Slashdot Mirror
The Future & History of the User Interface
An anonymous reader writes "The Mac Observer is taking a look at UI development with lots of video links to some of the latest developments in user interfaces. It also has links to some of the most interesting historical footage of UI developments, here's one of the 1968 NLS demo. From the article: 'Sadly, a great many people in the computer field have a pathetic sense (or rather ignorance) of history. They are pompous and narcissistic enough to ignore the great contributions of past geniuses... It might be time to add a mandatory "History of Computers" class to the computer science curriculum so as to give new practitioners this much needed sense of history.'"
The multi-touch interface demo on Youtube was interesting, I saw it a while ago.
:(
The thing that makes it different is how casual the interaction is compared to file & image programs today. You see the guy just touch the screen and rotate, zoom, and move images around and organize it, instead of opening up dialog boxes, secondary windows, or menus to access the functionality. It's very basic stuff, but you see how powerful it is, kind of like how Google Maps is compared to the old static kind of online maps.
It's like today's image programs are concerned with precicely doing something like zoom to exact levels(%100/%50/%33/etc), but this programs let you do it to "whatever zoom feels right", without worrying you with the details.
Hey speaking of which, I wish cameraphones had a much more fluid interface for picture organization, so I can add keywords, associate it with people on my contacts, etc... but what do they care, as long as they make money off the ringtones
FTA: The current state-of-the-art User Interface (UI) we've been enjoying has remained largely stagnant since the 1980s. The greatest innovation that has been recently released is based on video card layering/buffering techniques like Apple's Expose. But, there is a large change coming. Rev 2 of the UI will be based on multiple gestures and more directly involve human interaction. Apple is clearly working in the area as some of the company's patent filings demonstrate. Nevertheless, these videos might make Mac (and Windows) users experience a huge case of UI envy, as a lot of UI development (in XGL in particular) makes the current Mac UI seem creaky and old fashioned.
The guy seems to think that the stagnation of the UI is an entirely bad thing. It seems to me that when something works well, people like to stick to it. I really don't think the majority of people need multiple desktops floating around let alone a brain interface. The only widely practical new UI technology I saw was multi-touch interactive displays (or touch screens in general, though they have been around for a long time and are still not very popular). As far as his comment that the new-fangled UIs make the Mac seem creaky and old, well, that's his opinion I guess. Some would just say the Mac UI is useful as it is. Even some of the new features in Leopard seem unnecessary to me. It's never bad to innovate, just don't automatically assume every new cool thing is practical or useful for most people.
... and I could stop working and go back to university to get another degree full time and end up into research, where would the state of the art of the UI/human-computer-interaction field be? which degree would one want to pursue? where?
I've always been fascinated by HCI but have yet to be able to pursue this in a work-related setting (where I tend to write backend code, basically as far away from users as you could possibly get).
-- the cake is a lie
Reminds me of Alice in UNIX land. An oldie but a goodie.
What I am still waiting for is multi-pointer capable x11 (two mouses) and pressure-sensing mouse buttons.
Obama likes poor people so much, he wants to make more of them.
Your points are interesting. But they have already been largely mitigated!
1) Your points on overlapping windows is interesting. But KDE already addresses that. When I open a new window in KDE, it opens the new window over the area of greatest unused space. Overlapping continues, but as unobstructively as possible. Contrast that with Windows' means of opening windows about 1/4" below and to the right of the previously opened window, which almost assuredly wastes as much screen real estate as is possible.
2) Can't comment on shortcut keys and internationalization, other than to state that most programs let you customize this.
3) Multitudes of widgets with overlapping functionality. Care to elaborate?
4) "Jumpyness" is more natural! When you flip pages in a book, you go from one page straight to the next. It doesn't "slide", you flip it and that's it. How is page up/down any different?
I have no problem with your religion until you decide it's reason to deprive others of the truth.
Amazing how naturally he uses the mouse -- back in 1968!
Paleotechnologist and connoisseur of pretty shiny things.
On the shitcan of history, like the unreadable choice of default font on Slashdot, the Star Wars Galaxies NGE, the changes to Yahoo's stock message boards, and two recent changes to Google Maps, one of which has made broke printing impossible (users are now reduced to taking goddamn screen captures and printing those!), and and another one that auto zooms and recenters, instead of merely re-centering the map, on double-click, making navigation a time-consuming process of setting a desired zoom level, clicking to recenter, slowly loading a bunch of tiles you don't need, then unzooming back out, and loading yet another set of tiles.
In each of these cases, user feedback was nearly universally negative, and yet the "improvements" remain in place.
If this is UI innovation for Web 2.0, give me Web 1.0 back.
I graduated in 2003 with a BS in Computer Engineering and a BS in Software Engineering.
During my studies I proposed multiple times to do an independent study of the history of the computer field to count for 3 credits of my general electives. I was denied every time, even with support from the head the Engineering department. The liberal arts department continually stated that the purpose of the electives is to gain breath in knowledge. I finally took a (very interesting) class on Greek mythology.
I agree with the premise of increasing knowledge, but not the implementation. The college should encourage independent research when a student can blend his primary interests to meet a "credit based requirement".
What are your thoughts?
Understanding history of your profession should be as important as understanding your culture and your history. Your profession will become a part of who you are as well! Without context, you're clueless.
I looked at those videos, and most of the time the operator is spinning the damn cube, or waving a window.
Now you tell me, how much time are you going to spend doing that? I turned off the XP menu fade-in right from the start.
How about some useful stuff that actually helps organization, like virtual desktops did - a rather simple and cheap trick that didn't require everybody to upgrade to the latest octo-GPU CrossFile SLI-ed system.
I have two more:
1. The gas pump that once you pick up the pump the prices disappear asking you to "Select Product."
2. The ATM that the button that you used to press "Withdrawl" on the next screen would withdraw $200. Shouldn't that go to the smallest amount or a "Go Back" button?
That the biggest UI change yet-to-come has to do with moving from a single-user desktop metaphor to a collaborative virtual space that leverages a lifetime of perception of the real world. When computers evolve into a more transparent role in our life, layering this digital world on our physical world will be next. It's coming sooner than we think, will we survive that long though?
Several years ago I had the delightful privilege of talking about interface design with Jef Raskin (who designed many aspects of the Macintosh UI).
He pointed out that "the only intuitive user interface is a nipple."
Several days ago my wife and I had a new son, so of course I watched them learn (together) how to breastfeed. It was not obvious to either one of them how to make it work -- they had to explore and figure it out together.
It appears that Jef was wrong: even nipples are not an intuitive user interface.
Clippy came out of Bob, Melinda saw that it was good. But seriously though, about that same time period was the dawn of CDROM as a media type. Many magazines shipped with CDs, and each had a GUI. The gamer mags in particular had various custom GUIs for selecting their content. Some were based on shopping, some on Office or Home (ala Bob) Some on really weird stuff (Anyone remember the elevator ride to hell on old PC Gamers?). It seems those were some real free wheeling days of UI development, but none of that seems to have gone anywhere.
- Minutus cantorum, minutus balorum, minutus carborata descendum pantorum.
I think the throuble with tiling is that it simply doesn't work that well as a generic concept, there are simply to much applications around that are just to small to make sense in a tiled workspace, ie. a small calculator should overlap, not tile, since else he can't be seen in full and wastes a lot of screenspace. However in Blender or Emacs tiling works great, much better then MDI solution which present windows in a window, this is probally because Blender and Emacs deal with one kind of data only and don't have to work with hundreds of different applications which made have wastly variing requirements.
However, while tiling has a fair share of problems, our way to manage windows is also far from optimal, there is a lot of time wasted with moving windows around and aranging the screen in such a way that it is actually usable Apples Expose helps a bit, but real solution is probally to move to a fully zoomable desktop, so that one isn't restrited by screen borders, but can simply zoom out when more space is needed. This also helps a lot with orientation, since you can simply place everything side by side and still reach it and don't have to lower/raise yourself through a stack of windows.
How about an LCD Keyboard that actually displays those shortcuts so that you don't have to type them blindly in the first place? Might of course still take a while those are actually available and affortable, but the problem with shortcuts is certainly solvable in a better way.
That isn't because smooth scrolling is a bad idea, but because it simply was badly implemented. Now I don't necesarily blame the developers for that, because some things simply can't be implemented well with todays hardware, ie. when I press down I don't want the screen to scroll automatically for half a second, thats just not really a good way of doing it. However I also don't want the screen to just jump around, since that is extremly disorientating. So what could the solution be? How about a pressure sensitive scroll button, the harder I press the faster it scrolls and scrolling both starts instantly when I press and stops when I depress. Or how about a scroll wheel that actually scrolls smoothly instead of just sending up/down events on every click?
I don't think it's that you "can't get people to convert," it's that the designers haven't come up with a compelling reason to get people to abandon what they know. Someone who's worked with a mouse+keyboard, desktop-style GUI for (in some cases) 20 years or more, isn't going to completely retrain themselves without a darn good reason. Right now, there aren't many compelling reasons to switch.
In essence I'm agreeing with you; there certainly haven't been very many really radical designs, and because of that, there haven't been many designs that really offer the average user that much more benefit over what they're using now.
Offer something significantly better -- enough to cover the retraining cost -- and people will flock to it. There's nothing particularly enjoyable for most people about using a QWERTY keyboard and mouse, it's just what they're used to because it's "good enough." Come up with an interface that lets people enter data as quickly and accurately as they can with a keyboard, and move objects as easily as they can with a mouse, and view and comprehend data as quickly as they can on a monitor, and -- like my grandmother used to say about building a better mousetrap -- people will beat a path to your door. It's just that to date, nobody has really built that better system; at least not that I've ever seen.
Designers seem hesitant to go 'outside the box' because they percieve users as being tame, but really it's the users who are cynical about new designs, because most of them are nothing but lame rehashes; "difference for the sake of difference," which throw away optimizations painstakingly made over years (or in some cases decades) without offering much new. It's not until designers really go outside the box that they'll stand a chance of finding a Better Way, and when they do, the users will follow.
"Ladies and gentlemen, my killbot features Lotus Notes and a machine gun. It is the finest available."
You make some interesting suggestions, but there are practical issues.
Take your example of a stovetop. Your observation is a fairly obvious one, but the obstacle is that there isn't room for the knobs to be placed in a square, due to other overriding design considerations. That's why almost no one does that. There is instead a conventional arrangement that is the same from stove to stove, so you only have to learn it once.
That's one reason the current arrangement of shortcuts would be difficult to change. Everyone knows that <ctrl-o> is Open, <ctrl-s> is Save, etc. If someone decided to do it differently, he'd be bucking the trend of literally thousands of applications and is not likely to be successful. He'll probably just piss off his users instead.
You dismiss the non-intuitive aspect of this with a simple "but", but I assure you that intuitiveness is very important to users. They prefer intuitive over rational every time, and they don't want to have to be "power users" to take advantage of what should be a simple feature.
Your idea about the numbers will not only not work for menus with more than 10 items, it would be very annoying for the many applications where the menus are user configurable. Every menu rearrangement would also change the shortcuts. I can't see anyone being happy with that. However-- please note that where you can customize menus, you can also customize shortcuts. Why not do that if you're unhappy with the default arrangement?
But mapping shortcuts graphically in any meaningful sense obviously can't be done. At that point it's no longer a shortcut -- it's the GUI you were trying to short-circuit with the shortcut in the first place! The fact is there must be some mnemonic or some other very easy-to-remember system, or it simply won't get used. Users always take the simplest path. A shortcut is supposed to be simpler for often-repeated tasks once you're only moderately well-versed in the software. Make it too complex, and it will no longer be simpler than selecting from a menu with a mouse.
I think a more sensible hand-balancing arrangement for alt-tab would be to keep the key combination with the left side and let you select one of the windows with the right. It surprised me trying it just now (it hadn't occurred to me before) that Windows will not allow you to do this even though you get something that looks like a box full of icons to pick. I seriously doubt most people are interested in (or able to unless they're twitch gamers) to "instantly" hit a number key on being presented with an entirely novel menu. In any event, there's no reason to free up the mouse cursor for repositioning since the most likely target (the window you're trying to switch to) won't be available until the switch is actually made. There's no place to move it to.
So pay attention to your HCI class. I never took one myself. I'm just speaking from about 20 years' experience in writing apps for users who are often sitting in the same office. When they don't like something about the UI (graphic or otherwise), I get instant feedback.
And the brethren went away edified.
mine has two dials:
1. power
2. time
to start it you close the door, to stop it you open the door or ding the time dial yourself
though they are getting harder to track down. Why my toaster needs 5 buttons as well as the time dial i will never know, presumably they'll be getting clocks soon too.
If you don't risk failure you don't risk success.