Skip to content
 

Automating with… Screenshots?

This thing is absolutely awesome. It gives me the feeling of living in the 21st century and shows how we are going to interact with computers in the near future. Well, as soon as it gets “production-ready” and voice recognition is applied. But that’s a different story.

So what is it all about? Let’s say I need a timetable of the bus going to the university. In my reality these are very likely to be outdated right the next day I checked them. So instead of looking it up every morning, half-sleeping and struggling between drinking coffee and eating breakfast, I could ask my computer to send it over to my phone. Obviously, you could do that with general technical knowledge and some experience in OS scripting, but why to bother? There are much more interesting things to hack nowadays, right?

What is the alternative? You open up Sikuli, grab some screenshots, add a few lines of Python code (if you want to) and you’re ready to go. Yes, it’s that easy. Here’s the effect:

And here you can see a demonstration of sending PDF with the timetable:

Experimenting with Sikuli from Robert on Vimeo.

The sexiest thing about it is that you can use… Python Unit Testing! And that makes it really powerful tool (like e.g. Selenium). For instance, imagine distributing Django apps with GUI tests already there…

Leave a Reply