Swype demos at TC50- this is nothing new people


The TechCrunch 50 is ongoing on the west coast and I’ve been watching to see anything that looks exciting and new that is launching out there.  I started seeing excited coverage about a "new" text entry method for mobile devices called Swype.  Swype is a gesture-based entry system where you spell words by dragging your finger or stylus over a keyboard.  The system then knows what word you spelled when you lift your finger.  A lot of people were talking about Swype today and I saw Robert Scoble going crazy about it over Twitter.

The whole time I was seeing this talk fly around the web I realized that I have seen this technology before.  I felt I must be wrong as Swype is new and developed by the inventor of the T9 phone entry method.  The more I read descriptions about Swype the more convinced I was that I had seen this technique before and when I watched a video of the Swype presentation at TC50 I knew I was right.

Swype looks just like the text entry method developed by the IBM Research Center that was originally called SHARK.  I reviewed SHARK back in 2004 and was very impressed with it at the time.  It works mechanically just like Swype appears to work but SHARK predicts words based on the shape of the overall spelling gesture.


SHARK was renamed to ShapeWriter in 2005 and is still around today.  Look at the image above that shows how words are spelled in SHARK.  Take a look at ShapeWriter in action below and then look at that Swype video again.  This is nothing new, people.  Maybe TechCrunch should have researched Swype more diligently before letting them in TC50?


UPDATE:  ShapeWriter is also being developed for the Google Android project and was one of the top 50 applications recognized for that work months ago.  Here’s a video of that project:




Some of these comments about parallel development are true. I’m one of the founders of Swype. If you look at our original filing date of the Canadian patent, you’ll see it is well before SHARK/Shapewriter. They showed it publicly (many times, actually, at conferences, etc) before we did, but just because you saw it first doesn’t mean it actually came first. I’d encourage folks to try both and then judge from there.


Actually, if you do a little digging, you’ll find that Swype patented the idea before IBM came up with Shark. So Swype was first. Whether IBM copied Swype, who knows. Parallel development does happen.


Given how long IBM are takig to getting round to launching a solution for the TabletPC, I’d be happy for a competitor to get something to the market.

Especially since you can’t get hold of the old Shark software anymore.

The SlideIT software is only Mobile Win. Though I id like the PC demo. Shame yuo can’t copy/paste from the demo or I might have actually used it as a TIP.

In the meantime,I’ve install Inscribe and done s custom keyboard for it that looks like the ATOMIC one from IBM to get some practice in. Not pattern recognizing,(you have to pause or change direction on each letter) but at least I can get familiar with the key layout…


Henrico Dolfing

I placed a video of all three (WritingPad, Swype and SlideIT) on my blog. Spot the differences… I could not find them.
I bet the underlying algorithms are different, but this story is going to be continued when it comes to the intelectual property question.


I haven’t heard of this sort of thing before. This is exciting and will probably prove to be a boon for small gadgets. I currently find cursive more useful they picking at keys with a stylus on my Tilt – and I am thumb typing challenged.
I will agree that someone adept at thumb typing may not find this so useful.
Given a choice though, new adopters, like kids getting their first phone, will probably adapt to this quicker than thumb typing.


im not surprised that parallel development happens a lot in software…

Comments are closed.