I didn’t like the geometric shapes (called Identicon) I had as a default Gravatar image, so did a quick look around. Turns out – surprise-surprise, many tools exit that generate a random image given string (be it an email, or IP address, or just about everything). I’m on robot wave right now (as you might have noticed from my previous post), so I was so happy to find Robohash – hashing strings to robot images.
Now, beyond it’s obvious usage – to provide a picture for me, of course, the idea behind Robohash is very neat. When it assigns a robot (or a monster) picture to some text, it transforms it into a form much easier to remember for our human brains (it has to do with evolution that we are so good at recognizing faces, I’ve heard). This means that when we see the Robohash image of a familiar website we can recognize it almost instantly (faster than reading the URL, I think). More importantly, when we see the image of a malicious website that tries to play with our reading skills and imitate the URL of our familiar website we will recognize the difference in images much easier than, say, one different or extra letter in a long URL.
For this to work, though, we would need to have images next to the links we are pressing (for popular sites at least), and also to remember how the sites we visit usually look like. Sounds like a well-known issue in Informatics – trading processing time for memory. In this case, trading time to carefully read the URL for the inconvenience of having to remember, say, Google’s face. How software engineers go about the problem? They prioritize, basically – you have to give more memory to gain more speed, or loose speed to gain memory, so it all depends on what you value more.
I personally don’t mind spending these extra moments to make sure the website I’m visiting is legit by carefully reading the URL. But that’s just me and my poor memory. Maybe having website faces widely spread will make the web a bit more… humanoid? A good thing? =/
What do you think about websites having faces?