can figure out what they want to do; at this point, the interface can be designed for efficiency. Expert-friendly is the way to go.
Wrong.
This reflects a gross misunderstanding of what user interface design is all about. It's not about graphical user interfaces. It's about allowing low-risk exploration of the options with constructive feedback, so that the user will converge on something that works for them in a short time. Read Alan Cooper for some insight. Or play a good video game, thinking about how you learn how to play it.
Until the open-source community gets this, Linux on the desktop is going to suck.
Without commenting on UI design, I'd like to point out the Word Perfect Corporation had the best user feedback loop I've ever seen in any industry: a 1-800 phone number you could call as often as you liked that was answered by knowledgeable person EVERY SINGLE TIME. Frankly, a copy of Word Perfect was worth $495 for that alone. For pure OSS, a the economics of a 1-800 number and staff are probably unworkable, but I think this is an area where packagers and distributers could make some real money. I WANT to buy the $495 copy of Slackware, and so does my mother. I WANT to recieve ONE defined piece of software for my $495 (not 2000 pieces of freeware on 7 CDs). The reason I want this is so that when I have trouble with that one piece of software, I can pick up the phone, call 1-800-GET-HELP and have someone help me. By providing an out-of-band feedback channel, so that users having trouble with thier computer can report that trouble without having to use the computer, a telephone number provides immediate access to help no matter what the problem is. By limiting the software in the box, it becomes reasonable to expect real, in depth support. After all, is it reasonable to expect Red Hat to provide unlimited, toll free tech support for evey piece of software in thier distribution? No. No one company could understand all of that code in depth. If their distribution were OS-only (think NetBSD, not Win2K) could they do it? Yes, I bet they could. How did I get from OSS UI feedback to toll-free tech support? Simple: a distributor, having packaged OSS and accepted $495 for providing support could and should pass transcripts or recordings or trouble ticket data back to the developers. Developers could focus on developing without having to deal with (often stressed and annoying) users, but still have the benefit of feedback on what needs improvement. Users would have someone to listen to them and help them. Distributors could charge much higher per-copy fees and make more money. And of course, with OSS, if certain "features" were generating too many calls and cutting into profits, the distributer could always hire someone to six them. I think Red Hat is on the right track, but they don't charge enough to provide unlimited support, and the provide too much software in the box to provide in-depth support. The bottom line: Charge me more and give me less. My boss, my users, and yeah, my Mom will thank you for it.
When the rest of the world wakes up and realizes that there will be no real effort to make Open Source software useable for the masses they will take there money and go play elsewhere.
all persons, living and dead, are purely coincidental. - Kurt Vonnegut
Personally, I don't care about user-friendliness, since the command line is quite intuitive for me. It's because I learned to use it.
I know that not everyone wants to learn how to use a system. Linux developers can go ahead and try to please those who don't really get into their systems, I don't mind.
My concern is this: do I still get the old interface? Think about it. It's already happened. I'm using X right now, and it's hell, but there's apps for it that I need to use, ones that would otherwise be developed for the CLI that I love.
---
script-fu: hash bang slash bin bash
[ approaching AI ]
[buzzzz..] A: What does that mean?
B: That's a dial tone. It means you can punch in the phone number now.
[rwxr-xr-x] A: What does that mean?
B: Those are the permissions on that file. It means that the owner can read, write, and execute the file, but other users can only read and execute it.
The rest is left as an exercise to the reader. Intuitive is a bogus concept.
The more things change, the more they stay insane...
I long for the day I can sit in front of my computer and say:
"Computer, what do I have on the books for today?"
or,
"Computer, obtain and install all of the current patches and revisions for yourself. Give me a report when you are done."
or my favourite,
"Computer tell me why this scanner isn't working."
A good UI is the lack of it. When it comes to a user interface, less is definately more. Most users want to be able to use their computers without having to jump through hoops. They care not for complicated menus, or cluttered toolbars, jumbles of icons, and they certainly aren't literate enough for a command line.
The day the end-user can talk to his computer like he talks to his co-workers is the day that that the UI battle is won.
Bibo Ergo Sum.
First of all, we need to get over this idea that power and ease of use are tradeoffs. They're not, if you have a good UI designer at the helm. Unfortunately, I don't think we're ever going to convince the hard-liners of that. Why, I don't know; sometimes I wonder if it's nothing but simple fear.
Second, we all know that UI programming can be a dreary task. This is where graphical UI designers (like Glade for Gnome or Constructor on MacOS) come into play. Using these tools the UI design is far less boring (or time-consuming for that matter). It's even almost fun. Once we have programmers not dreading UI programming, they'll put more effort into it.
Third, UI guidelines for apps. It would do a lot of developers good to study the Mac, BeOS, and NeXT guidelines. These people spent a lot of money coming up with their UI's, and each has some very good points. From these, a coherent set of Linux UI guidelines could be created. There will always be anti-usability zealots who insist "no one's gonna tell me how to design my program!" but the fact remains, UI designs are there for a reason.
Finally, the metaphor. The more I think about it, the more I think the desktop metaphor needs to go. It's not appropriate on Unix-esque filesystems, because of the almost maddening complexity. It's not even appropriate on Windoze. MacOS only gets away with it anymore because so few people go near the System Folder. Unix also has the additional problem that filesystems are a lot more "fixed" in nature than MacOS; it's a lot easier to break something just by moving it. So basically, what we need is a new metaphor, one which is more appropriate given the filesystem. I don't know what a good one would be, though I've heard some intriguing stuff about using a tree before. I'm not sure I like it, but it's an idea, and those are what we need.
One last thing. People talk about just "helping people to get over the learning curve." They talk about a learning curve being a "one-time event." The fact is, programmers have a responsibility to the users to make that learning curve as small as possible. And learning curves are not one-time events, not if every UI you come across is different. That's why there are UI guidelines; if most programs look and feel pretty similar (given their purposes), then a very large chunk of the learning curve becomes a one-time event. Think about it this way: because of the Mac UI guidelines, I can pick up the fundamentals of almost any Mac program immediately, without ever so much as glancing at a manual (at the absolute least I'll be able to quit the program without looking at a manpage, which is more than I can say for vi or emacs). In about ten minutes, I'll probably have the basics down. And if I want to learn more about the program, I still can; people have this idea that the Mac UI is somehow limiting because of its guidelines. But the guidelines aren't what makes it limiting. It's the people who make the software.
That is only true for existing Windows users. Granted, there's a lot of those. But there's also a lot of people who have not used a computer at all, or have been exposed to other interfaces in addition to Windows. For those people, a Win9x-like UI (especially the start button) is going to be very strange. IMHO, the older Win3/Mac/OS2/Amiga/ST/etc.. interface where you click on icons, is a lot easier to grasp than that hidden menu where you have to do the "start" thing, and then make a menu selection, before you can see what programs are available.
If you think the Windows UI is good for novices, I suspect the "novices" you interviewed were actually experienced Windows users.
---
As copyright owner of this comment, I authorize everyone to defeat any technological measure which limits access to it.
Ummm... emmett? Maybe you need to read slashdot a little more? This was posted just over a month ago by CmdrTaco under the title Open Source's Achilles Heel. (Originally submitted by Tony Shepp).
--
--
"In Cyberspace, no one can hear you be sarcastic"
Windows is not intuitive. More specifically, most applications written for Windows are not intuitive. They suffer the same problem that most Open Source applications suffer from: some programmer slapped an interface on the front according to how they wrote the program.
This problem is endemic to the software industry. No one (or very few) are willing to spend the time and effort it takes to design a proper interface.
It's hard. I suck at it. Better to have an actual interface designer specialist design the interface, have the programmers write the back end, then stitch the two together.
All this Slashdot bitching and whining about how nothing is intuitive and everything is learned is completely missing the point.
:-)
I'll give you a working definition of intuitive: A feature is intuitive if on the first try it behaves as the user expects it to behave. See? Nothing about genes, nothing about instincts, nothing about learning.
Dragging a sheet-of-paper icon to a waste basket to delete a document is intuitive for me and probably for 90% of the Western world. We know what is a sheet of paper and what is a waste basket. It doesn't matter if we learned this: we have this knowledge by the time we touch the computer. However, say, for our friendly tribesman Mumbo from New Guinea the same action would not be intuitive at all because it's likely he's never seen a wastebasket in his life and doesn't really know what it means.
In other words, intuition is in the eye of the beholder. An interface is intuitive if I can apply my already-acquired, previously-learned knowledge to the new domain.
And, by the way, user interface design is a big and very complicated field. Most Slashdot readers are no more qualified to comment on it than to judge the performance of a surgeon by watching an operation. Of course that never stopped anyone, me included...
Kaa
Kaa
Kaa's Law: In any sufficiently large group of people most are idiots.
Windows is horribly unitutive and relies on so many just plain stupid techniques.
You mean like dragging a floppy disk to the trashcan to eject it?
DrLunch.com The site that tells you what's for lunch!
Ugh. I think I speak for all of us. ;-)
As to the question of "how do we make the GUI more intuitive?": we just need to extend the Open Source model. "All bugs are shallow", right? Think of a user as being one of those "solitary" developers, and set up a way for them to feedback into the system. Perhaps Linus can require that all GUI code come with a mechanism -- say an HTML page that includes a mail form -- to give feedback to the developers of that GUI. Or if the developers are loath to embed their email address, perhaps a forwarding box could be set up somewhere *expressly* for the purpose of receiving and routing feedback. The mail form that comes with the GUI could come with a Subject line already set up (or with some hidden info that encodes what the subject is), so that once the message arrives at the forwarding box, it could easily be sent on to the proper people. Thus the feedback loop is closed and it's a win-win situation for both developers and users.
DNA is a Turing machine. You, however, being dynamic and emergent, are not.
A good, intuitive user interface is not only totally relative, it also depends on the past experiences of the user in general. I, for example, find a command line interface intuitive, but I've been using Unix for almost 10 years. My parents, who have been using PC's for a few years now, find the Windows95 model (start menu, folder icons for directories, etc) intuitive. Some former co-workers used Macs exclusively for a dozen years--guess what they liked best?
But everyone knows this already. This is not news. But does matter.
We can try to find the common themes here -- if there are, in fact, any common themes to be found when you compare bash to the Finder -- but that will probably not be very fruitful.
We need a huge focus group to discuss user interfaces. We need to get the opinions of people who live in different countries and who speak different languages. We need to get the opinions of people with backgrounds in multiple OSes. So why doesn't someone put together an open source "Research" project that attempts to do precisely this? A site like slashdot would be ideal -- it encourages interaction, posting your own comments and feedback is easy and convenient, and it is widely read.
How does this strike people? Is this a viable things to do? I know this is one of those areas, like documentation, where many open source developers and users are traditionally lax. This is unfortunate -- it is one of the more important aspects of any program, open or closed source.
darren
Cthulhu for President!
(darren)
Well, if you want to get good user feedback about how intuitive an interface is whether its Open Source or not, you use the same method. You find someone who has never used any computer of any kind before and you have them use it and listen to their responses.
The problem with this in the Linux realm is that when a grandma sits down and comes back the next day (you can NOT help them in any way shape or form if you're trying to determine how intuitive it is, else it becomes a test of how good they are at listening to you) and tells you she absolutely hates everything about Linux, the testers tend to slay the grandmother in cold blood for speaking out against the holy relic Linux.
This is not a troll, it is a call to reality. if you wish to continue to improve something you need to know when it is time to destroy what you have and start over. Yes, you love your window manager, but to improve it sometimes you have to destroy it and ignore everything you liked about it and just go down a new path. The guy that answered questions on here a week or so ago, the user interface expert, he had some really great ideas. And he hated Linux. So why not take his ideas, destroy everything of Linux that contradicts it, and see if maybe what comes out on the other end is better overall?
Thinking about his ideas, I think it really is a complete tear-down and re-build sort of project. Not even the filesystem can be spared. Journaling? Who gives a flip, in his new concept of using computers (I hate the word paradigm) journaling is stupid. So scrap it. Start over. Quit referring to file by filenames and extensions and a bitfield of permissions. Don't look at those things in a new way, get rid of the old way! Store files based on arbitrary category names, you have a bunch of song files, well, call them song files then. You boot up and you want to look for that song file you have that was made sometime in the 60s by the guy whose name started with B? The machine should not provide this kind of functionality in an application, this should *BE* the system. And no, you can't search by regular expressions, give it up. What about progams and such? Container categories (like song is a category) would cover this, holding the executable stuff. Well, what if someone on the Internet calls their files "music" files instead of "song" files? Easy, when you go to download one of em, it tells you the new file has a "music" flag on it. You can either add "music" as a recognizable trait for files in your system, or create a translation system for your system that will always read "music" as "song". Community standards will emerge, but if they don't it doesn't matter, maps will take care of it. No more systems programming. Everything will be going through this very high level system. You don't open a file, you try to open a class of files or a property within the same container as the program.
Hell, throw in some of Ted Nielsen's ideas which boil down to hyperlinking between portions of a document. Defining document portions shouldn't be that tough and it makes it all possible (I haven't read too deeply into his stuff, but from what I've read this seems as if it would have solved his problems).
For once in your life, I beg of you Open Source world, innovate! Let go of the archaic, obscure Unix past that sucked donkey balls for UI and do something new. You will never (and I mean never) compete for the home desktop unless you do so. If you eventually do, it will only be because you finally became as bloated as Windows. Everyone bitches because Windows ran on top of DOS... hello? X runs on top of Linux, wake up!
Esperandi
Seriously and honestly, this is not a troll, and Open Source developers need to read it.
Unfortunately, intuitive seems to mean a Windows-like OS, for most purposes. Though I don't doubt that, through the sort of studies that Tog does, it's possible to develop truly intuitive OSs, that's not likely anytime soon.
No, what "intutive" really means is "like Windows." Having worked with all 3 major platforms for quite some time now, I've found that what most people really want is a Start button and Explorer as their OS. Disgusting, no?
Obviously, you and I aren't most people. That's why we're reading Slashdot. But trying to get your grandmother to use anything but Windows, which she's used twice, is going to scare her.
So that's how we do it, that's how we make an intuitive GUI. Imitate Windows.
Gosh, that's an awful thing to have to say.
This issue, and this article, have been raised on /. before. IIRC, the result was much heated debate . Wish /. was responsive enough today to let me get a search done. :)
:) Linux is, first and foremost, a hobbyists system. Then, it is a server-side OS. Then, and only then, it is an end-user workstation OS that looks pretty and holds your hand, and comes preconfigured out of the box.
In a nutshell, the point of the debate was: Who is Linux targetted at? The developers of Linux are the users of Linux - the users of Linux become the developers of Linux. This is the way it has been before the IPO of AllThingsLinux.com
I have to ask, has the intention of Linux changed? Is it no longer software of the people, by the people and for the people? Has it become a supplier-consumer relationship?
If it's still the former, then the developers are the users and vice versa, and it's a stupid argument. If the latter is true, and the developers are the Morlocks to the users Eloi, then what Linux is all about is dead.
Linux WAS about solving real problems. Performance, technical issues, doing things 'right' without market pressure. If the focus must shift to 'end-users', and to providing 'unwashed masses' with a comfortable experience, then that goes contrary to the spirit of Linux - at least as I see it.
Let them eat cake, and run Windows and MacOS, I say! If they want to use Linux, they'll have to learn regular expressions.
In fact, out of the box is exactly where this end-user convenience should come from. Let the people making money on Linux distros add that value. They're the ones who depend on a growing user base. "Hey you, in the red hat! Are you listening?"
The core community is it's own user base and doesn't seem to care too much about auto-configuration wizards and user-friendly dumbification. If they did, those features would be there by now. See how it works?
-- What you do today will cost you a day of your life.
I originally tried to put this in my own words, but I just couldn't do it any better than Jakob Nielson, so here's his take on this issue (his emphasis):
"The "paradox of the active user" is a concept introduced by John M. Carroll and Mary Beth Rosson (then at IBM, now at Virginia Tech) to explain a common observation in several user studies done at the IBM User Interface Institute in the early 1980s (later confirmed by many other studies, including my own): Users never read manuals but start using the software immediately. They are motivated to get started and to get their immediate task done: they don't care about the system as such and don't want to spend time up front on getting established, set up, or going through learning packages.
The "paradox of the active user" [PDF,66k] is a paradox because users would save time in the long term by taking some initial time to optimize the system and learn more about it. But that's not how people behave in the real world, so we cannot allow engineers to build products for an idealized rational user when real humans are irrational: we must design for the way users actually behave."
Source: http://www.useit.com/alertbox/ activeuserparadox.html
The paper is old, but still very relevant. It was written before Gooey Tarbabies achieved World Domination. I was really surprised to discover that many of my current user interface issues have actually been thoroughly documented and processes for (potentially) surmounting them outlined.
Why is it that since we've known about this for so long, so little apparent progress has been made?
My short 2bit answer is the evil upgrade treadmill - everybody is so busy preparing for and researching the Next Big Thing, they don't have time to refine and polish the tools already under our noses.
Some things are not intuitive. The only time a user interface is "intuitive" is when its model of what you want to do is the same as your model. Users do not share models very often, so we invent "common" approaches, but these simplified approaches *cannot* model the more-complicated tasks!
Frankly, a lot of the "desktop" market is people who would not be able to administer sendmail if it understood English, because they don't know *what they want it to do*. They just sort of want, you know, the thing where the other thing isn't done unless it's supposed'ta.
This can't be resolved by making interfaces intuitive, any more than we can make graduate-level math accessible to children by using "intuitive" words and pictures to describe it.
Eventually, we have to accept that part of what we want to do is educate the users a bit more, so they can figure out what they want to do; at this point, the interface can be designed for efficiency.
Expert-friendly is the way to go.
My blog: http://www.seebs.net/log/ --- My iPhone/iPad app: http://www.seebs.net/seebsfrac/
He's not saying that we need to imitate windows to create a better GUI; he's saying that that is what we are doing already.
To provide something new, we need to have a more direct link to what the users of a particular application want. This is something that the commercial houses can and do spend a lot of money on (Aqua, anyone) but in the OSS realm, so much effort is spent in just making the thing run, or better yet, play nice with others (not that there's anything wrong with that), that any UI improvements are an afterthough at best.
We are in a position to take the UI paradigm in any direction we want, but this opportunity is raerely being seized.
At least, that's what I got out of it. Anyone else?
--sugarman--
Firstly, "intuitive" is a slippery word. There's "intuitive for power users", "intuitive for somewhat experienced users", and "intuitive for newbies". Some would say the last category is the empty set.
Metaphors are the key. Read John Lawler's 1987 lecture "Metaphors we compute by" for a quick intro on metaphor and metaphors in computing. The situation has unfortunately changed very little in the last 13 years. George Lakoff's book "Philosophy in the Flesh" shows how metaphors actually are the basic working units of the mind, and that all basic metaphors are based on sensory-motor concepts - in fact, sensory-motor neurons very probably do double duty as metaphor processors.
So, as long as one bases GUI metaphors on basic sensory-motor stuff - things like "time is space-like" (progress bars), "nearer means on top" (overlapping windows), and so forth - you have some chance of being more newbie-intuitive. I've used a prototype ancient GUI where scrollbars worked the other way around; the thumb moved down to scroll up for some reason the designer considered compelling, and it was pure hell to use. Needless to say it didn't survive.
Now, something I've noticed in most (thankfully not all) replies here is a restricted understanding of what a GUI should do. Yes, having icons represent files is useful; installing by running a single program and marking off some options is useful; using menus is often useful. But that's only a small part.
I've released an application for the Mac OS recently. As long as one uses the standard system calls, one gets the expected GUI functionality for basic items - that is, menus work like a Mac user thinks they should, windows drag, roll up, close and whatever. But I spent very little time on that - thanks to a neat C++ framework called PowerPlant - and spent much time making sure other things worked as Mac users are used to.
For instance, placement of the exact hot spot in cursors is important. Shift-click selection of long sections of text is important. Exact timing and graphical progression of drag-and-drop is important. Wording, defaults and back-out options in dialog boxes, sound cues, selection behavior - the list is very long. And the extra time spent on subtleties was rewarded by a five-star review where the reviewer said "I love using this program, but it's often hard to say why unless I watch very closely".
When I'm using CAD software I'm often forced to use Windows NT, and frankly, it's terrible. People argument there are Microsoft UI guidelines, but it seems most people either don't know them or think they're unimportant. For instance, in the PCB layout program I'm using right now, if you select a library component to place on the layout, you select the component from a list - OK, the cursor changes to a translucent image of the component - also OK, but if you want to click on the scroll bar to place the component somewhere else, the component is placed _under_ the scroll bar with no visual indication that the scroll bars are inactive! If you click on a tool palette the component is placed under the palette! If people don't think this sort of flagrant misbehavior is important, are they likely to worry about more subtle things? No way...
So, what the original article says about "skins" being just skin deep is very accurate. Most Linux programmers seem naturally disdainful of graphic interfaces and therefore are slapdash in implementing one. And reaching consensus among a large group about any particular GUI feature is nearly impossible. This is definitely a field where the "absolute dictator" method is the only one which may work - granted it may also fail spectacularly.
The point of OSS (to me) was never to make it easier for anyone to use. It's to make it easier for you to use.
Is there a difference? You bet!
Windows 98 has to be written to the lowest common denomonator of user. Limit configurability and leave that to the programmers. It made it easy for anyone to walk up and use.
OSS (take GNOME, which I'm using now) is highly configrable. My bottom panel is probably guaranteed to not look like anyone else's, and if someone else walked up and used my machine, they'd be lost trying to figure it all out. Heck, people get messed up enough since I have "focus follows mouse" and "auto-raise" on.
Does this make OSS harder to use? There's a learning curve to figure out the interface, but not a whole lot. The concepts between Windows 98 and GNOME are similar, which were mostly developed 20 years ago.
Take another fun OSS example - Emacs. There's a steep learning curve, but those that get past it swear by it and use nothing else. It's probably a dismal failure when looked at in a general usability standpoint, but for each person that customized Emacs to work the way they do, it's a success.
-- Ever notice that fast-burning fuse looks exactly the same as slow-burning fuse? I didn't... (Edgar Montrose)
...ahem...
Command line power
yet some prefer point-and-click
What's intuitive?
.. or isn't it a bit ironic that an article about user-friendliness is hosted on the Web site for Sendmail? Is this the same Sendmail whose configuration file often looks like incoherent line noise? A friend of mine, having once examined my sendmail.cf, commented that it looked like I had written it by pounding my fists randomly on the keyboard. I simply smiled knowingly and nodded, too ashamed to admit that that was exactly how I had configured it ..
What has happened is that Windows has become part of peoples' learned expectations.
Much as with English, people have learned about its warts and aspects of non-intuitiveness, and have figured out how to work around them.
If something is truly intuitive, this means that there is no learning process required. The thing in question simply works the way people expect.
UNIX has the merit that it is a simple enough computer system that some people can get a sufficiently accurate model of its operation so that it becomes possible to predict what it will do.
In contrast, as soon as you strip the veneer of "learned things about its behaviour" off of Windows, it's a richly complex system whose behaviour is much more difficult to predict. (The lack of source code or other disclosures don't help either...)
The above approach to "intuiting" about Windows and UNIX take a rather different tack than the usual, assuming that the individual will actually try to deeply understand the system. It replaces the "black box" with one that is expected to be understood.
Rather unlike the usual model of trying to know nothing about what is going on behind the curtain...
If you're not part of the solution, you're part of the precipitate.
This is really the oldest problem in the free software (Open Source/label of choice) community. There are authors that elicit and get feedback, but they are ususally getting feedback from other programmers, so it's just an inbred feedback loop.
People growl at the thought of having to edit a text file to make an adjustment or configuration. Geeks say "Awesome - text file" and whine to people to just learn how to do it.
Teching the masses is not the point and will never be the point. The masses will not learn and any software that is predicated on a painful interface will be opverthrown by software that is pretty and easy to use.
It's not what any of us wants to hear, that free software utterly fails to companies like Microsoft that spends millions of dollars on research with people who are not geeks, who are not programmers, who are not even proficient with computers.
Money is not the point. research, forethought, and feedback is the point. There is reason that Apple and Microsft have fixed UI models. They recognized that the biggest weakness of most programmers is intuitive UI design.
Programmers and geeks as a whole are extremely intuitive people, and that intuition allows us to make tremendous leaps of the non-obvious. Most users are not capable of these tremendous leaps and fail to understand. These users are not stupid or even lacking in intuition. One component of intuition is past experience, and if they have not spent a long time around computers then they do not have the necessary reference material.
There are UI guides out there, and there have been some efforts by the GUI people to get coders to follow them, but we all know that they are herding cats because we won't follow anything that somebody else is going to impose. <sarcasm> I don't need those frigging kernel patches that king linus and prince alan keep trying to shove down my throat.
To get back on subject. If we really want to take over the universe (duh) we are going to have to figure out how to make software that is easy to use, friendly, and intuitive to -non-computer users. There is no compromise.
Thank you Redhat, Corel, and all of other distributors that have gone to great lengths to make is easier to install Linux. I mean there are tools around now so that I don't have to manually program my monitors frequencies into XF86config!
enjoy,
chris
-- I need more coffee. It's Monday. There is no such thing as enough coffee on a Monday.
Am I the only one tired by titles of 'It's [whatever], stupid'? This always sounds to me like a somewhat veiled attempt from the author at placing themselves above the reader from the get-go, by claiming they have such a clever bit of information that the reader should feel 'stupid' for not knowing it.
More on topic, I have to say there probably isn't such a thing as an intuitive control. If you think, for instance, that a mouse is an intuitive device, you should see 80 year-olds who never touched a computer before try to figure them out.
Rather than considering so-called 'intuitive' controls, the goal should be to develop methods which are built upon existing and well-known ones. Ultimately, nothing is intuitive in controlling a computer (unless we developped Herbert's genetic memory in the last 50 years), but rather part of a slow learning process.
So don't shoot for intuitive; shoot for ease of learning. Mounting and umounting a drive, for instance, isn't intuitive to a Windows user. But it's easy to learn, and once you catch the principle, it's acquired.
Is that more on topic?
A common misconception is that Linux GUI's are hard too use. For what the average user does, this is a very false statement.
.ini's are created, and the monster known as the registry is altered.
Most people use only 3-5 of the icons on their Windows desktop, without ever delving into any of the menus. AOL, Word, Internet Explorer is all they know.
For most people, it's all they need to know.
So when we talk of usability of Linux GUIs, there is nothing "hard" about starting Linux programs from the GUI. What gives average users trouble is installing the actual program.
For Windows most installers automatically put the icons on the desktop and the user never has to worry about fiddling with any settings.
There is an immense number of technical things going on with the installation of a Windows program, but the user never ever sees it. DLLs are copied,
The same should also work for installing Linux programs. Install scripts which hide the ugly technical stuff from the user, place an icon on the desktop, and thats all.
Instead of a "better" GUI, Linux needs a better install procedere which lets the user click and go.
One possible solution (and feel free to rip this apart as unworkable): Create a consensus document that sets forth interface guidelines, like Apple has for Mac developers. Not to create a rigid dogma, but to provide a starting point for interface design, so that one person or group isn't starting from one end of the spectrum while another group starts at the other. To create a common language for UI design. As for user feed back, well, a web based feedback forum, like /. boards, could allow end users to comment on interface features. This problem is not one that is insurmountable. It simply needs to be recognized.