The Team's Take: Why Android Worked
Androids: The Missing Pieces, Part II
Illustration by Daniel Sandler
Like the previous article in this series, the content below was originally written for the book, Androids: The Team That Built the Android Operating System. But, also like the previous article, it was eventually cut for brevity and focus. I’m reviving it here for those that are interested in not just smaller quotes from these people, or in my analysis of what they said, but in their more complete explanations from my conversations with them. I hope you enjoy hearing from them as much as I did.
As background for the book Androids: The Team That Built the Android Operating System, I conducted interviews with most of the people that worked on Android prior to 1.0. These were free-form conversations in person, on video chat, or over email, but there was one question I always asked at the end: Why do you think Android succeeded?
I provided my answer to that question in the final chapters of the book (or, really, throughout the entire book), and summarized key aspects in the first article of this series. But I so thoroughly enjoyed hearing a variety of well-articulated answers from the people I interviewed that I thought it was worth sharing some of those answers verbatim.
Here is just a small sample, from three of the people on the early Systems team: Brian Swetland, Ficus Kirkpatrick, and Rebecca Zavin. I may try to share more in future articles, since all of the interviews offer a unique take on things. Then again, that’s kind of why I wrote the book, and maybe four years writing it was long enough. So no promises.
Why do you think Android worked?
Brian was the first engineer hired onto the team, seven months before Google acquired the tiny startup. He got the initial kernel running and led the growing systems team through 1.0 and beyond.
I never believed that Apple would win. I don’t believe they can because of their Model T — you can have it any color you want as long as it’s black.
The greatest thing about iPhone: there was a lot of vendor interest because people were worried about how to compete with Microsoft and stuff [before the iPhone]. [Then our] phone rings off the hook after Steve’s [Steve Jobs] thing. Because Apple’s not going to license you any of that shit — now what do you do?
I was pretty sure it [the prevailing phone platform] would be something opensource-y, just because of backlash from Microsoft after the 90s and people being paranoid about that. But why not Maemo, or all these random Linux things? Because the desktop Linux stuff that became those platforms are the most anti- “product” things in the world. They make the fundamental mistake over and over of Win CE, where Microsoft took the Microsoft UI on the desktop and squished it down onto the phone. So many of the Linux [platforms] tried to do the same thing. And people don’t want that in a phone. It’s a platform vs product thing.
The other reason why people choose Android vs Maemo or whatever [Brian referred to a diagram from one of the slides from the Android pitch deck, shown below]:
Slide from the original Android pitch deck showing the TI Linux-based solution at the time (2005)
You can use TI’s OMAP chips to build a Linux phone. You need TI’s OMAP and then forty components from forty different vendors of middleware, and you put all of these together and you integrate them all and then you’ll have a Linux phone. And that was just absurd.
Android aimed to build the full stack. Now you can take it and build a phone with it. And it was actually possible, a real phone that did things.
And sure, Android has its warts and it’s not perfect — nothing’s perfect. But it gets the job done. While everyone’s thrashing around with random middleware, Android made some decisions and built a product around it. They didn’t focus on how to replicate the Linux desktop, because Linux desktop is not what people want on phones.
[Also:] Droid helped. Verizon put some significant amount of cash behind the Droid campaign. And that obviously helps. You get TV commercials and billboards. The whole Droid Does campaign.
Ficus was the last person hired by the startup; a week after he started, Android was acquired by Google. He worked on kernel drivers, like one for the camera, but eventually crawled his way up the stack, working on messaging and eventually (post-1.0) Android Market.
Anytime anything works out, if you don’t acknowledge the huge Luck factor, you’re kind of a jerk.
It’s a continuum. I certainly wouldn’t say it’s 100% luck and we were just standing in the right place at the right time. But I think all of us need to acknowledge that some amount of this was being in the right place at the right time.
I struggle with this — how much was luck and how much was brilliance or cunning. I just wanted to acknowledge that there was a lot of good fortune. It’s not terribly interesting or insightful to talk about.
The cases where the decisions we made played a difference, which is the heart of this question. I talked about this in the podcast [Android Developers Backstage, episode 56]: the partnership based approach is the fundamental thing that made it work. By setting up an incentive structure, many more people than our team or Google were incentivized to make Android successful. It was essentially a recruiting strategy for getting so many of these OEMs and carriers behind Android. Again, the fortune part there was that there was a gap there to be filled. So they were open to hearing that message.
Also, just being pragmatic. Android had a very deadline-based culture. People who work on services software don’t understand that you can’t say that it’s done when it’s done because there’s $100 million supply chain at Christmas, so you’re going to ship whatever you have on that date. Very few people, maybe video game developers, can understand this principle. All of these things we dither and talk about like mistakes or regrets are ultimately dominated in importance by the fact that we just got something out. And we were able to keep things marching forward by, maybe not knowing when to put our pencils down, but being forced to put our pencils down was a really important factor.
The last thing was: we worked hard. We worked really hard. There were times in the first four or five years for me where it was essentially all I did when I was awake. I know I was not alone on that.
Rebecca was on the systems team, working on the kernel, drivers, and the first Droid device. She was one of the few people on the early team to transfer from elsewhere in Google. She worked on the platforms team at Google, which turned out to be good experience for her work on the Android Systems team.
On the systems team, we got a bit of a window into the OEMs and how they were functioning. So when we did these engagements with HTC or Motorola or whatever, even the silicon vendors like Qualcomm and nVidia, we got a pretty good sense for organizational dynamics there. And I can tell you that a company like Motorola or HTC had like 6 OS projects going. Everybody had some kind of Java on Linux thing already, and Symbian, and a feature phone thing that was different.
It was kind of a mess.
You could really see the bureaucracy, weird stuff like “Oh no, that’s the kernel team, not the systems team.” Weird pockets of different fiefdoms in the company that couldn’t interact properly. So there were a lot of resources spread really thin at the OEMs at the time. These guys basically cut their losses and went full Android.
[Also, Android] lands at exactly the right moment in order for these companies that suddenly are going to have to move to smartphones. It gives them something to stand on, versus rolling their whole own thing from scratch that they just weren’t equipped for. One minute you’re a calculator and the next minute you’re a computer. The modem is complicated, but you’re mostly getting that directly from Qualcomm or Motorola or whoever. But the rest of the software is buttons and a menu. The equivalent of a cordless phone today.
So it was really really good timing.
We were really ready for it. Microsoft had Windows Mobile. I don’t think anyone was really compelled by it. And it was expensive. And there was a lot of bureaucracy involved in doing it. For everybody.
Google came along and said, “I’m going to give it to you for free. You’ll just take it and run with it, it’ll be a much lighter touch thing. I’m not going to have this huge bureaucratic engagement program. It’ll be easy and we’re going to do something aggressive.” And I think Google was uniquely positioned to be credible, like they’re actually going to do something.
Android pre-acquisition, that would have been a pretty tough sell. But you’ve got Google behind it, and the ability to apply these Google resources around, the Google applications, and Search, and Gmail, and Maps. Maps became huge around the same time. Today, one of the primary use cases for a smartphone is location. I know where you are and I can help you get places and find the things you want when you’re not at home and connected to the Internet.
All of those things collectively, kind of a mix: Everybody needed something suddenly, and there was a cheaper, better, faster option.
So, in Summary…
Ah, heck, read the book. That’s why I wrote it. But in the meantime, I hope you liked being a part of the original conversations that led to the story. I know I did.
If you find this topic interesting, I’d invite you to check out the book, either in paperback or ebook form from No Starch Press (soon available on Amazon, other online print and ebook stores, and even orderable from your local bookstore as well), or in the upcoming audiobook from Tantor Media. All profits from all formats are being donated to tech-related charities.