Arnaud Coomans once built a version of Instagram that ran on a flying toy house.
He called it Upstagram, a nod to Up, the Pixar movie about a guy who takes his home aloft. Adorned with toy balloons, Coomans' creation flew over Paris, snapping photos and uploading them straight to Instagram for all to see. Coomans and his cohorts didn't get approval from Instagram, but the stunt worked.
"Basically," he says, "we hacked Instagram."
Coomans is an Instagram employee now, an engineer at its offices in Menlo Park, California, within the headquarters of parent company Facebook. But his work hasn't changed all that much. Once again, he finds himself hacking together a new version of the photo-centric social networking app.
This time around, he's squeezing it onto the Apple Watch.
This isn't quite as adventuresome as breaking into the Instagram API and rebuilding the app for a sky-high Raspberry Pi. But it requires at least as much creativity. As the first wave of reviews show, the Apple Watch isn't an iPhone that sits on your wrist. It's something that works in tandem with your iPhone, something that depends upon a tiny screen and its own controls, something that's tethered to your body and suited to shorter, simpler tasks involving even less of your attention.
As Coomans and his colleagues Ian Silber and Ryan Nystrom explain, Instagram's existing iPhone app is entirely unsuited to the Watch. The Watch doesn't include a camera, and no one's terribly interested in browsing a long stream of photos on their wrist. When Apple unveiled its much-anticipated wearable last fall, the Instagram braintrust knew it needed an app on the device (call it the power of Apple). But this required dreaming up a new kind of app.
They settled on a simple idea, and as Coomans and crew built it, they found themselves paring it down to something even simpler, discarding so many things that would have made sense on a phone. The result is a Watch app that flashes photos when your favorite people post to Instagram, and provides a simple way of responding to those photos. You can send a "like" or a few emojis. That's pretty much it.
Coomans and crew are among the few outsiders who have built Watch software and actually tested it on the new device, which is slated to hit the market at the end of the month. Their experiences show just how different this device is, that it's best used as an adjunct to other devices, and that the world's developers will need a bit more time to determine just where the thing can be most useful. The trick, Coomans and his colleagues say, lies not just in reshaping existing apps for the device, but in trying to find entirely new applications.
When Apple CEO Tim Cook unveiled the Apple Watch in October, Silber and six other Instagram designers sat down to brainstorm ways they could use the device. At the same time, elsewhere in the building, company co-founders Kevin Systrom and Mike Krieger kicked around their own ideas. Ultimately, both groups settled on the same notion: a notifications app. "What if you could know right when someone posts---and not just know that they've posted, but be able to see it?" Silver says.
For Silber, it was a validation of the idea. Two groups---independent of each other---reached the same conclusion. But it also shows that for existing apps like Instagram or Facebook or Twitter, the realm of possibilities on the Apple Watch are slim.
The Watch is designed, in large part, to juggle notifications for existing smartphone apps. It also makes sense for tracking your exercise, paying for stuff, presenting a boarding pass at the gate, perhaps even hailing an Uber. Beyond that, the applications aren't immediately obvious. As they built their notifications app, Silber, Coomans, and Nystrom explored other ideas, trying to expand the scope of the project. But they kept returning to that core idea. "It was a learning experience," Coomans says.
This is partly because Apple's coding tools don't allow for a whole lot. As Coomans and Silber explain it, Apple has restricted what you can build, either to reduce the possibility of bugs or to focus the purpose of the Watch. "You expect some things to be there and they're not," he says. They wanted to drive "likes" with a double-tap, for instance, but that's not possible.
What's more, they quickly realized they weren't really building software for a watch. They were building software that runs on an iPhone and wirelessly serves stuff to the Watch. That restricted them too. "Any code that you run---there is a slight delay before it reaches the screen," Coomans says. "You have to take that into account."
So, they thought about a tool for exploring photos across the Instagram network, as you would on your phone. And they considered a tool that would "discover" photos other people were posting in your general geographical vicinity. But ultimately, this stuff didn't make sense. They also created a way of viewing your Instagram profile, but even that didn't fit. As Coomans explains it, you'll spend seconds doing something on the Watch, but not minutes or hours. "It's a short-lived experience," he says.
Ultimately, creating the app was more about pruning than building. "We slowly cut down more and more," says Silber, "until it was basically the photo and who posted it."
The app does provide a way of commenting on photos. This too is simple. You can tap a "like" button or use a kind of mini-custom-keyboard that lets you quickly send what amounts to "emoji sentences." The team toyed with the Watch's voice controlled interface, but they say the keyboard is faster. And the app, like the Watch itself, is all about speed.
That may seem modest. But that's the point. They not only cut the interface to the bare minimum, they limited the number of notifications it sends (you can choose who's on your favorites list). You don't want too many interruptions hitting your wrist.
Silber and crew are happy with the app they created. They believe it's useful---something above and beyond their existing app. Mike Wehrs, the head of US operations for Appster, an app builder that is also exploring Watch apps, agrees. He frowns at the idea of receiving all social media updates on a watch, but if you pare things down, he says, it can makes sense. "You need another level of sorting," he says. "You need to limit things close friends rather than just friends."
Does an app like this change your world? Maybe. Maybe not. But if all your favorite apps do something similar, it might. The big lesson from Silber, Coomans, and Nystrom is that the coders of the world need more time. "It's similar to when the iPhone first came out. Now, there are companies, like Instagram, that exist just because of the phone. That might be the case with the Watch too," he says. "We'll see what happens."
The main difference is that app makers shouldn't think big. They should think small.
