Attention device makers: You can’t just bolt touch onto existing products

Finger painting is one of my favorite things to do on my iPhone (you should check out my Brushes.app Flickr set).

Ever since experiencing iPhone multitouch I've wanted a touchable computer screen, not for mundane tasks like document editing or window manipulation, but to paint.

So at Fry's Electronics I tried an HP TouchSmart computer. It did not go well…

Two problems:

1. See how the line lags behind my finger? The faster I go the more the line lags. It's not terribly noticeable in the video, but when you're using one of these in person you can feel it. I've had this problem with every pen tablet and touchscreen I've ever tried. It's even true of my beloved Brushes.app.

I'm guessing it's hard for today's software and hardware to render brushstrokes in realtime. Whatever the cause, it's vital that all input devices respond instantly so that they feel like an extension of your body. This is doubly true when the input device is your body as with multitouch screens and accelerometer controllers like those on the Nintendo Wii.

Still, the input lag is nothing compared to the usability problems I faced because…

2. You can't just bolt touch technology onto existing software and hardware. In the video, see how hard it is for me to select colors? Or how it takes several tries to touch the button in the dialog box? Or how the "close window" button doesn't work at all?

This is why Microsoft has been pushing tablets for years with zero traction. You can't just bolt touch onto a foldable Windows laptop. The software and hardware must have touch baked in from scratch.

Look at Microsoft mobile OSes. The UI conventions are all lifted from Windows — the start button, program launching, window manipulation, everything — and the experience, while familiar, is horrible.