Chris,
I am not (I hope) a stupid moron. But I have seen inventions that I should have thought of but did not.
One of these is the passive holographic optical element backlight for PDA screens.
In the pioneering PDA project I was on, we spent months trying to make an electroluminescent backlight work. We could not for many reasons. Finally we went for the usual reflective layer.
Then Polaroid pitched us a reflective layer that was a holographic optical element. It worked amazingly well.
I felt really stupid.
Not only had I just wasted months, but I had worked with holographic optical elements before.
The use of an aspheric surface as a HOE never occurred to me.
-Ben
Only problem here is that apples are not oranges, and as such, hardware isn't software.
Let's make it simple. I spoke of the PDF viewer i use. It's called "kpdf" and comes with the KDE Desktop. If you are viewing a PDF file, among other things you can do the following:
- Scroll/pan around the document
- Zoom in and out
You can do that in different ways too. The scroll/pan can be done using the scrollbars of the window, or the left/right/up/down arrows next to the scrollbars, or by click & hold down the left mouse button and then moving the mouse.
The zooming can be done by using the +/- icons in the toolbar, by selecting a predefined zoomlevel from a dropdown-menu in that toolbar, or by clicking & holding the middle mouse button and the moving the mouse.
How would you now replace the pointing device (mouse) to get the same functionality by using a different input/pointing device? By this i do not mean the use of icons/scrollbars, but a "direct" method using the available pointing device. That device would be a touchscreen capable of distinguishing 1 to n fingers.
I'm pretty sure that everyone with even the slightest clue about programming will come to the same solution to that problem.
Once you come up with that, you can of course add all kinds of fluffy wording about mobile/non-mobile device, single-core or multi-core CPU, etc. That wording does not change the actual method in the slightest.
Now it's clear to see that it is just a bloody obvious use of the available input methods. However, the situation would be different if the touch would not be "1 to n fingers" but only single-touch. Then it would be obvious that "if it could just detect more than one finger at the same time", but it can't do that. If you now go and invent a physical device that is capable of acting as a multi-touch input device, even transparent so you can put it over a display, then i have no problem with you getting a patent for _that_ particular invention.
Because while multitouch input methods were known of before, you would be the first to come up with that particular type of hardware implementation. But the actual _use_ of it should get no patents at all, because that is a pure software implementation. And for that matter, bloody obvious how to do things given actual capabilities of the physical device. Just replace mouse button 1, 2 and 3 by number of fingers 1, 2 and 3. It simply doesn't matter if that touch-device of yours would then be used on a big plasma TV, a graphics tables, a PDA or a small smartphone.
That's the thing about computers. They are universal. Program X can run on a multitude of different devices and platforms. It may not be feasible to use certain methods on different devices. Like, big scrollbars, toolbars and icons in them are of little use on a small mobile device's screen, but it's no biggie to use them on big screens. The other way round it doesn't make much sense to use certain gestures/touch interfaces on big screens, because it makes it impractical to do so (remember the light-pens?), but they make lots of sense on mobile devices.
But if you want, you can do such things, of course. It's just code. The program doesn't care much about it. That's why i see all that mentioning of "mobile device" etc. in that patent as nothing more than fluff, meant to distract the examiner from the core of what it is about: using one finger does X, using two fingers does Y. Which would perfectly translate to mouse button 1 does X, mouse button 2 does Y. I guess without all that silly fluff around it, such patents would get rejected right away.
Greetings,
Chris
ETA: To add some fun about multi-touch stuff for the electronics guys,
check out this page (or Google for "led matrix touch" to find much more stuff about that method).