Simon’s Blog

I see from Techmeme that today people are discussing the latest “innovation” in Windows Mobile 6.5 – the honeycomb touchscreen.  I can’t get to the original article – the site appears to be down right now; however I looked at the discussion.   It seems that I’m not quite so impressed as others; for a couple of reasons (btw – as a caveat, I should say I can’t be sure about the details of the MS software implementation, because I haven’t seen it.  It may be that the MS implementation of this idea is materially different from the IBM implementations, such that the points made below do not apply to MS ).

Firstly, IBM patented this in 1996, so it’s hardly anything new or original (see the patent at the US Patent Office).  From the IBM patent, the invention is summarized as:

A computer in which entry of data from an area provided on a display, touchscreen or the like is facilitated by the provision of a “soft keyboard” having keys configurable as hexagonal cells. Hexagonal cells enable display of larger areas for engagement by the finger or other input pointer used by a user of the computer of this invention.

Pretty clear, I think. Of course, while the fact that IBM already patented this invention more than a decade ago renders this feature in Windows Mobile 6.5 uninnovative (is that a word?!), it doesn’t make it a bad idea.   Which brings me to my second point; because I don’t think this is that great an idea.   The IBM patent throws some more light on the reasoning behind the invention.  From the IBM patent…

Wasted space in soft keyboards is a serious problem, especially on small screens such as are provided in mobile client systems and especially when there are a lot of keys (e.g. alphanumeric keyboards)…. Hence, soft keyboards need another shape for keys besides the rectangle. The new shape must have the following properties… The new shape should approximate a circle or ellipse as closely as possible. It should avoid ninety degree corners like a rectangle since this is what wastes space… The new shape should interlock so that space between the keys is not wasted.

This section demonstrates clearly both the reasoning behind the invention, and also why the idea is, in my opinion, not so great.    Here’s the mistake I think both IBM and Microsoft made – ” The new shape should interlock so that space between the keys is not wasted” Errrrrrrr…  NO!!  Space between keys is NOT wasted space.  Quite the opposite in fact.  Space between keys is actually vitally important in keyboard/keypad design because it reduces “user interface errors” i.e. it reduces the number of times that the user hits the wrong key on a keyboard/keypad by mistake.

IBM almost got it right when they said, “The new shape should approximate a circle or ellipse as closely as possible.” However,  it shouldn’t approximate a circle/ellipse (e.g. by making it a hexagon), it should actually be a circle/ellipse.    With circular/elliptical shape keys on a touchscreen, the usable key surface area is maximized, while simultaneously enabling space to be left between the keys.

Just thought I’d mention…  Anyway, if you’re a maker of mobile devices, you’re very welcome to have that idea for free   Not that I’m claiming that pointing out  that close-packing keys on a touch interface (hardware or software) is a bad design idea, is particularly insightful.  I’d just be happy to see some people (other than folks at Apple) implement some good ideas relating to mobile handset design for a change… ‘cos I tell you, there are some shockingly bad handsets on the market at the moment…