Lessons from TV Land: My years as a front end developer for TVs

When I first started freelancing back in 2008, I had a part time contract with the global television manufacturer Sharp Labs, primarily designing and developing apps for this very new thing known as “Smart TVs”.

As I worked there, the market seemed to radically shift month-by-month, with new competitors arising, and falling. Roku, Apple TV, Samsung, Vizio, the list wasn’t long but everyone was vying to define the space and saw it as the next big thing. Everyone was launching their own app stores, as we all saw the power of Apple’s App Store rise and become wildly successful.

You Know What They Say About Assumptions

There was common “knowledge” about what people wanted out of their Smart TVs, which is fairly suspect given that they had only existed for a year or two and very few people had them. Viewers (not users) wanted to just sit back and let it wash over them. YouTube and Vimeo’s TV apps both play in what’s known as living room mode, which is less about searching and more about discovery. Don’t make me think at its finest.

I found the whole thing suspect for the same reason that you should feel alarm bells going off if someone tells you that mobile users are always “on the go”. As if we’re all New York executives, jumping into a taxi, just trying to check our flight status and nothing else. I’m pretty sure the web industry has learned that you can’t assume what people will and won’t do with technology; I browse the web on my phone, even when my laptop is an arm’s reach away. And I know I am not alone.

On the go! So important! -- Click for photo credit.
On the go! So important! — Click for photo credit.

The way we interact with technology, and especially TV and streaming media, is undergoing a huge shift. Many people are no longer watching shows when they “air” on TV, and instead opting for a DVR or streaming media like Netflix (not to mention the shows that are ONLY streaming). In this context, the viewer must choose said media, and suddenly they are, actively searching and recording and saving. DVR and on-demand interfaces are nothing new, and I think we can all agree that these systems have a long, and storied history of absolutely sucking beyond reason.

There are folks out there re-thinking how we interact with media—Windows Media Center has been in this space for a long time and actually has a pretty nice interface (at least the last time I looked at it). And that’s not about just sitting back and setting Bill Gates wash over you; it’s about finding the thing you want to do and doing it. Video game consoles have been winning at this forever.

Progressive Enhancement, Accessibility, yeah probably

People say that users don’t really browse the web on their TVs. Of course users don’t want to browse the web on it or do anything other then watch a video–trying to navigate the web on a TV is pretty much near impossible. No one does keyboard support well. No one. And especially not for D-pad (the arrow keys). Sounds like the old “We don’t need to optimize for mobile, we don’t get much mobile traffic”. Maybe because your site is unusable on mobile?

When we first launched the Smart TVs, the browsers only supported a subset of ECMAScript 3. A subset. Any site today depending on modern JavaScript is likely unusable in those TVs. I still have one of them, and I will probably have it for a long time as it has hardly seemed to age (other then that Twitter application I wrote that is defunct thanks to API changes). TVs are not like phones; people usually don’t upgrade their TVs every two years, and unlike computers, they also don’t upgrade their browsers, the manufacturer controls software updates. And they ain’t the Chrome team.

Will Responsive Save the Day?

You may be thinking to yourself: my site is fully responsive, so it will look fine on a TV, right? If Brad Frost’s site can look good on a watch, certainly my site can look fine on a TV. My answer to that? Define “good”.

We designed everything as if the viewer is a mostly-blind 80 year old with an 800×600 monitor. It’s all about the “10 foot” UI—can you read it from ten feet away. Certainly most web sites fail this test, and it’s not like you can use media queries to target TVs. While the screens are physically large, the pixel width is not, so CSS media queries will not help you here.

A good way of thinking about this is presentation design: think about how large the type is on a Powerpoint slide. Real big. Not many words. Simple. Now imagine your user has to navigate this slideshow with a remote control. They have up, down, left and right. Were your website actually just a slideshow, then that would probably work out just fine. But I’m guessing it’s not.

So this begs the question: should we be expected to make our sites work well on TVs? I’ve never heard of client asking for it, we’ve never claimed support for it, and QA would be an unfathomable blackhole. The implementation of browsers, at least on first generation Smart TVs, was not in step with what’s required to be a standards-compliant, up-to-date browser, which is makes it impossible for the web community to support.

There are more and more web-enabled devices, consoles, TVs and other unknowns. Many of them lie outside of the general working knowledge of your average front end web developer. What does that mean to us? Should we care?

I can’t answer that question for you. Certainly progressive enhancement is part of the solution. I think this circles back to the the thing that makes the web so amazing: that so many different types of technology can access it and process it. That a watch and a TV can use the same code and display it is just crazy. It’s crazy cool, but it’s also pushing the boundaries of what responsive is and can be.