10/GUI

October 13, 2009 9:22 AM

Since SRI and Xerox invented the GUI and the mouse in the late 1970s, technology has leaped forward, but the way we interact with our computers has stood still. "10/GUI aims to bridge this gap by rethinking the desktop to leverage technology in an intuitive and powerful way."
posted by Plutor (66 comments total)

Why did I know that this was going to be about multi touch technology before even visiting the site? For those who want to cut through the tech bullshit, here's what 10/GUI is about:

10/GUI is probably one of the most dramatic reimaginations of the desktop user interface I’ve seen in a long time. This concept proposes a multitouch interaction system that does not require a multitouch screen (and thus does not have to deal with all the problems such a screen causes), but instead uses a multitouch area near the keyboard. The proposed graphical user interface makes full use of this multitouch area.

More from TechCrunch.
posted by Foci for Analysis at 9:29 AM on October 13 [3 favorites]

I think it's a little bit disingenuous to call it "tech bullshit". I think it does a fairly succinct job of not just presenting the ideas but also explaining the rationale. And it's actually hardly technical at all.
posted by Plutor at 9:34 AM on October 13 [2 favorites]

Seems to merge many of the ideas that have come from mobile devices - touch interaction, simplified GUI vs full windowing and move it to a desktop model. One funny thing is that I don't think I saw any text entry in the demo, although the final render shows a combo keyboard-touchpad. As long as I'm not forced to touchtype on a virtual keyboard it sounds good to me.
posted by GuyZero at 9:37 AM on October 13

The hype to innovation ratio here is pretty high. One can already do much of the first half of that video with the newer multi-touch MacBooks, though third party software is and will be needed to get at some of the fancy stuff until the OS catches up. But I can already, for example, swipe with four fingers between desktops. Multitouch mice and desk-pads are already on the way, probably by January, from Apple and Microsoft. None of this is very far future, and it'll definitely be fully integrated into OSX and Windows long before Xerox finishes revising that ugly logo.

Worse, it appears that in the second half of that video, Xerox has invested a lot of energy in reinventing Expose. (I haven't dealt with "piles of overlapping windows" for years. Center-mouse button, done.)

I wonder who will break the bad news to them?
posted by rokusan at 9:38 AM on October 13 [1 favorite]

A radical rethinking spoken as blandly as possible.
posted by yeloson at 9:38 AM on October 13

I already use a version of "con10uum" you might have heard of. It's called the taskbar.
posted by battlebison at 9:39 AM on October 13

Why have sensors for the middle, ring and pinky? The motor control of those is much less fine, and offhand I can't think of anything that I use them for independently on a daily basis (other than, oh, typing)
posted by leotrotsky at 9:40 AM on October 13

I think they have good ideas, but man, them complaining about the mouse being a constricting, linear, serial based interface falls a bit flat when they promote their concepts through sketches on a communication format (video) that forces me to parse their concepts in a constricting, linear, serial based fashion (time).
Another UI beef I have is that it takes 2:00 minutes of background before there's even a hint at what the solution is. As a technical guy I fall into the trap of describing the problem before the solution too, but it is not the best way. It's much better to say, here's my solution and here's what problems it solves. Doing it the other way is tying the presentation of the concept to the conceit of the concept developers and the learnings they had to go through to get to the solution.

It's a shame, because I'd love to have the UI they describe, as long as they aren't the ones providing the UI, if that makes sense.
posted by forforf at 9:40 AM on October 13 [4 favorites]

Has BumpTop been linked here before?
posted by chunking express at 9:45 AM on October 13

Yes it has.
posted by chunking express at 9:46 AM on October 13

MIDI was specified in 1982, ten years after the Xero Alto. So these guys have 10 more years to create some new, groundbreaking, perhaps even wave-based music solution for the PC and splice it with their groundbreaking video!

/end of obnoxious-midi-music rant
posted by Laotic at 9:46 AM on October 13

GoogleOS?
posted by TwelveTwo at 9:46 AM on October 13

I like having many, many terminal windows open (we're talking at least eight) spread across two windows. This is clearly not for me.

Also, less tactile feedback on something I would use several hours a day would not make me a happy camper, I guess making it a big button like Apple laptops would be a solution.
posted by amuseDetachment at 9:46 AM on October 13

Err... spread across two monitors.
posted by amuseDetachment at 9:47 AM on October 13

it appears that in the second half of that video, Xerox has invested a lot of energy in reinventing Expose

This, this and this. Expose is wonderful, simple, intuitive, requires no fancy new input mechanism, and comes with every Mac. It's revolutionized my computer usage. I no longer care how many windows I have open or how they are stacked. Once I linked my scroll-wheel click to Expose, I just click, they separate, I select the one I want. Done and done.
posted by hippybear at 9:48 AM on October 13

Separating the touch from the screen: Great idea.
Taking a full five minutes to explain why do that: Not so great.

I liked the examples, but the linear window manager seems more like they are overcoming hardware UI problems rather than solving actual human ones. I notice the concept video didn't show a person interacting with it. They went on (and on and on) in the first half about how traditional mousing is easy on the arms, but I doubt using this thing would be. Not to mention the fingers.
posted by DU at 9:49 AM on October 13

Leotrotsky, in the OS X applications to date, those third and fourth fingers seem to only be used in a present-or-not fashion. The presence of four fingers indicates a special function (like a right click), and one isn't expected to show any dexterity beyond placing four points down. I haven't yet seen an application where the third and fourth fingers are expected to do anything: the two finger 'pinch/pull' (a la iPhone) is the fanciest move yet deployed, along with the 'spin' variation. Again, two fingers, or (easier) one finger and a thumb.

I've been out of the UI biz for a decade, so I'm not up to date on the literature. I can't say if this is the long-term goal Apple's shooting for, or if it's just the easy first step to weld onto the existing OS. Evolution over revolution, etc.

One sideline of all this: such interfaces make an absolute mockery of many disabled users.
posted by rokusan at 9:49 AM on October 13 [3 favorites]

That's a deep, deep keyboard, and I don't think most people will have room for it, either on their desks or on their laps. Positioning would be tough; you'd either have to reach too far forward for the keyboard or pull your hands too close to your body for the touchpad.

Just did a little experiment now, and if I put the keyboard at a place that allows comfortable typing AND touchpad work, the heels of my hands touch the pad while I type. That can't be good.

Now, if they could make a keyboard whose keys had very small gaps between, and the touchpad surface was comprised of the key tops, with some sort of toggle method...hmm.
posted by davejay at 9:52 AM on October 13

such interfaces make an absolute mockery of many disabled users.

This.
posted by davejay at 9:53 AM on October 13

such interfaces make an absolute mockery of many disabled users.

This is true of many computer UIs in general and very true for smartphones. There's a research project at Google on designing a smartphone UI for blind people.
posted by GuyZero at 9:57 AM on October 13

They went on (and on and on) in the first half about how traditional mousing is easy on the arms, but I doubt using this thing would be. Not to mention the fingers.

Yeah, using a touch pad on a laptop really kills my hands, and I can't see how doing a bunch of fancy multi-touch stuff would make that any better. I like using a nice ergonomic mouse so that all I have to do is move my arm to move the cursor, with a minimum of wrist or finger movements. I already have enough potentially damaging repetitive finger movements using a standard keyboard, so I wouldn't want to switch to a mouse alternative that was even more RSI-prone.
posted by burnmp3s at 9:59 AM on October 13

an intuitive and powerful way.

Starting with the most obscure techno-babble for a name. fail
posted by blue_beetle at 9:59 AM on October 13 [1 favorite]

Y'know, I always thought the benefit of the single mouse pointer was that we humans, with our bicameral vision, focus on one thing at a time. Having a bunch of others floating around doesn't seem to have any real benefit if I'm paying attention to one thing.
posted by kafziel at 10:12 AM on October 13 [1 favorite]

Yes let's give the movement towards intuitiveness a totally confusing name that makes no sense when you first look at it.
posted by autodidact at 10:15 AM on October 13

Interesting. One of the benefits of large form multi-touch interfaces is placing your hands on the screen feels intuitive. The first time I used one of the systems, I walked up to the screen and started moving and resizing objects with no direction from the developers. An interface like this would have a much steeper learning curve than the multi-touch displays that are currently out there, but I wouldn't want to work with one of the current systems as my main computing device. They are great for conferences, training, and demos, but not for day to day computing tasks.

Having developed software for large form MT, the trick is to get the interface to feel as natural as possible. The disconnect of having the MT pad separated from the screen is going to make that a lot harder.
posted by ryoshu at 10:18 AM on October 13

Not sure that I really buy the whole horizontally scrolling windowpane bit. I suspect you'd actually need to play with a working version of it to know how well it worked, Still, it's nice that they are thinking about this stuff and presumably constructing such a thing.
posted by Artw at 10:24 AM on October 13

I always thought the benefit of the single mouse pointer was that we humans, with our bicameral vision, focus on one thing at a time.

In reality the mouse was the best thing we could built in the mid-70's that allowed someone to point to something on the screen. Lightpens were too expensive and not much better in terms of accuracy. Part of this whole idea is trying to find what we can/should do with new hardware that's come along in the last decade especially as it comes down in price. Ergonomic concerns are typically fitted around whatever tool we have - we're not really building tools around ergonomics unfortunately.
posted by GuyZero at 10:25 AM on October 13

Why did I know that this was going to be about multi touch technology before even visiting the site?

I was expecting a 3D desktop as well - I guess people have stopped beating that horse now or something?
posted by Artw at 10:26 AM on October 13

I was prepared to hate this, but I'm actually quite impressed; they've picked up some good lessons from the mobile space, without in the process abandoning most of the advantages of a windowed desktop (as so many attempted reinventions of the desktop do). I also give them a lot of credit for not trying to put it in 3D (as so many attempted reinventions of the desktop do).

There are still a couple of use cases I can think of where this interface would be inferior to the existing windowed desktop:
* Image editing (or other apps where you're working with a bunch of small, discrete files simultaneously) I might need a dozen images visible at once to compare them, and they wouldn't necessarily fit as a stack of narrow, full screen height windows. What do you do here? Go back to an MDI interface for those applications? That would abandon a lot of what this GUI tries to accomplish. Treat each as its own full-size app window? That would, frankly, suck.
* Tasks that require looking at more than two applications at once. Two side-by-side this can handle quite well, but if I'm building a website I might need a text editor, a browser or two, a terminal window watching the server error log, another with an FTP session open... with the existing windowed GUI I can arrange those all as i like, but with this 10/GUI I'd have to keep scrolling across a bunch of horizontal windows or else make them all uncomfortably narrow.

One obvious area for expansion of this idea, that might mitigate the above problems: use the vertical dimension for different workspaces. I'm really surprised they didn't include that (though possibly they didn't want to clutter the core idea with too many bells and whistles.)

Why have sensors for the middle, ring and pinky?
This is actually one of the better-thought-through aspects of this: the interface does different things depending on how many fingers you use. Fine motor control tasks you would presumably do with your index finger(s); any task that requires more than two fingers at once doesn't require fine control, it's mostly slinging windows around.

I already use a version of "con10uum" you might have heard of. It's called the taskbar.
Hmm. No. Not even close.

rokusan's comparison to Expose is much closer to the mark, but as much as I love Expose -- I feel crippled when using a machine that doesn't have it -- I can see some ways in which this might be an improvement on it. (Better use of Fitt's law, and having the 'windows' in a set order makes it easier to remember where things are: I often find myself hunting through all the thumbnails in a busy Expose window trying to find the one I'm looking for.)
posted by ook at 10:42 AM on October 13 [1 favorite]

Not exactly sure how my friend with only three fingers on his right hand is going to be able to manage this.
posted by medium format at 10:43 AM on October 13 [1 favorite]

Y'know, I always thought the benefit of the single mouse pointer was that we humans, with our bicameral vision, focus on one thing at a time. Having a bunch of others floating around doesn't seem to have any real benefit if I'm paying attention to one thing.

The invocation of four simultaneous sliders at the beginning was immediately convincing to me, and I suspect it would be to most other people who've used music sequencers or digital audio software, where there's an entire sub-industry built around custom physical control surfaces. A multitouch surface controling a couple of points simultaneously would be a great alternative. You'd lose tactile feedback but you gain an awful lot of flexibility. People are already interested.

I get the complaints about the "hype to innovation ratio," the music in particular seems a bit too "Welcome to the FUUUUUUTURE!1!!", and yes, other people have been and are working on these ideas. There are still some interesting thoughts here, there's certainly stuff I've never seen before, even though my MacBook Pro has a multitouch gestures too.
posted by weston at 10:49 AM on October 13

Err... spread across two monitors.

It would be interesting to see this idea spread across an arbitrary # of monitors in a way that builds logically on what was already demonstrated. I think the comparison to mobile devices is very apt in that this UI is really set up primarily for single tasking, the ability to run multiple apps notwithstanding. Certainly for the developers around where I work who have three 24" monitors going at once all in portrait mode, this isn't going to work as demonstrated.
posted by GuyZero at 10:57 AM on October 13

Oh man, at the crowded windows part I had the thought that this was a complicated lead up to a joke where the new linear system was actually a command line interface, and the touch pad just a new fancy keyboard. I wanted them to go on about how another new awesome innovation was raised buttons on the touch pad that sense your finger presses. I was prepared to bust a gut laughing.
posted by Mister Cheese at 11:05 AM on October 13 [1 favorite]

This makes me so looking forward to the first Apple tablet release...
posted by DreamerFi at 11:06 AM on October 13

I don't do overlapping windows and I rarely use a mouse. I use Awesome WM instead.
posted by Hastur at 11:07 AM on October 13 [1 favorite]

The multitouch control pad is pretty cool, and I don't think I'd have a problem acclimating to it.

That Con10uum windowing scheme, though...ugh. I know it's popular to slag our current messy windowing systems. But, dammit, it's a pretty darn human system. People ARE messy, non-linear creatures. Lining everything up in a neat line like boxcars on a track and having you scroll through the line to find what you want just smells like an over-thinking developer's wet dream. Nice, neat, orderly. Exactly what people aren't.

Keep the multitouch control pad. Lose the windowing system.
posted by Thorzdad at 11:08 AM on October 13 [1 favorite]

Lining everything up in a neat line like boxcars on a track and having you scroll through the line to find what you want just smells like an over-thinking developer's wet dream. Nice, neat, orderly. Exactly what people aren't.

The success of the iPhone proves that actually people are a lot like this. Or that they like it a lot at least.
posted by GuyZero at 11:20 AM on October 13

I'm like that when I am looking at a phone sized device.

(also there are things I just wouldn't do on an iPhone. Most of which require either large amounts of typing or wrangling multiple windows)
posted by Artw at 11:26 AM on October 13

It seems like this interface would be impossible to use if your multi-touch device didn't match the aspect ratio of your display, and I'm not sure how you account for that in the multi-monitor age.
posted by Loser at 11:26 AM on October 13

It seems like this interface would be impossible to use if your multi-touch device didn't match the aspect ratio of your display

Pen tablets (WACOM et al) have pretty much solved this already; you either just scale the touch aspect ratio to match your display (which quickly becomes unnoticeable unless there's a vast difference in aspect ratio), or you just letterbox-crop the usable area of the tablet.

also there are things I just wouldn't do on an iPhone.
I more or less expect "computers" to evolve into two different beasts: one based on the existing windowed-desktop UI, which will eventually be used pretty much only by nerds for development work, and one based on the iPhone or some other radically different interface, which will be used by everybody (including the nerds) for the majority of tasks that don't require wrangling dozens of windows.

The idea of trying to use a single device with a single interface for everything we use computers for already seems a bit quaint.
posted by ook at 11:41 AM on October 13 [3 favorites]

There's a research project at Google on designing a smartphone UI for blind people.
A virtually-blind friend tells me that the new iPhone 3Gs has a really nice VoiceOver interface that lets you use the phone without the display. It tells you what you're touching and so on.
posted by fightorflight at 11:53 AM on October 13

My Palm Pre does exactly this. Except they're called "cards." And they blow my mind with their awesomeness and I would pay top dollar to have cards on my pc.
posted by Baby_Balrog at 11:56 AM on October 13 [2 favorites]

I enjoyed the presentation. It had a few interesting ideas, and I think the concept is intriguing.

My heart sank a bit though when I saw the device attached to a keyboard. I have enough trouble with touchpads on laptops registering input when it shouldn't. I foresee a flurry of circles appearing randomly on the screen while I'm typing.
posted by Dr-Baa at 12:01 PM on October 13 [1 favorite]

Like Hyper-Cards?
posted by Artw at 12:01 PM on October 13

While they tout the ability to have all ten fingers as manipulation devices in the operating system, they don't show us uses past four. What exactly was the point of the left hand, just to show that the 1-4 tasks work ambidextrously?
posted by graventy at 12:11 PM on October 13

Software is developed in levels - just because you don't have a specific use for the secondary hand doesn't mean you don't build API support for it. The fingertip UI mockup just shows what data would be available.
posted by GuyZero at 12:17 PM on October 13

[e-mail]
date Tue, Oct 13, 2009 at 8:33 PM
subject Repetitive Stress
mailed-by googlemail.com

hide details 8:33 PM (0 minutes ago)

Hi.

The idea is fantastic, but it doesn't address one of the main problems that causes rsi's - having the hands constantly pronated in front of the user. Obviously preventing rsi's isn't your only objective, but would be great to see you try and address this issue.

Having a fancy structure isn't going to be economically viable, but how about giving the facility to use 2 10/gui devices simultaneously, side mounted at an angle so the hands don't have to pronate. This could potentially also lead to a 'travel version' made up of 2 smaller touch screens.

All the best with it.

Not Supplied
posted by Not Supplied at 12:35 PM on October 13

I started the video prepared to hate it — the keyboard and mouse have stuck around for 25-odd years because they work dangnammit, so get off my lawn — but I'm pretty impressed. I'm impressed in a way that I haven't been with a lot of other "improvements" to current schemes, like Expose. (The computer I'm using right now has Expose, but I never use it; I just command-tab.)

I question whether people will find a multi-touch UI as easy to get used to as a mouse, though. I remember what it was like to train totally unfamiliar users on a mouse (Apple used to have a whole interactive tutorial that explained and taught you how to click, double-click, drag, etc.; imagine a computer coming with that, today) and it could take some people a bit of time before they were able to just naturally move their hand and make the pointer move wherever they wanted it to go. That's with just one point moving. Getting to the point of total comfort, of just thinking and doing, with 10 points of control on a totally-blank touch surface seems like it could be very difficult and off-putting to some. I'm afraid the learning curve might be a bit high for the perceived benefits, and it might not get adopted for that reason.

One improvement I can think of that would make things easier, if more expensive: make the touchpad a touchscreen, in addition to the user's normal heads-up monitor. Rather than just a blank touchpad without any feedback, either have it mirror the main display (at much lower resolution, just enough to let you see what you're doing) or have it display some sort of simplified/high-contrast version of what's on screen, so that you can see what your fingers are resting on, just with your peripheral vision.

With most people's desk layouts, it's possible to see your fingers on the keyboard in the very bottom of your vision. If that 'keyboard' (or the area right above it—we're still going to get a real keyboard with this thing, right?) were actually a touchscreen, and it just displayed nice big blocks or circles showing touch-sensitive areas, I think the whole system would be a lot easier and more intuitive to use.

The display in the touchpad wouldn't have to be very high resolution; heck it might not even have to be in color. But just something to give it some feedback, rather than being purely one-way interaction, would make it a much more useful and approachable device. (Tactile feedback in real-time, like a Braille terminal, would be even better, but I think that would be impractically expensive and probably difficult to maintain. Touchscreens are at least a commodity product these days.)
posted by Kadin2048 at 1:24 PM on October 13

We already have a ten-finger haptic device that can do all of this and more. They even allow for chording and multiple simultaneous input. Yup, it's called a keyboard. More importantly, you can actually use them to control windowing environments without that stupid pointing device you're using now. It's much, much faster and more efficient. The only problem is that most people don't know that there's an alternative to the slow, clunky desktop metaphor compositing window systems that they use.

The answer is dynamic, tiling window managers. wmii is the one that I use. There's also dwm, awesome and several others.
posted by signalnine at 1:58 PM on October 13 [1 favorite]

Not Supplied, I'm still looking for a proper sideways mouse so that my hand doesn't have to spend ten hours each day twisted flat.

I found this one unusably clunky (huge and hard to move), while every other one I've seen has been of the 99-cent Hong Kong knockoff variety.

A good one really needs something that hooks around the back of the hand, because otherwise moving your (right) hand to the right means either leaving the mouse behind, or requiring way too much curled-up thumb grippiness: so much that I'm sure it'd cause as many RSI injuries as the more natural position alleviates.
posted by rokusan at 2:01 PM on October 13

They even allow for chording and multiple simultaneous input. Yup, it's called a keyboard.

Ctrl-Shift-+ is less appealing then "pinch" - plus on a touchpad/screen you get a lot more granularity in zoom levels vs zooming on the keyboard (at least how Photoshop implements it). Also, since people empirically have trouble learning or remembering keyboard shortcuts, why not try something else?
posted by GuyZero at 2:02 PM on October 13

I'm pretty sure this multitouch device will have a lot steeper learning curve than a few keyboard shortcuts. It's a lot more difficult to learn, "Make a motion like THIS" than "hit these specific keys."
posted by signalnine at 2:16 PM on October 13

Rokusan, this one is pretty good, and it goes round your hand. I've had one for a while.

To get the perfect hand position though, you'd need a keyboard without a numeric pad like a portable one one. I haven't invested yet...trying to think of a better interface that's within my technological power to make.
posted by Not Supplied at 2:21 PM on October 13

2nding the text entry being a problem. Virtual keyboards just require too much focusing on the typing area and don't have the feedback I expect.

A testament to the collective power of the ideas in the video (aside from not falling asleep to the narrator) is that I found myself imagining how you might put the multitouch sensor above an existing text keyboard, perhaps with ergonomically placed armrests slanting out so as not to block the keys below.

That to me felt right and completed my mental picture of how it could eventually be used. I hope they consider a way of keeping regular keyboards for the text part; any other existing text entry system simply doesn't have its advantages.
posted by Hardcore Poser at 4:15 PM on October 13

I found it a bit interesting until the words "At last ...". Then I realized they were trying to sell me on something.
posted by DarkForest at 4:25 PM on October 13

Around 2005, Apple bought out Fingerworks. (wiki)
Fingerworks was selling keyboard/mice based devices which were 1 or 2 touch-slabs, with a few raised dots to indicate the home row. You could type on them, start mousing, pinch to zoom, etc. We now see this in Mac trackpads and iphones.

The keyboard-mice things apparently worked quite well, were rsi-friendly, and users liked them. They occasionally pop up on ebay for several hundred dollards.
posted by sebastienbailard at 5:51 PM on October 13

Here's a good discription of it:
http://ask.metafilter.com/126463/Easytouch-keyboard-and-mouse#1806923
posted by sebastienbailard at 5:52 PM on October 13

I'm waiting for a USB-guitar-oriented GUI.
posted by ZenMasterThis at 6:25 PM on October 13

The success of the iPhone proves that actually people are a lot like this. Or that they like it a lot at least.

The success of the iPhone Windows proves that actually people are a lot like this. Or that they like it a lot at least.
posted by Jeff_Larson at 8:32 PM on October 13

Shocking, people, en masse, like more than one thing. All I'm saying is that it shouldn't be dismissed so quickly.
posted by GuyZero at 8:36 PM on October 13

I'm not convinced there's a problem to be solved, at least not with any kind of urgency. The first input that comes to mind is the driver's cockpit in a car - the wheel, pedals, gauges, and even the mirrors and gear shift haven't changed much in decades. The wheel and pedals haven't changed in 80 years. It seems like a very different circumstance at first, but the interaction required to drive a car - the subtlety of steering at speed, the larger movements required in a congested city, fiddling with knobs - are at least a little similar. They work pretty well and people adapt to them quickly. They also work for people who drive just a little (with a small amount of training) and for people who drive a whole lot with basically no variation.
posted by pkingdesign at 10:11 PM on October 13

pkingdesign: I think the problem with that metaphor is that, while the task of driving a car really has not fundamentally changed in the last century—sure, speeds have increased, but the actual task being performed would be recognizable to someone who cut their teeth on a Model T—computers have changed rather radically with each new iteration of input device.

Unlike with a car, where the task is basically predetermined and the interface must only let you complete that task as easily as possible, the tasks that computers are put to depends on their interface.

Just think of the transition from text-based, mouse-less computer systems to modern multitasking, windowed systems. There are a lot of things you can do on a modern system that would be very difficult to do, or at least would require more training and be less easy for casual users, with just the keyboard. (Sure, some people do it, but the public's failure to embrace any of these mouse-less alternatives is indicative of their learning curve.)

The reason there's such interest in new interfaces isn't because there's a real great dissatisfaction with current setups, for current workloads and tasks. The keyboard and mouse combo works just fine for work processing, web browsing, graphic design ... just about anything that computers are used for today, because those tasks and the applications that are used to complete them were developed for the keyboard/mouse paradigm. Likewise, keyboards alone were just fine when all the software was designed for them. But that's not to say that there aren't other tasks, maybe things that are kludgy or inconvenient now, which might be better suited to some new interface.

It's not until those new interfaces are actually developed that we might begin to realize their benefits. And it's for that reason that a lot of people are interested each time there's a development. In the past, we've seen changes to UI hardware drive the whole industry in dramatically different directions. That alone makes spotting the next interface paradigm before it really arrives a very big deal.
posted by Kadin2048 at 12:02 AM on October 14 [1 favorite]

I'm another who uses a tiling window manager (Ratpoison, but I'm looking at others) which suits me better than the regular desktop metaphor, when I had a Mac I didn't really use Expose because it was just another part of the same metaphor but I liked this demo. It certainly made me more excited than the GNOME Shell (ugh...).

The extended areas at each side where a nice touch too and would be even more powerful if you could configure them to suit your needs (which I'm sure is the intention).

It made me think about all the different ways I'd like to interact with my machine which can't be a bad thing.
posted by littleredspiders at 12:24 AM on October 14

One sideline of all this: such interfaces make an absolute mockery of many disabled users.

That was my first thought, too; some days I find it hard to use the mouse or type, so a multi-touch interface would be basically unusable. But then it occurred to me that all the window manipulation they perform with finger touches in the video can be done with extra buttons on the mouse (side-right for rearrange, side-left for zoom, both sides for move, whatever) or with pressure points on something more portable (like, say, a 360 controller) so I'd be okay anyway if the interface was properly configurable for disabled people (unlike, side rant, nintendo's stuff: hey, shake the controller to spin mario? and it's, like essential to play the game? and you can't remap that ability to, like, button x? fuck you, miyamoto, now I can't play your game). But then it doesn't seem particularly revolutionary if all those actions could be performed on mouse buttons just as easily.

Even if I were able I'm not sure I'd be that impressed with this new interface anyway. It seems perfect for devices with limited screen space or resolution, like games consoles or phones (and not being an iphone owner, the interface reminded me most of the "blades" on the original xbox 360 interface) but doesn't seem particularly relevant even to my relatively small 22 inch screen, which dedicates a smaller amount of screen space to its application list and doesn't expect me to be able to read sideways text, either.

I do think that reducing the level of abstraction in a user interface is a real step towards inclusivity and usability (witness the popularity of the wii, and the level of apeshitness directed towards project natal), but again I'm not convinced that this does: how is moving two fingers together to move in more or less intuitive than holding a button and moving the mouse up? But actions are trivially taught to new users, but both actions still need to be taught.
posted by ArmyOfKittens at 12:34 AM on October 14

Both actions are trivially taught, sorry.
posted by ArmyOfKittens at 12:42 AM on October 14

I think the problem with that metaphor is that, while the task of driving a car really has not fundamentally changed in the last century—sure, speeds have increased, but the actual task being performed would be recognizable to someone who cut their teeth on a Model T

Actually, the guys from Top Gear investigated this exact question, and discovered it isn't quite so cut and dried.
posted by hippybear at 9:46 AM on October 14