Slashdot Mirror
GUIs for Everyone
An anonymous submitter writes: "A former Microsoft and Creative Labs interface designer has an interesting diatribe on the approach of Linux GUIs on the desktop. Thomas Krul has three Microsoft patents for human factors research into digital interfaces and graphic software functionality. Probably most known for the interface work he had done on Softimage DS and its web site. Though not a technical read, it does provide an interesting note on the approach for Linux on the desktop." And headless_ringmaster notes that Jef Raskin, the guy who designed the first Macintosh and author of The Humane Interface, has a SourceForge project putting his ideas into action.
When I was in grad school, I did a paper on the Windows interface from an end-user design perspective, and it sucks. Surely there are other ways to handle a GUI that might make sense.
Other people who've weighed in on this subject include prominent researchers like Jpseph Goguen, Terry Winograd, and Eben Moglen.
Right now I'm not proposing a solution, either, but I am working on understanding the problem.
I'm not a geek, I'm just a clever script.
I don't know, I hate flashy.
Here's an example: do "regular users" prefer the new look of WinXP, or the old one? My mother-in-law, in setting up a system for an elderly friend of hers, set the overall system to the Win95 look-and-feel, after I showed her how. She also had the very good idea of clearing off the desktop to a blank background, and putting the icons for 4 or 5 apps right there, so the newbie could avoid the start button altogether for now.
Anyway I hate the new fisher price look, and am grateful that they include the ability to rollback...which of course raises the spectre of using the same GUI for the next couple of decades and becoming an old fogie....
But I don't think the Win95-ish interface is that bad, frankly. The taskbar was actually a throwback to the earliest version of Windows that had the "running programs" all in one place, but that isn't that bad of a thing...running programs should look different from program launchers in my opinion. (That's a mistake I think OSX makes, kind of mixing the two)
Maybe I'm too short sighted about the future of GUIs, but I think th status quo is pretty decent. And for as long as Windows is the dominant desktop, the more Linux acts like it from the UI, the better, since learning new UIs is a pain. (Paradoxically, by making XP look all new and flashy, they may have done Linux a small favor, by opening people to the idea that it doesn't *have* to all look the same as it has for the past 7 years....)
SO YOU'RE GOING TO DIE: The Comic for Dealing with Death
I am a little sick of articles and comments that bash current GUIs for being derivative, without coming out with new ideas. It all fine and good to say that we need something new and exciting, like the GUI was to the commandline, but it hardly does any good to complain about there not being something new if you don't present your ideas on what the new paradigm should be.
The most creative thing I have seen are 3D desktops, but those don't seem to be a major improvement over virtual desktops. I guess the next big thing should be computers that you can converse with(not neccessarily with spoken speech) and just tell to do a job, which would be great if we could do it.
I guess I am just tired of people complaining about WIMP derivatives. If there were better viable ideas out there, we could do them, but I haven't heard any.
If anyone would like to enlighten me as to what the next paradigm should be, I would be happy to encourage and help it's developement, otherwise stop complaining until you have an epiphany.
Spencer Ogden
I think the biggest failing behind Windows (and by implication the Mac that it so blatently stole from) was that it hid the Command Line Interface (or shell if you prefer)
GUIs are well-suited for simple tasks, and are good for the important-task-infrequently-used items, but for items of moderate complexity, nothing beats dropping into a shell.
But by hiding the shell (and making it clunky, as per Windows and DOS) or by removing it entirely (Mac) there is now a huge class of computer users who expect *everything* on the computer to be availible via GUI widgets. The concept of communicating with the computer via a type of language is completely and utterly foreign to them, and is viewed with fear and distrust.
But to ignore the shell is to ignore the greater part of the power of the machine!
It's like all the books in the world were suddenly converted into comic books, and all literature was abandoned. Not that there's anything wrong with a comic book, but they don't deal well with Shakespere or Gibbon.
Celebrate the shell! Bring back the CLI!
DG
Want to learn about race cars? Read my Book
That said, let me address his points: The mistake I see this guy making in his logic is assuming that OSS makes large-scale innovations. In reality, I've noticed that OSS projects tend to borrow a basic framework and when innovate in smaller steps. Linux looks like Unix, KDE and Gnome look like Windows, etc. The difference, of course, is the small changes and nifty add-ons that make any given system more configurable, useful or whatever.
The real strength of OSS is the rate of evolution, not in the ground-up creation. I'm convinced that it takes a small group of well-led, motivated people with an original idea and good planning to make truly structural leap -- think Be. I haven't seen an open source project do this *yet* (not saying it's impossible, however).
So, instead of just doing is shallow-understanding critique of open source development, he should have been discussing a way to allow open source development to make these sorts of large-scale fundemental leaps. That would have been useful.
Every year during my review, I just pray the words "slashdot.org" aren't mentioned.
I think the main isssue here is not the GUI but if your hardware has drivers for Linux and if all of the software that you want to run will run on Linux.
For instance, Photoshop is not available on Linux. Some CAD and 3D software is also not available. Some of the popular games are not available. When you see those things for Linux, you will have popularity on the desktop.
Notice, I didn't mention M$ Office. There are alternatives for that on Linux. When you see Adobe, Autodesk, and others develop for Linux, business will switch due to cost. Then the consumer will switch too.
Here I am reading the article scrolling with my mouse wheel. I get to the bottom of the first column, instead of being required to move the mouse, grab and drag the scroll bar, or repeatedly scroll the wheel back up, he provides a quick link to jump to the top.
Very simple, yet elegant. You don't see things like that often. Small little things like that can greatly improve the end user experience.
Ascalante: Your bride is over 3,000 years old.
Kull: She told me she was 19!
The apps being mixed in the dock is a legacy of NeXT UI principles. They figured that with modern computers that preemptively multitask and with endless virtual memory (theoretically), the user shouldn't care or think about whether a program is running or not, he should simply use the right tool for the right job. There is even less of a difference between running and non-running apps (just a tiny grey ellipsis) in NEXTSTEP. I dig it, and I like the principal.
There should be a moratorium on the use of the apostrophe.
Max V.
NeXTMail/MIME Mail welcome
...for some of Creative Labs's crappy interfaces. I hope he chokes on a "Flash For Dummies" manual. I can see sooooo many errors in usability on his web page it makes me laugh. For the newbee not versed in UI usability, here's one for ya... do you really want a slider that has values 10,20,30,40,50, where the diffeence between the values is a few pixels? how the hell does the user easily adjust to 38? And any reposne of "But they don't need to" is wrong, and although I love vi too, you need to read up on User Interface ***USABILITY*** best practices. Start a book club maybe, and make this guy a member.
Herein lies an important tenet of usability testing, which is Jakob Nielsen's "First Rule of Usability:"
Don't Listen to Users
You may think the Windows interface is OK, but your saying so is no substitute for observing you in action. Chances are--and no offense intended--you probably don't get along as well as you think you do.
And you have to have something to compare it to. When compared with the Macintosh, the Windows GUI is much slower. Just, Ask Tog. Finally, as MaxVlast points out,
- the user shouldn't care or think about whether a program is running or not, he should simply use the right tool for the right job
It goes by many names, but this concept is what Alan Cooper calls "Goal-Directed Design." Design the system so the user can do what they want to do. The underlying technology should be transparent.How about a GUI that acts like a person. You fire up the computer and it starts shooting out information at you, if it's just idling. Perhaps have a little face with a bubble with stuff in it, and maybe use text-to-speech to spit out the words.
"The time is sdfjsdf. This just happened in the news. The following holidays are near. Your hard disk is running low on space. You haven't run this application for a while"
Just random crap... and when you're using an app, the apps can give that person stuff to say... there could be difference classes of information. If you're browsing the web, the computer could be more friendly and informative to you. If you're coding it could just sit there and shut up. If you're writing a paper it could perhaps search the internet in the background for related information and pop up ideas or something.
I'm not sure how to elaborate any further, cause it's just an idea floating around my head, but it could be neat.
SuPz.orG
If I were a billionaire I would buy Trolltech and CUPS and eliminate the last of the objections to their licensing and pay people to stop developing GNOME and to develop KDE or cool new apps on top of Qt and KDE.
What Linux does need is more high quality applications, open source or not.Having to test and integrate an application with two pointlessly competeting desktops is just an annoyance to application development. Adding a 3rd cutting edge GUI isn't going to help. Having an app that integrates with KDE but not GNOME or vice versa is equally annoying.
What is it about desktops and window managers that invoke such passion in geeks anyway, compelling them to continually develop new ones, most of which land in the bit bucket, Enlightenment .vs. Sawfish for example
I really need very little from my desktop or some wizzy next gen replacement. I just want something that launches the applications I use and I want more cool applications to use, not a new GUI paradigm.
I disagree -- they need a way to tell the computer "I want to write a letter," or at another level "I need to get this done so I can go to lunch." That's a very important and fundamental paradigm shift that we as engineers are responsible for assuming; rather than contemplating the nature of "programs" and "processes", the user is best off focusing on accomplishing his/her end goals without even knowing that the concept of a "program" even exists.
I would argue that Win 3.1 and 95wer flashy for their time. As for 98, NT, 2k, they just extended the look of 95, but MS didn't really have to push the boundaries of GUI design with them since they had such a commanding market share.
If Linux is to take off on the desktop, an outstanding GUI is needed to attract attention to it and give everyday users that don't understand the underlying system's benefits a reason to switch. Flashy doesn't need to be non-functioning. The two are not mutually exclusive.
I love these trolls - they complain about the author's lack of any solutions and then don't offer any alternatives themselves.
The author is trying to spur discussion for new interface ideas.
So - I'll try suggesting something.
How about a search-engine based UI?
Here's a use case:
You are presented with a prompt - it says "What do you want to do today?" (ala MS) I enter in "email Bob Johnson about the party on Wednesday". The computer then responds with an email form - it has already entered Bob Johnson's email into the "To" field and has put "Re: the party on Wednesday" into the "Subject" field. The cursor is in the contents of the email with my signature already entered at the bottom and a greeting at the top.
Once I am done - I click on the gigantic "Send" button.
I never even see an email application - just the form to create a single email.
Programs would be installed into a database along with keywords and use cases - this is where the search engine gets all of the info.
Each use case has an associated wizard or application or form for the user to fill out. If the search comes back with more than one entry - it presents the user with the entries so that they can choose.
The web (and various web search engines) could meld with your machine. If the search through the local database turns up nothing, it will go to the web and give you some results.
Other use cases:
-- Burn MP3s
-- Create a picture (opens the GIMP)
-- What's for dinner? (search for a good recipe)
-- What's the weather like? (web search)
-- Troll on Slashdot
-- Buy movie tickets (web search)
-- Write a resume (loads a word processor)
-- Balance the checkbook
The computer will create convenient shortcuts
to the use cases that the user frequents to
further customization.
Applications would be defined by their use cases. That way, when someone talks about a
piece of software, people can discuss the use cases it adds to their system.
Oh well. I just thought I'd offer something instead of whining about the author's lack of solutions. What do you think? Offer constructive criticism - don't just troll.
My idea for a 3D UI is this:
First the input device is a head tracker and a standard wheel mouse. The output device is a 2D (monoptical) or a 3D (bioptical) head mounted display.
Think of a sphere that surrounds your head. That is the new "desktop". The applications are standard 2D applications that we know today. The windows are anchored to the sphere such that their plane is parallel to the tangentental plane at the center of the window. Forground applications are fully in front of background ones.
The mouse moves along the sphere until it visually falls on a window, then it moves in 2D within the window. Grabbing the title bar and dragging moves the center point on the sphere, and thus adjusts the orientation to still be parallel to the tangent.
Holding a special key (alt, maybe) and rolling the wheel expands and contracts the sphere. Holding another key and rolling on an application rescales the application. This is different from dragging the sides and corners as that changes the size the application thinks it has and thus changes layout and may obscure some information on the screen. Rescaling just allows you to make something small that you have to contract the sphere to get it closer to see well.
This design has a few good advantages:
- the user can place applications that are similar to each other close together, so that, for example, looking close to straight forward you have your work applications, while to the left you have websites. Changing context just involves rotating your head.
- the user can place less important applications to the sides. The looking straight forward is the most natural position to be in. Applications that aren't important harder to look at areas. For eample a stock ticker may be above and to the right, and you can check it by glancing there. Also, you can take advantage of the human peripheral system that has been tuned to detect movement over providing clarity. A stock alert that pops up there will be noticed by the user but not interrupt the application they are working on unless they choose to look.
- Because the sphere is actually a 2D surface in 3D, it can use normal 2D tools, such as the mouse, to navigate on. Yet, it still allows the user to arrange things in a 3D space, without actually worrying about how to move in the 3rd dimension.
- Since the user will typically only place windows where they can physcally rotate their head to, the windows all end up being within reach fairly quickly.
- It's natural for humans to interact with the world by standing in one spot and rotating their head. It isn't as natural for us to fly in all 3 directions.
- No changes to existing applications need to be made. They don't have to know they are being projected in a 3D world.
If linux were to have this, I doubt I would ever go back to windows (much like I can't go back to IE because of what it lacks over Mozilla). Now, I can go between them without caring because they aren't very distinct, featurewise.
-no broken link
I used to do exactly that.
Then I would go use someone elses computer and be almost completely lost.
Now I try to maintain a computer that is as near vanilla as possible so as to be able to sit down and use a vanilla machine when presented with one without swearing and cursing or hitting the wrong key/expecting a certain macro to work.
Just a different approach to a different problem.
42 - So long and thanks for all the fish.
This is exactly right. Changing the underlying data representation is the first step to enabling truly new GUIs. As Gelernter says, it simply doesn't make sense to use a 1960's era data model (the hierarchical file system) on 2002 hardware.
Also, while radical approaches like 3DUIs don't make a lot of sense on top of the traditional file/folder storage model, they become much more compelling when the file system becomes a relational database.
And you are absolutely correct that Microsoft is pursuing this opportunity with a vengeance. By battling for the "Windows desktop", most Linux UI developers are fighting yesterday's battle. Instead, they should be looking forward and trying to beat Microsoft to a truly next-generation environment.
Nooface
In Search of the Post-PC Interface
It involves a keyboard and a piece of paper.
I'm being serious.
Want to write something? Pull out a Bluetooth keyboard, and an 8.5 x 11 touch-screen OLED, what I like to call "Bluetooth paper". Start typing on the Bluetooth keyboard, and watch your text appear right on the paper, with quality as good as a laser printer. Or you can dictate it. Or you can handwrite it. It's completely up to you.
Want to check your email? Press a key sequence, or say "email", or write "email", and your email is shown right on the paper. Flip the paper over to see the second page, flip it over again in the same direction to see the next page, flip it in the other direction to go back.
Want to print something? Put the paper near a printer, press a button on the printer, and whatever's on the Bluetooth paper will be printed out on the real paper; a permanent copy.
Want to surf the web? Type in, or handwrite, the URL; the page will load up, viewable on the paper. If you've got another sheet, it can split itself, showing content on one page, and navigation on the other. Touch a link, and it opens up.
Now, tell me you wouldn't want to use an interface like that. The OLEDs and keyboards (of course) are in production today, even if the paper's a bit expensive. All you'd need is a device that would intermediate, that would accept input from whatever source and broadcast the raw pixel data back to the paper. It could be in a hub-like box, in a cellphone, even in a wristband. Anything.
To make it work optimally, you'd need the Bluetooth paper to be a touchscreen. That's not possible yet, but it will be soon; until then, you could use a wireless Bluetooth "remote control", or trackball. Also, you'd need to embed a Bluetooth chip in the OLED; again, if it's not possible today, it will be by this time in 2003.
Revolutionary? Not quite. It's simply making computers more natural. And until what I describe is widely available, we need to make existing computers work more like that. One wonders, why aren't all current desktops running WinCE or Symbian? Both of those OSes are powerful enough to run productivity and email apps, and WinCE is powerful enough to run games, too (if the Dreamcast could use it, so can desktops). Imagine if someone could press the power button on their PC, and have a list of applications come up *instantly*, because the OS is installed in ROM! It might mean multitasking isn't as powerful as it is now, but no users use multitasking anyway; just us geeks, and our boxen are not desktops, but workstations.
So, in the short term, what should we do? Extend the LinuxBIOS project to be a full-featured OS with a Palm-style interface, that can load applications off a hard drive, but caches the most frequently used apps (browser, email, word processor) on flash for fast access. Obviously, X is completely out of the picture; really, gtkfb should be appropriate. Start shipping 64MB flash cards, in USB2, FireWire, and IDE versions, with LinuxBIOS, some GTK launcher applet, Galeon, Balsa, and AbiWord preinstalled; you could charge, say, $150 for the initial device, $20 for future upgrades on CD-ROM (or free download). And make very liberal use of AutoPlay for the CD-ROMs; for example, if someone wanted to play Alpha Centauri, all they need to do is pop in the game, click Install, and *everything* happens for them; in the future, all they need to do is pop in the CD-ROM and it loads. For system upgrades, you pop in the CD, wait for a dialog that says "OK" and ejects the CD, take the disc out, and watch it restart itself.
And better still, we could ship a computer, with a custom mobo (or at least, a mobo with a custom BIOS), that has the whole thing built-in to the computer; so it's even faster than IDE, in fact instantaneous. And that computer could be quite small and cheap. Why? Base it on VIA's VPSD Mini-ITX mobo. Smaller than FlexATX, it clocks in at 17 square centimeters - quite possibly, the world's smallest x86 mobo. It has an embedded processor, and sells for $125 from PriceWatch (including shipping). About the only thing it doesn't have onboard is RAM. You could sell one of these things for cheaper than a Dell, and that's including a 15" flat-panel monitor! As long as it had game support, I imagine lots of people would buy it.
The problem with all the other devices that were like this was that they didn't run standard apps. This box, being a real PC, would run standard apps; it could run most any console or GTK program, even if it required a recompile. The killer app, though, would be games. Sell the box in two editions; regular, and gamer's edition. The game one comes with a GeForce 4 Ti (or the latest card at the time), VGA-to-RCA converter cable, and no monitor.
Sounds like a console? So it is; essentially the Linux version of Xbox. But it can also be used as a regular computer; considering that, it wouldn't cost very much at all, and importantly, neither would the games. No subsidised loss-leaders here.
So, enough of my rambling. Between all these ideas, we should be able to do *something*. So why aren't we?
You mean like in IE, the favorites list appearing in the sideline or such?
multi-row tabs,
Which damn nearly everybody uses, unfortunatly. They ARE good for some uses, but they have to be used
and their latest shennanigan, the adaptive menus that constantly change position on a user
You mean like when I right click some wheres the right click menu appears ALL THE WAY on my screen as opposed to appearing half OFF OF my screen? I always considered that a boon myself.
Or do you mean the frequency of use menus that display items depending on how often (or if at all) they are used?
Quite handy actualy, once they have adapted themselves to the user they rarely change at all, and you can set the time out limit for how long you want something to stick around before it is put into the invisible bin. I find them to be very useful, especialy since I tend to have a few hundred tools for a LARGE variety of purposes installed at any one time, and I may need any one of them at any one time, but the overall chance of me going to any particular tool at some instance is actualy veeeery small. So I do not
It is customizable, or can be turned off compleatly, but even in its default mode it is quite handy and useful, and I am VEEERY picky about my UI elements, the fact that I have let it stay around at all on my machine is a minor miracle.
Oh, and another example of consistency.
The start menu is itself just a directory full of links and other directories. If you give one of those directories the hidden attribute, it will also disappear from the start menu. Very nice. Rather useless, but still very nice.
Oh, and the Interface Hall of Shame is, err, run by somebody who does not realize that the ability to do anything from anywheres is VERY powerful.
- An application uses the common Open File dialog simply to allow you to specify the file to be worked on. Why is it then, that the Open File dialog allows you to rename files, delete files, create new folders, send files to the printer, send a fax, save a file to a floppy disk, try to convert a bitmap file to an Excel spreadsheet, edit a file with a different application, create an e-mail message. and so on, all while the calling application is waiting for the name of the file you want it to work on? This is bizarre!
I use that feature ALL the time, if I made a typo when saving a file (the common save file dialog box has the exact same properties) I can just go to the save as menu again, right click on the file and rename it.Which is a TON more convienent that having to open an explorer window, navigate to the directory, find the file, and THEN rename it, and then close the explorer window and return to my application. w00t. Especialy since the save as dialog automaticaly goes to the directory that a file was last saved to from that program. Yaah.
- One particularly bad design feature of the common file dialogs is that they require the use of a pointing device to access certain functions.
Now that I will agree with, they are HORRID for screen reader users and users are only using the keyboard, ick.Word's open and save boxs are worse though.
Need help treating your acne? Come here!
The "taskbar" introduced the idea that there would be a thing on the desktop representing a window that stayed there whether or not the window was open. Almost all modern interfaces copy this (OS/X and the new Linux ones) and as far as I can tell MicroSoft invented it. Before that everybody, including MicroSoft, thought of "iconization" where a window was replaced by a small representation, and the small representation disappeared when the window was opened.
If anybody can point to prior art where there was something created when the window is created that did not disappear when the window was opened I would like to know, but as far as I can tell this is a real invention by MicroSoft.
I also want to commend them on figuring out that text is much more important than icons, and getting rid of the large icons, especially in the taskbar, and supporting large amounts of text. This was a dramatic reversal from contemporary designs then and they should be commended for it, though I guess it isn't really an innovation.
I believe having Alt+Tab navigate to closed windows (not just opened ones like it did in CDE) is an innovation as well.
I think MicroSoft should be criticized for some stuff that now pollutes Unix and Windows and often is considered "user friendly" so it is impossible to fight them: click to type replacing point to type was very bad. Clicks raising windows completely defeats the whole purpose of overlapping windows. MDI and tiled windows are a horrible abonimation that was created because of the clicks raising windows. Tying all the app windows together so you can't insert non-app ones inbetween also defeats the purpose of overlapping windows. And icons on the desktop (why not in a window that can be raised?)