Mobile and tablet apps change all the time, but there is no public record of the way an app’s user interface evolves with each new revision. What we need is a version of the Internet Archive’s Wayback Machine for apps, but unfortunately because of the siloed nature of this class of software, it’s not possible to simply deploy bots to create one for us.
It occurred to me that one viable alternative would be to crowdsource something similar to the Wayback Machine by creating an app that would let any user upload screen grabs to a central archive on the Web. That sounds much more manual than the Internet Archive’s approach, l know, but in fact I think it could actually be fairly well automated.
Introducing the Appback Machine, Hopefully
Here’s what I’m thinking. Contributing users — basically anyone who wants to take part in the project — would simply take screen grabs from their favorite apps, as many as they want. On iOS, that can be done from anywhere within the operating system by pressing the Home button and the power button at the same time, with the resulting screen grab getting dropped into the Photos app where it’s available to any other app that cares to access it.
Our hypothetical archiving app — let’s call it “Appback” for now — would simply be a tool for getting those screen grabs from the Photos app and uploading them to the Appback server. Easy, right?
Wait, don’t the screen grabs need to be tagged and identified with the name of their originating apps? Yes, they do, but this is where the automation magic would come in.
First, it would be pretty simple for Appback to discern which of the pictures within the Photos app are screen grabs (as opposed to ‘natural’ shots taken with the phone’s camera), and pass that information along to the server.
It’s even well within the limits of current technology to analyze the picture data from each screen grab and make a guess that, say, these screen grabs are from Foursquare, these are from Instagram, these are from Angry Birds, etc. This sort of identification could be done on the Appback server, and if there’s any ambiguity, the server would just send a message right back to the user and say, “Here are the two or three apps we think this is from, can you help us identify which one it actually was?”
If there are no matches, the user would be asked to type one in, with automatic matching against names from a database of known apps from the app store. Over time, as the corpus of identified screen grabs grows larger and more accurate, this disambiguation process would happen less and less often across all users, reducing the contribution process to simply a matter of uploading screen grabs.
To motivate users to contribute, Appback could institute some kind of points or rewards system, so the users who are first to post screen grabs for a new app or a new revision would receive recognition within the Appback system. With enough users contributing, it wouldn’t take long at all to build a fairly robust, Web-based gallery of screen grabs for the most popular iOS apps, which would be a huge benefit to app designers and developers everywhere.