Learning from my actions

I'm working on my first project that involves a touchscreen interactive at the moment. It's got me thinking about how a person learns to work a system with their hands.

You know how people stand at street crossing repeatedly hitting the cross button, even though (a) it only needed to be triggered once (b) hitting it more doesn't make the green man show any faster and (c) it's really bloody annoying?

I'm like that with web sites and services that are slow to respond. I click - nothing happens - I click again - and again - and then something happens - and then I'm not sure which click triggered it - and then I'm lost. Image magnifiers are classic for this - the time taken to serve up the image creates too great a lag between action and consequence, and without a loading graphic, you can't be sure anything is happening. So - click - there you go again.

I had an experience like this in a gallery over the weekend with a touchscreen interactive. I turned the sound off (because I don't like intruding on other people in the gallery) and pressed things and moved things. I could successfully 'close' things and move elements round on the screen - but I couldn't tell to what end. Possibly the sound was essential, but there was a prominent 'sound off' option. In the end, I couldn't tell who was at fault - me, or the machine. Most likely it was me.

So what I've been thinking is a very simple thought. To successfully manipulate a system that's new to you, it needs to give you prompt and quite obvious responses.

Today on Pippin Barr's blog I came across the procedural drawing tool Harmony. He'd used it to draw the face of god ....





Harmony is an online drawing tool that uses algorithms and the HTML5 'canvas' attribute to contruct an image that is a bit more than - and a bit more surprising than - what you're doodling with your mouse on the screen. Or, as Barr puts it, "What that means in practice is that you draw, and it kind of fills things in in interesting ways around your drawing."

Playing with Harmony, you learn from your experiments. You learnt that fat movements draw sketchy lines, that often die out, that slow delibrate movements build darker lines, that bringing lines closer together causes them to reach out to each other. You learn (because there is no undo, only clear) not to rgret your mistakes too much. Or perhaps you learnt to be more careful, if you want results like these.