...quite a cool idea. You tend to aim toward the 'fat' part of the key - which is closest to deadspace anyway. With a big square key, you just push it somewhere in the center; the visual cues with this direct you away from 'difficult' areas.

Seems really obvious once you see it, but I don't recall anyone having come up with it before, so apparently it's not all that obvious after all... :)

Or is there not enough cash in it yet?

... until they get sued by Toblerone...

OK. I can believe that it might be an improvement, but why would there be any need to display the keys as triangles on a virtual keyboard ? Surely the underlying sensing mechanism could use triangles even if the keys are still square ?

Maybe AAPL could do that on just not tell the guy !

One of the Nokia 5800's onscreen keyboards (the mini-qwerty) has triangular keys - well trapezoid anyway. I imagine the reasoning is similar.

http://www.allaboutsymbian.com/images/features/5800/screenshots/ui-details22.jpg

"Baker believes that the virtual keyboard’s full potential will only be realised if it’s integrated into every area of the iPhone that uses a keyboard, such as SMS and email, rather than developed as a standalone application. For this to happen, he requires Apple’s aid, something he described as "very difficult to get"

er.. support that isn't required to implement a different keyboard for Windows Mobile. I guess now that he's told everyone his idea, we can expect a flurry of triangular keyboards for WM. Perhaps he should have implemented it there before attempting his patent!

Is he assuming that the sense area of an iphone keyboard (or any touch screen) is equal to the displayed area of they key? Just becuase the key is square does not mean that the area that is used to accept input is also square so his "invention" may already be implemented, and dawing triangular buttons makes no difference as its the "sense" area that matters.

Obviously a physical button is as big as a physical button, but a "virtual" button can be bigger or smaller than the screen area that its drawn on.

This feels like one of those times in Dragins den when the inventor suddenly goes ah.....err......oh........

Im not investing.

I reckon I should be able to mask the onscreen keyboard on my Touch Pro in just this fashion. Shouldnt take to long either

> For this to happen, he requires Apple’s aid, something he described as "very difficult to get".

Presumably, he also "requires" Apple to pay him buckets of money to license his patent. He just decided not to mention that, and spin it as "OMG Apple should take this and implement it or else they are EVIL!!!". And you fell for it.

Completely unnecessary on a virtual keyboard, where which key has been pressed is decided by the software.

What ever way you look at it, all he's doing is making the target area for any key smaller, and therefore training yourself to be more accurate at typing, through lots of early failures to hit the right key.

His screen dump is hilarious - it's all keyboard and no where to type!

Very nice. I will take a few of the them if they come as a physical device with a usb connector. Probably help my typing a little and certainly wont hurt it. I wonder how it would fair with the over 150wpm crowd.

Maybe iPhone doesn't support keybaord replacement but I believe Windows Mobile does.

Only iPhones matter.

Good idea though.

@Hywel Thomas, David Edwards:

The point of this layout is that people will aim for the top centre of the key. Therefore, if they miss slightly, they are much more likely to hit "dead space" than the next-door key.

A triangular sense area with a square graphic wouldn't be very useful - people would aim at the absolute centre of the key, where they are much more likely to miss and hit "dead space".

-----------

Overall, it looks like a pretty good idea. Hope he gets his patent!

It's not necessarily the area of sensitivity that's important. By displaying the keys as triangular, the user is visual guided to press the key at its widest area, effectively missing any other keys. You can then increase the sense area of the key and still unambiguously detect which key the user intended to press.

Making the sense field triangular while keeping the keys square will cause the opposite effect: users may be inclined to press any part of the key not necessarily captured by the input sensor, and get frustrated when the input is not recognized.

-dZ.

I don't see the point of displaying triangular keys, and the touchscreen is a grid and would have to stay that way, also it is obvious in using the iPhone that the keyboard software is already very clever indeed and seems to intuitively know which key you meant to press by how your finger lands and of course by what letter is likely to follow another.

"So, for the time being at least, the Crocodile Keyboard for touchscreen phones remains just a dream."

So, because it can't be easily done on the iPhone, it follows that it can't be done on any touch-screen phone, ever?

That's the kind of specious reasoning that would get a reporter fired back in the days when editors took pride in their content.

I hope someone comes up with a deformable screen, that allows tactile responses. I hate looking at a keyboard while typing.

Oh, and it took me five seconds, to come up with a better design. I noticed that the space around the keys is, so you do not accidentially hit other keys. Well, wanna know what else prevents that? Bigger keys! ^^

So essentially, they are just making the keys bigger, but not all of it touchable. I would make the triangles point upwards, and fill a third of the empty spaces too. But still only react when touching the inner, downwards pointing triangle. And enforce minimum sizes for both triangle areas, in the patent license.

Because with those keys, companies will just try to fit more keys onto the screen, until you can't hit them properly again. Which was the source of the problem in the first place. :)

The human eye/brain is very good at estimating the centre of a circle. Why not make round buttons?

This is a great idea! It's no good having a square touch area over triangular keys, or a triangular touch area over square keys, think about it. With the touch area and the keys both triangular, you can touch the top part of each key and be sure that no other key will accidentally be pressed.

That was my first thought too. Show rectangles but detect triangles.

However, showing the triangles helps the punter aim better which is the key thing to do here.

I wonder whether there are even better shapes than triangles. I can see triangles being good for physical keys, but perhaps a more Y-shaped scalloped-edged triangle would be even better for virtual keys.

Increasing the amount of dead space will increase the number of times people don't hit anything, making any reduction in acidental keys presses pointless. I would suspect that hitting dead space would be way more annoying as well.

Stupid idea

Not sure about the dead space. Why not Windows Mobile 6.5-style hexagons for the display, with the edges of the hexagon less sensitive? If you want a space filling shape and you want users to aim at the middle of the shapes, surely a hexagon is the obvious tesselation? Triangles just make the keys hard to read.

"Made with triangular honey from triangular bees

and triangular e-mail from triangular keys"

I actually think this is a great idea - the problem with square keyboard virtual buttons on a touchscreen is that the user isn't really sure where to aim, because your fingers aren't pointy, but the part that comes into contact with the touchscreen is the underneath of the finger, towards the front of it, not the actual front it, due to the way in which the device is held, and the angle of viewing.

A triangle on screen would, I think, significantly help users aim at the top of the triangle, as nobody is going to try push the tiny angular point at the bottom of the virtual key, with the result of hitting it in the middle.

It's amazing how many years it takes for simple things such as this to be created!

Don't forget about Jailbreaking! If you post it in the jailbroken appstore, MANY users can use it. It isn't as mass users, but it's still a fair amount!

This isn't worth the effort. A smaller key graphic means smaller legends. The iPhone and its ilk already have small displays; making the virtual keyboard even harder to read is a non-starter.

As for the accuracy issue: the iPhone implementation displays an enlarged version of the key it thinks you're touching. If it's the wrong key, you can see it's the wrong key and simply slide your finger over to the correct key instead. Missed the "F" key and got a "G"? Slide the finger a little to the left and the "F" appears instead. Let go. The app gets the "F".

Virtual keyboards are compromises almost by definition; there's no tactile feedback and even if the display were deformable, it'll still be small and fiddly. A chordal design would be a better fit for this form-factor, but for occasional use, the virtual keyboard seems to be good enough.

Personally, I find Apple's implementation pretty good compared to the one on my not even slightly missed Windows Mobile 6-based HTC brick.

(Note that my replacement for the HTC brick is... a Nokia 2630. Nokia's Series 40 GUI is pretty much bang-on; Apple aren't the only ones who can get it right, though they do seem to be the most consistent.)

Remember the old Bowmar pocket calculators? They had keys that were about half an inch square, with only a small space between them.

Then pocket calculator manufacturers realized they could provide as much space for people's fingers, and yet put more keys on the calculator, by making the keys smaller but having space between them, since the same empty space would serve for the two keys that adjoined it.

So this principle is not entirely new, although an effective area that is like an upwards-pointing triangle might fit the fingertip better than a square.

<> shaped keys. You still get plenty of dead space around it and so on ;)

That is all.

... couldn't you just have a smaller sensing area in the centre of the key? That would surely do the trick without the gimmicky triangular keys. Oh hang on, I think I've just answered my own question. The gimmick is the main thing isn't it?

Been done long ago, and it worked fine. Anyone else here remember teletypes (ASR-33, etc)? Lovely clunking keyboard, gave your forearms a bit of a workout and no-one got RSI on them.

Bah! My K’Vort class Bird Of Prey had triangular keys all over the bridge, especially for the weapons stations! A true warrior worries not about the shape of the key, for victory with honour is all that is important in the world of glorious text entry!

Mines the one made of targ leather with the Mek’leth in the pocket, purely in case some hoodies try to steal my mobile…

I have to say, my intuition says that this isn't any use on a small keyboard, where the primary issue is space, but I would be prepared to believe him if I had seen a study that showed error rates and typing speeds compared to a more conventional layout. If it is implemented exactly as described, it may well prove worse, as the contact area for each key is lower.

I'm still waiting for some 'genius' to figure out that the keyboard keys are not in alphabetical order. It takes ages to type anything since you have to spend most of your time hunting for each letter. If they were in alphabetical order then at least you'd have an good idea where to find them on the keyboard.

Christ, you people are one tiny step away from trainspotters, I swear.

You poke a touchscreen keyboard you expect a key to be pressed, the one closest to where you poked.

What is this 'dead space'? Just an annoyance meaning you have to be more accurate than necessary.

If users guided by the graphics tend to consistently miss in the same direction then the sense areas can be offset from the graphics to compensate with no need to change the graphics.

What would be smart is for the device to learn how a user tends to miss and dynamically adjust the sensing area, real time calibration of the user. Not difficult to do and IMO too obvious to merit patent protection.

Didn't I see you at the end of the platform at Crewe the other day?

Since typewriter keyboards have been around over a hundred years and keys have had many shapes over those years, it is immaterial how keys are shaped, and therefore should NOT be patentable. How the keys actually cause letters to appear, however, should be patentable. Also NOT patentable are key configurations and the shapes of keyboards. If this dude is granted a patent for this frivolous nonsense, then it shows that there's yet another patent office out there with its organizational head up its organizational butt.

"I'm still waiting for some 'genius' to figure out that the keyboard keys are not in alphabetical order."

No Frank, it's the *alphabet* that is in the wrong order (along with the keyboards in all other countries, of course).

What the crocodile keyboard does is to put the largest part of the key where the largest part of the finger can contact it without overlapping onto another key. The triangle shape eliminates the lower part of the key where an overlap of the shape of the finger would hit the lower part of the key above the intended target.

True, there are differently shaped fingers and angles of attack, but the general idea of eliminating the contact area where there is none desired is done easily with triangles. Other shapes still put contact (or sensing) areas where they are not needed. Ovals between ovals or other shapes might do something similar, but triangles are easy.

I assume that was a joke Frank...

Looks like an excellent idea. I plan to start work on a refinement, involving heavy use of a pumice stone on my fingertips....

I'll take nine to have his babies, just the one though.

I tried to find him via...

http://v3.espacenet.com/searchResults?locale=en_GB&IA=David+Baker&compact=false&DB=EPODOC

but could not find it and the chat-room function is not implemented.

"For this to happen, he requires Apple’s aid, something he described as "very difficult to get"."

Yes... I would assume that Apple, being most voluminous in the area of screen fiddling patents themselves, will be a bit miffed at not coming up with this one themselves so that will be tough fuck. Perhaps he can link up with some iPhone programmer to get an implementation on the iPhone so Apple can reject it.

To everyone else who does not, at a bare minimum, consider the concept, wherever it comes from, to be a bit 'in your face' cool may I just say I have an IQ above yours so keep practising on your Nintendo DS.

Oh BTW... not wishing to seem biased but if Matey David has a UK patent on the idea then it has to be quality.

<trivia>

Actually, in the early days of typewriters, the keys were in alphabetical order, but the mechanics of the device kept jamming and the load was disproportionally skewed to one area of the mechanics. Someone, I know not who, rearranged the keys to slow down the typists and effectively "load balance" the keyboard, which left us with what we have today.

</triva>

It's fun to learn, because knowledge is power!

Power to crush all those who oppose you.....

I think you'll find the Klingon Empire has been using triangular keys for some millenia now.

They certainly were in Klingon Honour Guard, circa 1998.

Pirate, 'cause Klingon pirates would be something to see, yaaaarrr...

Voronoi regions would be the best general solution. You cold add dead space on the boundaries, but why would that be better than the closes key?

(And it would result in hexagons for a hexagonal grid.)

The keys were arranged so that the neigbouring keys would be the keys with the least chance to be the next letter in an english word.

This was done because on old typewriters the arms blocked if you pressed neigbouring keys.

this had nothing to do with slowing down typists.

QWERTY is much easier than a alphabetical keyboard.

Ive used a ALPHABETICAL keyboard before. And you spend more time searching for the letters than anything. its also uncomfortable

QWERT is designed to use without too much wrist movement.

the cat sat on the mat. - QWERTY

zut eqz lqz gf zut dqz. - Alphabet

http://www.comfort-software.com/e107_images/akeyen.jpg

Which is easier to type?

"QWERT is designed to use without too much wrist movement."

FAIL!

QWERT was actually designed to keep appart the letters commonly used together within the mechanism of a traditional keyboard to prevent the whole lot jamming.

Before world+dog settled on QWERT other layouts were tried. When QWERT was first popular people who were used to other layouts complained that it slowed them down because they had to hunt for letters and thus began the myth that QWERT was actually designed to slow down typists in order to prevent the typewriter jamming.

The idea that QWERT is designed to minimize wrist movement is nonsense. As is the idea that other layouts are necesarilly slower. Sure you will be slower if you switch to another layout, but it's nothing to do with the layout being inherently slower and everything to do with the fact that you are having to learn a new layout.

a triangular key logger

Since the old "the letters are in the wrong order" thing is going round again, even if Frank was joking, people have tried alphabetical modern keyboards (and indeed a lot of touch screen devices do offer them), and they're really no better - and usually worse - than QWERTY. The usual problem is that the alphabet is in order linearly, but that keyboards are split into rows; knowing where to look makes sense only if the row length is standardised (or, admittedly, if you arrange the letters in diagonal stripes so "ABCD" map to "QAZW" on QWERTY). Even if everyone standardised on such a keyboard, it would still require a bit of hunting and pecking for new typists and be awkward (given the mapping of letter frequencies to fingers) for experienced typists.

Other layouts can be at least slightly better - I use QWERTY on one keyboard and Dvorak on another (stops me mixing my computers up) and I really do find Dvorak to be more comfortable and probably faster for some tasks; it was also reasonably easy to learn. Other layouts, like Maltron's, are supposedly better still. However, the (Western) world uses QWERTY (or variants thereof), and it's not that bad - it's probably not worth the hassle of everyone changing. Anyone having a problem with it just needs to practise a bit, like the rest of us had to, and not be deluded into thinking that alphabetical ordering would help.

Back on topic, I can't see that the key shape is going to matter much. If you want typing on a touch screen to go faster, find a way to give some touch feedback (e.g. small bumps where the keys are). Fingers don't see, they feel; a fancy keyboard is no good if you can't see it behind your digits, but something to keep your fingers centred would be a much greater help. I can type on an unlabelled keyboard - I once used a keyboard that someone had swapped the keys around on (into, as it happens, alphabetical order - along with more evil changes such as swapping the numeric keypad from calculator to phone layout) and didn't notice until I looked to see whether the punctuation was in UK or US style - but I crawl on a laser keyboard.