10/GUI — an Interface For Multi-Touch Input

Naznarreb writes "R. Clayton Miller has an extremely impressive GUI concept he's calling 10/GUI (video; written description here). Essentially, it combines the high-bandwidth input possibilities of multi-touch interfaces with the ease and immediacy of a mouse. The video is quite interesting, and, for me at least, pretty jaw dropping. This is a dramatic re-imagining of the current mouse/screen schema, one that I think has significant potential."

Comments

Not for desktop pc's, but

[sopssa]

Theres still a few problems though. For one, mouse is an incredibly precise input device - you can pretty easily move it along same pixel axis, or get it precisely to a specific pixel. It's hard to do that with your fingers because the area they touch is a large one, it's not easy to just move your finger by one pixel and your hand tend to shake a little bit too. If you look at the video, you see everything in the interface is quite big and even a few small windows take lots of place.

Other problem is that now your both hands lay on the wide touch area and you dont have a keyboard. If you put them side to side, you'll only have one hand on the touch area and dont get the full power of it. Moving hands between them all the time is inefficient. Typing on the touch area gives no feedback and again takes your hands of the "mouse".

It would also be quite impossible to play FPS or other kinds of games with this type of setup.

So no, I still dont see touch interfaces replacing the usual keyboard+mouse combo anytime soon. However, I would love to have this kind of system in my living room (either just for the tv, or the computer thats connected to tv screen). It's clumsy to have keyboard or mouse in living (at the moment I have MX Air -mouse, which is okayish), but this would be perfect for such job. Not for a desktop pc replacement though.

Re:Not for desktop pc's, but

[tepples]

it's not easy to just move your finger by one pixel

Place finger on surface, then roll finger without lifting or dragging it. I do that all the time on my laptop's trackpad. Besides, you don't need to use pixels as the fundamental unit of movement if your input device can detect movements smaller than a pixel. Putting something at a subpixel position is even easier with modern GPUs (and even Intel GMAs) that power compositing window managers.

It would also be quite impossible to play FPS or other kinds of games with this type of setup.

Even RTS or rail shooters?

Re:Not for desktop pc's, but

[LBt1st]

At the very end of the video they show a keyboard positioned over the touchpad. So I don't think they're trying to eliminate the keyboard entirely.

Still you've got many valid points. The mouse is still a much needed tool for many tasks.

Re:Not for desktop pc's, but

[smitty777]

Another problem is the assumption of 10 digits. For example, this might be more difficult for someone that is missing a digit or is paralyzed in an arm. In that case, the mouse would have a definite advantage.

Re:Not for desktop pc's, but

[theurge14]

It would also be quite impossible to play FPS or other kinds of games with this type of setup.

Are we all really that stuck in our ways that this is how we judge possibly revolutionary ways to interface with our computers? By how well it stacks up to an already poor approximation of shooting guns in a computer game?

Re:Not for desktop pc's, but

[Mysticalfruit]

Fine then, have a mouse as well for super detail work.

I watched the video and I found in very interesting. As someone who's sitting in front of a 24" monitor and I've ~30 windows open, I totally get the clutter thing.

With that all said, what I saw was lots of talk and lots of eye candy. If you go to the 10/GUI website it's completely devoid of any details about hardware, what OS it's going to be supported on, etc.

Until there's more details, I'm calling vaporware on this...

Re:Not for desktop pc's, but

[Svartalf]

Actually, he's described enough of the framework in the video to do it fairly easily with the stuff that's currently available.

Seriously. You'd use a capacitive or similar touch-pad and do multi-touch against that backdrop as an input source. All one has to do is apply something along the lines of this and modify it to understand his local/global edges of the touchpad and then implement his window management system on top of one of the lightweight WMs out there as a fork.

However, while that would require a smallish amount of work, one has to wonder if he's got patents applied for or a copyright on the "look and feel" that he'll let it all happen and then submarine the whole thing when it becomes "the big thing". If he's letting anyone have access to it, or letting FOSS projects have it under FOSS terms and proprietary under similar RAND terms, then I'd say let's see how the idea actually works. If not, I'd say give it a pass. It's interesting enough to evaluate if he's barking up the right tree or not- but only if he's not merely setting himself up as gatekeeper so he can extract rents on a potentially useful interface paradigm.

Re:Not for desktop pc's, but

[cwgmpls]

I understand that people with disabilities have computer input needs too. But I don't understand why the fact that we are all differently-abled should prevent people from using their abilities to their fullest capacity.

Does the fact that some people are missing a digit or have paralysis in one arm mean that no one should propose playing a piano with ten fingers? Does that fact that some people don't have feet mean that pianos should not have foot pedals? Of course not.

While we should move forward with good interface designs for people with disabilities, I don't see why we should stand in the way of people using the abilities they do have in a novel, more productive way.

Re:Not for desktop pc's, but

[timeOday]

I switched from VI to emacs about 10 years ago primarily because of emacs' ability to manage multiple windows (by which I mean buffers in this post, not separate X windows), and I still think it offers most all the benefits of 10/GUI! You can have columns of buffers (windows), but also rows (addressing your concern), but they're not just arbitrarily strewn around. You have a high-bandwidth, but precise input channel that doesn't block the screen, called a keyboard. You can easily access a labeled list of open windows. No, I don't expect a resurgence of emacs to wipe out the gui, but people who are designing windowing systems should at least be familiar with how emacs worked. (Maybe emacs borrowed it from lisp machines? I don't know). Granted, keyboards aren't exactly multi-touch (except for modifier keys like SHIFT), but keyboards do utilize the ability to move several fingers at once to achieve high-bandwidth input.

Re:Not for desktop pc's, but

[fprintf]

If your keyboard is "on your desk" then it is probably not the best ergonomic setup. I do this, but that is because I am tall and my knees hit any keyboard trays. I would *hate* having anything below the desk. So for me, it might be the best arrangement to have a split touchpanel or keyboard.

For example, I might like to have the touchpanels on either side of my keyboard. I can put my hands over there. Either that, or split the keyboard (ala Microsoft Natural) on either side of the touch panel.

Overhyped

[ledow]

How can you over-hype a one-paragraph summary?

Five minutes into the video and I'm still none the wiser as to how this is supposed to be an improvement in the use of my computer, or more comfortable, or easier. The "real-world" demo towards the end doesn't seem at all impressive and leaves out an awful lot of computer uses (we'll start with gaming, because it's easier to pick on multitouch for that).

Why is everyone determined to sell me multitouch but can't actually show a decent use that justifies the price/hassle/upheaval/software development costs?

Re:Overhyped

[kevinNCSU]

Because everyone grew up watching Star Trek which clearly shows all computer interfaces are supposed to become glass touchscreens dangit!

stretching to type

[koekepeer]

I love the concept, but I imagine myself stretching over the touchpad area to type, which wouldn't be very ergonomical. I can also imagine that the base of my palms would rest on the touchpad area occasionally as I type.

It's an interesting implementation

[r_jensen11]

But I'm not sure how many people will be wanting to drag their fingers across a surface for 8 hours a day. One of the benefits of a mouse over a touch service is that there's less friction for the hands- all of the rubbing of objects is between the mouse and the surface.

Can I see this replacing the mouse? No.

Can I see this supplementing the mouse? Yes

Can I see this being placed with a mouse and keyboard? No- the combined three objects would take up too much space (who really has that much desk space?)

Re:It's an interesting implementation

[ceoyoyo]

I use a track pad all day. It's no problem. I even still have fingerprints.

Re:It's an interesting implementation

[MMC Monster]

Apple's multitouch trackpads on their current notebook lines have it right. In fact, they are so good that I wish they would sell a stand alone trackpad to add onto a desktop keyboard. Using gestures to scroll around a window and two finger click or hold and drag are often much faster than moving around with a mouse.

Not that I would ever get rid of a mouse, except (potentially) on a media system with a limited physical keyboard.

Window management

[phorm]

It doesn't really seem like an improvement in window-management for me either. Sure, window overlays are a bit cluttered, but then again there's only so much information one can process at a given instant.

I tend to have a *lot* of items running as I multitask. A web-browser, document, several terminals, perhaps a coding window, and others. Having windows aligned horizontally it going to be a PITA if I have to zoom out every time I need to jump from #1 to #15. In that event, a taskbar really is quite a nice thing and "just works". Perhaps rather than having a left (contextual) and right (global) menu, they could also have a bottom/top (taskbar) menu.

Needs a curved surface

[Anonymous Coward]

The flat surface is not ideal from an ergonomic point of view. The touch-sensitive surface should be curved so that the user's hands can be held at a more natural angle. Preferably two domes of soft, touch-sensitive material, and two small raised dots on the top for tactile positioning. Hm? what? sorry, I drifted off there for a moment... what were we talking about?

VT100

[thomasdz]

I like my GUI text-only, 80 characters wide by 25 lines long. The way Ghod intended.
oh yeah and 7-bit ASCII only... none of this fancy schmancy 8-bit extended code-page goop.
and GET OFF MY LAWN! Damn kids with their game boys.

Replacing current business work interface

[Dripdry]

I disagree with an earlier poster who says this can't replace the desktop interface we know.

He points out that it is inefficient because one has to move hands between keyboard and pad.
It seems to me that this interface can be manipulated with just one hand. that's how the mouse works now. you take your hand off the mouse to work with the full functionality of the keyboard, why couldn't 10/GUI replace that?
Better still, why not have both! The pad can sense a mouse and act as a mouse pad. If you need to use it as a pad move the mouse off, or perhaps use both the mouse AND the pad at the same time (one hand each). I can see a lot of possibilities there. It could clutter a desk, sure, but I'm sure we can solve that problem.

Second, but probably more importantly, I can easily see this for the work I do, which I imagine is similar to many other people.

I use my system (two monitors) for business. I have data on one side of the screen and sometimes excel or word on the other screen. In addition (here's where 10/GUI could be useful) I'll have a pdf open in the background which i need to quickly scroll through ("in adobe, quickly!?" you ask?). With one hand on the mouse I can quickly zip through Morningstar data, and use my free hand (on the pad) to scroll through a document, then quickly zip back and forth, scrolling and zooming as necessary. Right now that's just using a mouse and it can be tiresome to move around with just that little pointer (especially morningstar! oh it would be nice to have a touch interface for that...)

Finally, I need to have a "document scroller" or whatever it is that I can resize and move around, as I find myself with documents that sometimes need to show different parts of their data on screen. Basically I need to be able to "undock" documents so i can use them effectively (like papers on an actual desktop) and then redock them when I'm done.

That would be a beautiful interface that I do feel would save me some time and frustration. I would buy that for my business.

Virtuoso Users only!

[starglider29a]

Yes, this is slick. Yes, it's an improvement. Yes, this will happen. But...

Having seen people have trouble with pressing control and clicking at the same time (to deselect a single item), I foresee a chilly reception, user frustration and a training issue. 10GUI is like playing Mozart among people only able to manage Chop Sticks.

I see this as stratifying feature... the have's and have not, the able and the un-able. I would request this for my workflow, but the run of the mill admins would be stuck with the keyboard. Aside from the social aspect, there is the difficult task of convincing the boss that "you need this, even if the others don't". Good luck with that.

I have grown to hate the windowing paradigm for all the reasons cited. I'm not convinced that the linear arrangement is an improvement. I'm more in favor of multiple monitors, the main screen for the primary task and satellites with multiple windows for ancillary tasks. 10GUI doesn't address this.

going in circles

[jipn4]

Apple got its multitouch technology by buying a company called FingerWorks. FingerWorks' primary product was just like what 10/GUI describes: a multitouch surface that could either replace they keyboard or the mouse(pad). It largely failed in the market.

People use the keyboard and mouse because they really work well. If people did want more DOF, it would be easy to add more sensors to a standard mouse, for example to record twisting, pushing, and other pressures, but even that isn't catching on.

Another idea that keeps bubbling up is the idea of pressure sensitive keyboards; they also keep failing because the resulting interactions just become too complex and add little benefit.

The real flaw in all these devices is the assumption that the limiting factor in communicating with machines is they "bandwidth" of they keyboard and mouse. It really isn't. Generally, people can think no faster than they can type and mouse, and speeding up the keyboard or mouse any further is pointless.

Re:going in circles

[Tom]

As an owner of the FingerWorks TouchStream (the keyboard/mousepad multitouch combo product), I disagree with your assessment. The reasons the keyboard failed are many. One of the most important ones, I think, was that it was too early. Yes, it had multi-touch, but no application supported multi-touch. So the only thing you could use it for is a virtual keyboard. The keyboard was good, but lacked haptic feedback, thus making touch-typing very difficult.

If applications would support multi-touch, a product like the TouchStream could be a success, because now it gives you an advantage over a conventional keyboard.

Re:going in circles

[jipn4]

There is no such thing as "zero force typing"; it's physically impossible. And hitting a hard surface while typing is subjecting you to more strain and requires you to use more muscles than a regular keyboard.

Moving your hands between they keyboard and mouse is generally a good thing. However, if it really bothers you, there are many existing keyboard designs that use regular keys and have a built-in trackball or trackpoint. You can also use footmice.

If the Fingerworks product works for you, you should probably stick with it. But it's probably not a good choice for most people, and it's probably not even the most ergonomic choice for you.

Application-centric workflow is a problem

[argent]

Application centric user interfaces are already a problem. On both Windows and Mac these days there's an increasing level of application-centric organization, and that breaks the task-centric workflow badly. I normally have a separate workspace for each task, with windows from each application all visible simultaneously. I can surround each primary document with windows of all sizes, to the sides, above, and below. The 10/GUI control model looks very very good, and would work well for a multi-desktop window-oriented workflow, but the Con10uum user interface would be a huge obstacle multi-document workflow.

All 10 fingers

[MpVpRb]

You know what other interface uses all 10 fingers...a piano.

Virtuoso pianists can make magic with the interface.

Most people never really get it.

This is the fundamental limitation of all these schemes. The mouse is easy enough that anybody can use it. The more manual dexterity that's required, the more it becomes like learning to play the piano.

Re:All 10 fingers

[kitsch]

It's not the number of fingers that makes playing piano hard. It's the combinations. The chords are the hard part. Also the independent use of each finger. An interface like this and most other multi-touch interfaces use simple clenching or releasing motions. These are movements that anyone with fully functional hands will have already mastered. They are baby movements.

As for FPS games I don't see the issues I'd say my finger tip is roughly the size of a baddies head. Panning would be a breeze. Group selection in RTS games would be much nicer. If you combined this with a pen for drawing then I could easily see a mouse becoming obsolete.

As for palm interference I doubt that is a hard thing to treat as noise in the input.

The biggest problem is replacing a keyboard. That wont happen any time soon. Though for many tasks (like Gimp/Photoshop, or 3D animation) you could replace the small key combos with larger key areas. Replace 3D mice with circular panning areas and such. I would prefer that the touch screen to display an abstraction of the interface -- window outlines, key areas. Nothing too detailed and distracting -- maybe an e-ink level display.

An incremental hardware update putting on airs

[SpaceToast]

I blogged this last night. Short version: fail.

Problems:

1. You just doubled the amount of space I need between myself and the monitor.

2. Multitouch allows for more kinds of interaction: true! However, this interface steals ALL of them away from use by the applications.

3. Left and right sides of the screen aren't discoverable. Might as well be top and bottom -- i.e. bottom of the screen for application launching (call it a "dock") and top of the screen for context-specific options (a sort of "bar" of "menus").

4. Linear spatial overload of windows is no better than two-dimensional spatial overload of windows. Labelled zoom-all-the-way-out cheat no better than Expose and application switcher.

5. Where does file management fit into this scheme?

Lukas Mathis calls 10/GUI "one of the most dramatic reimaginations of the desktop user interface I've seen in a long time" but on examination it's an incremental hardware update with no real interface breakthroughs. Keyboard + mouse has gone on for far too long, as has the W.I.M.P. interface. A better direction would be a tactile multitouch surface which can be anything it needs to be, including a keyboard (for any language), coupled with a GUI that represents tasks and actors rather than objects in a space. 10/GUI does nothing about window and document clutter, squinting, scanning large lists, or making the computer's workings and status an organic part of its presentation. The video may be a slick investors' reel, but shows no real progress.