How Google bought Android
arstechnica.com
arstechnica.com
Also, not sure if it's covered elsewhere in the book, but it would've been interesting to get to understand Google's motivation for doing this deal. My understanding is that Google understood mobile was important and given that, at the time, network operators had a lot of control over mobile software, the fear was that they could control access to the search engine and, hence, exclude or take control over Google.
> This was back in 2005 or so, and we were in meetings with Android, and this was before Android was a thing... they had a prototype of their Android phone, and this was before iPhone... but we were debugging something with SQLite and we were plugging into the phone and we were running the debugger on a workstation which was pretty amazing. Nobody else could do that... here we were, we were debugging an application in GDB on a phone that was on the public network, and this was utterly mind blowing. Nobody at Motorola, nobody at Symbian, nobody at Nokia had anything close to that, and in that one moment, I knew that Android was going to be huge.
Not to mention that in 2005 Nokia already had the 770, which was basically a mobile desktop GNU/Linux device, with Gtk+, Gnome and everything. You could run gdb on the device itself.
To the person who made the claim, it was the first, because of their limited view of the world. I always take claims of being first with a grain of salt for this reason.
That said, I completely agree. I definitely used remote debuggers and on-device debuggers on mobile before Android/iOS existed.
Then one would say 'it was the first i've seen'
We have language for a reason. Despite what you hear on daytime talk shows about 'your truth', back in reality, actual truth is objective
It is the first they know about. It may be defined differently.
You cannot claim absolute truth on most things you can't can't claim absolute understanding of the language people use to express things.
If you expect absolute truth in a variable world you're in for disappoinment.
From the description it sounds as if gdb was running off-target (same as it would for debugging Palm), except perhaps it connected via gdbremote-over-TCP instead of a 68k debug stub via RS232 or similar. And in this case, the phone's TCP stack is running over the phone's radio.
> basically a mobile desktop GNU/Linux device
From the sound of it, this device likely could have matched these claims.
> ...whereas the engineers that other companies, they had the big breadboard prototypes, the big full sized prototyping board, and the phone would run on that, and it was not connected to a radio so they couldn’t actually use it as a phone.
(he clearly says "plugged in" in the audio track, which does not hint to a over-the-air solution).
According to the people who deviced [no pun intended] the 770, it was never meant to be a product. It came originally as an internal prototype / developer machine, to make ongoing evaluation and development for maemo platform and its common applications easier. Nokia was obviously planning a polished consumer tablet. Someone high up decided the investment in hardware and supply chains for a developer-only system was not financially sensible, but if it was a product that could be sold to cover at least some parts of the setup cost, that would change the picture. (Incidentally, I believe, but can not prove, that it was also to gauge interest in tablet devices in general and to lay the ground for the future Nxxx series market.)
Let's keep in mind, Nokia's consumer facing business was a PHONE company. We'll get back to that.
The first consumer grade Nokia tablet was the N800. I vividly remember fixing its media player UI and working around various set-in-stone roadblocks in the GTK+ middleware. I also remember how our entire team were on a crunch schedule to get the feature completeness and bug rates to an acceptable level for going gold. The launch date had been set. We actually were told a few weeks before the internal deadline to stop working on all changes: our latest weekly progress report showed we had passed the launch requirement and the PMs didn't want to risk regressions from otherwise benign bug fixes.
N800 had an FM radio. It had wireless internet and Bluetooth. It had a surprisingly functional front-facing video camera. It did not have phone capabilities. Neither did its upgrade, N810.
The N8xx devices were incredible pieces for what they did, at the time. Techies loved them. The rest of the world didn't. Nokia was a phone company, and consumers associated the brand with the ability to make and receive calls.
The first member of that family to have the missing capability was N900. Based on rumours and drunken vents in a sauna, the certification process to get that type of device with a baseband modem accepted was not a pleasant experience.
----
To this day, one of my proudest achievements in the maemo projects was to hunt down the root cause for a really nasty UI lag problem in the media player. Nokia was a design-heavy company, and someone in the design team had mandated a very specific type of GTK+ element to use in the problematic section of the player UI. The element did not support the necessary actions or properties, so in order to satisfy both the UI specification and the latency requirements, Nokia would have had to patch in a whole load of additional functionality to the element itself. This was a public API, and supporting that kind of bastardisation in a core library would have been an absolute nightmare.
This bug was present in the run-up to N800. It was still there for N810, and it got bounced to me because I had argued about the infeasibility before the N800 launch. Once I had laid out the full technical details of the root cause, with a list of possible ways to fix it and the associated costs of each, I got the most satisfying resolution:
"WONTFIX. We do not dare fix this bug."
"The Untold Story of SQLite"
If nobody at Motorola, Symbian, or Nokia could do (specifically) that, it just means that they didn't have GDB ported to their OS. But they absolutely certainly had other remote debugging tools; it's preposterous to suggest they didn't.
This is also why I personally believe Microsoft failed but Google was ultimately able to succeed in competition with Apple.
[0]: https://www.google.com/books/edition/Losing_the_Signal/oE-TB...
1) their "device simulators" could not be relied upon
2) certain devices running certain versions of their OSs would have APIs function or not (i.e: a library call may work in one OS version and not in another, but only on one type of device)
3) every time you wanted to run any code on a device, it needed to be signed by blackberry's signing servers - which were often down or slow (waits of 40-60 seconds were common). (it no longer exists, but we had http://isthesigningserverdown.com to help us know)
4) sometimes the devices just restarted, either when you tried to load something or sometimes when you made certain API calls. Blackberry devices took MINUTES to start.
5) logging infrastructure was basically non-existent... logs went to a hidden log page, (ctrl+lglg to access it and view/scroll on the bb device) - we wrote our own network logger lib to basically `tee` to the builtin logger and back to our desktop consoles.
the first 3 points combined meant you often had to iterate on real hardware, waiting 40-60 seconds to deploy each time, only to have it explode in your face by restarting and then frustrating you by not being able to figure out what was going on.
I'm sure i've forgotten much more.
https://twitter.com/stroughtonsmith/status/13567772811889377...
>Google worried that if Microsoft made it hard enough to use Google search on its mobile devices and easy enough to use Microsoft search, many users would just switch search engines. This was the way Microsoft killed Netscape with Internet Explorer in the 1990s. If users stopped using Google’s search engine and began using a competitor’s such as Microsoft’s, Google’s business would quickly run aground.
https://www.theatlantic.com/technology/archive/2013/12/the-d...
And Bill Gate blame it on Anti-Trust and monopoly lawsuit against them.
It was. I used it. They ported the Win95 desktop (right down to the task bar and start button) to WinCE and a screen the size of a couple of thumbnails. The GUI was horrid, but evolved rapidly. At the end WinPhone pretty, very uniform even across 3rd party apps, faster, and used less power than either iOS or Android. I'd say they executed WinPhone GUI framework extraordinarily well.
WinCE on the other hand never worked reliably on any device I came across. They stuck with it till well after iOS and Android had left them in the dust, and then transition to NT made the Python3 transition look well done - you had to bin your old device. WinCE blue screened more than Win95; a remarkable achievement. This was true for Microsoft's but doubly so for OEM's as MS made getting the source as exercise in pain, so debugging why an interrupt disappeared up it's own arse was basically impossible - you spent 1/2 your time reverse assembling MS's machine code so you could decode the stack.
I suspect the fate of WinPhone can be put down to that. iOS and Android went with robust, mature base OS's (BSD and Linux), that came will all the batteries required in the box and had all their rough edges knocked off by being ported to every platform under the sun, where Microsoft stuck with their home baked embedded OS, WinCE, which felt like a multitasking version of MSDOS / CPM, and they had to develop everything themselves. By the time they fixed that mistake, it was too late.
I'd say open source crushing all comers in the new mobile space is when it came of age, which is to say everyone from C levels down decided it was here to stay. That had nothing to do with technological superiority. When the largest software corporation on the planet pitted their embedded solution against open source, and was wiped from the floor, even a banker could see what the new lay of the land was.
But that still really would be the antitrust suits themselves since 1) MS got themselves into that mess in the first place and 2) Bill Gates shouldn't have let it distract him so much from potentially vital projects. 3) employee accounts of the internal politics at the time pretty much meant that most new products were doomed to fail from the outset if management on the Windows teams thought they were even the least bit of threat to them.
Gates even admitted he messed it up and lost $400bn with mobile operating system. https://youtu.be/W5g4sPi1wd4?t=709
There is a different story/side to this statement.
Fixed that for you.
Exactly.
> ... it would've been interesting to get to understand Google's motivation for doing this deal. My understanding is that Google understood mobile was important and given that, at the time, network operators had a lot of control over mobile software, the fear was that they could control access to the search engine and, hence, exclude or take control over Google.
That's sort of hinted at in the article:
When Android met with Google, Larry Page observed that it would make sense for Google to acquire the small company, to help them build a platform that would enable Google to enter the mobile market.
Perhaps we all need to buy the book to find out ...
Motivation was that Microsoft at the time was developing Windows Mobile or whatever it was called and Google got scared that they will put Microsoft Bing as a default internet search engine on millions of Windows Mobile phones that were suppose to conquer mobile phone industry.
It's funny how Steve Jobs was mad at Google because Android was chipping away Iphone's dominance but in the reality Android was about and against Microsoft not Apple and Iphone.
Search for an article on HN or Google I can't find it atm. But the article talks about just like I said how Google was developing competition for Windows Mobile then famous 2007 Iphone presentation happened and Google engineers realized touch UI and UX is the future not QWERTY plastic keyboard and then they pretty much copied most of the Iphone and iOS features.
Edit: here is the article: https://www.theatlantic.com/technology/archive/2013/12/the-d...
The way he explained it, the mobile phone industry in many markets, particularly the US, used to be such that the customer was owned by the carrier.
iPhone changed the status quo in 2007: Apple could pick and choose carriers on their terms, the smartphone class device became a primary network interface for the customer, and the customer was effectively owned by the device manufacturer. This was a huge threat to Google, but also the carriers.
Android launched in 2009 as Google's hugely successful ally in their response to commodify their complement.
Google politically aligned themselves with carriers and the status quo and maintained a 'half-open' rapid-change policy which has created constant API and language churn. Great potential futures from recent smartphone history like Samsung Dex (phone-based dockable Linux workstations), wireless mesh networking as a first-class connectivity paradigm and standards-based IOT control never had a chance to mature.
By now we could be running UUCPesque media feeds over local ad-hoc wifi with free educational resources, reliable dockable workstations in our pocket and a self-organizing community economy rewarding social and environmental values. Instead we pay for VPNs and watch TikTok while our right to repair, understand or modify are wholly eroded under the auspices of an "open" platform shepherded by well-paid corporate lawyers and doublespeak.
The phone industry is absolutely terrible now. It's like a conspiracy against society focused on capturing consumer attention, surveilling consumers, using them as intelligence nodes for building global wireless infrastructure maps and media OSINT and actively preventing off-carrier cooperation among the population. We desperately need open mobile hardware.
There's no money in it. It's dead on arrival.
I'd be glad to be proven wrong.
A commercial challenge is that everyone wants the slim integrated thing as a faux-fashion accessory and there's significant costs and there are power and waterproofing concerns with mobile hardware which are technically more difficult to guarantee with a modular architecture.
OTOH, as consumers increasingly demand a right to repair and get sick of forking out for the same phones with slightly better specs every 2 years, it might just begin to work out...
You can build your dream on Android right now. It provides all the basic infrastructure to make it possible. The reality is that not enough people share your dream to make it viable.
Never doubt that a small group of thoughtful, committed citizens can change the world; indeed, it's the only thing that ever has. - Margaret Mead
In what way? Everything you said can be built on Android right now.
> Attempting to simultaneously cajole multiple large-scale device manufacturers to add a non-standard addition at breakneck hardware and OS release pace would be virtually impossible.
This is not necessary to implement the idea.
> The fact that drivers are bundled with OS releases thus making third party system upgrades extremely tedious to create for any given hardware revision is not an accident
You don't have to use that hardware. Android supports hardware that has open drivers.
Anyways, Those who appreciate the openness of Android it seems like Brian Swetland[1] is the MVP in that regard.
>From Korea, the team flew to Taiwan, where they met with the CEO of HTC, Peter Chou. Nick remembered the meeting: “Peter mentioned something about exclusivity for our first device, which Brian overheard. By the time we got back to our hotel room, Swetland had threatened to resign because ‘I didn’t join Android to become another Danger.’ I was concerned because Brian was so critical to our success, but when I saw him the next day everything was fine.”
>(Swetland said, “I don’t recall the discussion, but certainly believe it could have happened.” His memory of Danger was fresh and strong at that time. The dynamic at Danger of being beholden to the carrier and manufacturer for product decisions was not something he wanted to repeat. He was strongly in favor of Android’s vision for an open and independent platform. He threatened to resign several times during his time on the Android team over decisions that would have resulted in a closed platform.)
In hindsight the camera OS was also a brilliant idea. I primarily use my phone to take pictures, make calls and send text messages. SMS can do that well enough.
> Google turned what Andy’s team thought was a meeting of Android pitching to Google into a meeting in which Google was pitching to them instead.
There were some other companies doing interesting things - one was a startup called Savaje that had a complete java based phone environment up and running. However, they weren't a silicon valley company and didn't have Andy's connections (or reputation) from Danger.
And can you tell us a bit more about the choice of Java to create apps? Wasn't it frustrating for all the C++ coders in your team to use such a "slow" language for apps?
As for Java, it worked well enough for Hiptop at Danger on a 24MHz ARM7TDMI platform. We felt we could use a similar approach (use Java more for "business logic" and do the heavy lifting in native libraries and services) to get sufficient performance on the 200MHz+ ARM9 platforms we were looking at, and take advantage of having a real MMU for process protection and eventually supporting native code (the latter, more contentious).
Someone still hasn't built an equivalent to LuaJIT. Quickjs is in the same spirit but still interpreted, the way LuaJIT does C FFI is just pure awesome[1].
I think it's absolutely awful that you and others in this thread are actively discouraging this kind of wonderful discussion.
I think you are going against the hacker ethos and the Hacker News ethos.
It's still gauche, you're not going crazy.
People like me are untermensch to folks like you.
When you're given everything from peer bonuses to spot bonuses to first class flights to Hawaii to unlimited external validation from everyone else - it's not surprising.
Please explain specifically how I've done that if you've going to throw around accusations.
You seem to be reading a lot into the fact that I worked at Google in the past, but I didn't design the hiring process, the data access rules, the cultural vocabulary, or any of the other things you seem aggrieved by. People at giant companies don't have influence over those things, and if you go through my comment history you'll see that I have mixed feelings about working there.
If the mere mention that someone worked at a particular company gives you a feeling of inadequacy, I can certainly sympathize. But you need to realize that those feelings are coming from your own issues and not something I wrote in my comment. It's not necessary or sufficient to work at Google to have a fulfilling career, it's only necessary in this case since the codebase I reviewed is not public.
That's clearly the subtext here, and here [0]
https://news.ycombinator.com/item?id=28172705
Ya'll clearly think people like me aren't worthy of being "in the know" (and probably of living at all to be honest) for describing this particular doc. Openly flaunting your immense privilege is usually considered gauche, for the record.
If this is your read on the situation, you have problems. First of all "y'all" is multiple different people on a website who don't know each other. You can't cite a comment by one person to support your claim that another person believes something, and "that's clearly the subtext" isn't clear to me from any comment in this thread.
For the benefit of other readers, I'll just state it explicitly: this doc describes a TAR file that was shared with me when I worked at Google, that includes source code owned by the company that was never released. That's why the doc isn't public, it's not because I decided you "weren't worthy of being in the know" or of "living at all" - that interpretation is completely ridiculous. There are docs I've written at my current job that aren't public either, and that's not because I believe my colleagues constitute some kind of master race.
The key takeaways for me were:
- The UI looked like an old fashioned flip phone, as in the screenshot from that article
- Everything was JS, as discussed elsewhere
- The functionality wasn't particularly impressive to me, it had a half-finished calendar, contacts app, etc... And some graphical 2D demos
What was most interesting was just poking around the screenshots and commit log which I obviously don't have.
It'd be like asking a solider in 1915 what war they were fighting and they respond "WW1" (as if they new in advance that there would be a second world-war 30 years later).
In fact, when I was using the demo it felt like the color flip phone I used to have around the mid 2000s, and in retrospect it's hard to tell why Android became what it did today. There wasn't much unique "smartphone" functionality in that demo from today's perspective, although I didn't see the original materials about how it was positioned.
(But that was the last time I purchased an iPhone: I lost $hundreds when it got bricked and Apple said the moisture sensor was triggered so they wouldn't replace it. It never came close to getting wet, and I later got a minor settlement through a class action because the sensors were faulty, meaning anyone in the unfortunate coincidence of a broken phone still under warranty & a faulty triggered sensor got screwed.)
This was my experience as a software dev starting out at trad. hardware companies in early 2000s. Software actually drove much of the potential upside, but traditional management was slow to realize this, having perhaps made their millions making hardware. They just didn’t get software, we’re fairly software illiterate. Software became an afterthought to a hardware centric dev process, but actually was what could make or break a product.
This kind of classification only makes sense in retrospect. Blackberries and Danger Hiptops were pretty much the best phones you could get in 2005. Sure, keyboardless phones existed but they weren't necessarily better.
The market was also incredibly fragmented both on the hardware and software side. The hard divide between feature phones and smart phones hadn't yet been delineated in a general sense (the Windows phones were just kind of clunky and weird, and Blackberry basically had it's own market segment and represented "smart phones" in the press). Phones just kept getting higher and higher end, and adding more and more features.
https://www.cnet.com/tech/mobile/best-mobile-phones-of-2006/
The most anticipated phone for 2007 was Nokia's N95, the press called it a "multimedia device". It came with GPS, quad-band, 5MP camera, Wi-Fi connectivity, an accelerometer, a Web Browser, a big screen, apps. It looks vaguely like an older iPod, but has a pop out numeric keyboard.
https://en.wikipedia.org/wiki/Nokia_N95
At the time it was considered one of the best phones ever released and sold 10 million. At the time it was sometimes called smartphone, but I think today we'd call it an advanced feature phone.
Android thus came out in this market, and the phones it originally targeted looked more like the N95 than the iPhone. When Apple introduced the iPhone, they rethought the form-factor and simplified it, getting rid of most buttons and making touch not suck. But functionality-wise the iPhone was about on par with the N95, and so was Android. However, the first iPhone didn't support user apps, had limited connectivity and sold only 6.1 million devices.
When Android started supporting phones with iPhone-like form factors, it wasn't a given that it would become the predominant form factor for phones. The first few generations of Android phones still features lots of vestigial keyboards and buttons and rollers and sliders and such carried over from the N95 and blackberry style phones that Android devices were also looking like. Once the iPhone style form factor started leaving everybody else in the dust, Android phones (and the OS) simply dropped that stuff also.
It was all just a bit of a pain to DO. Mount on a PC, copy video or ebook files over to certain directories, operate through the numeric keypad, etc... the cumulative friction of all that compared to the iPhone was pretty groundbreaking even if it in a lot of ways it was just a lot of things being "a little bit" easier.
But I still held on to it for a few years because it took a while for the iPhones to really be able to do all the same things.
Apple and Google now both run app store hosting millions of apps, they both maintain their own development environments and adopted more advanced languages to make it easier. It's easy to submit apps (relative to Symbian days) and even nontechnical users commonly install and use dozens of them. The development tech and available APIs are being added and improved at breakneck speed, relatively speaking.
I remember the day this was announced and even though I never bought into the Apple ecosystem I remember feeling that it was a game changing move.
At least as of Android 8, the OS still supports hardware keyboards (and likely the rest); my current phone has one and can scroll by swiping on the keyboard.
I think Apple's stroke of genius was in the software and UI/UX.
Which, in the end, helped us a ton with Android -- post iPhone announcement everyone wanted to compete with that, nobody knew how, and the carriers eased back a bit on their absurd requirements documents full of random features nobody cared about (WAP?).
At the time iPhone was announced we were already running on the prototype of what became G1 (Dream), with multitouch. We just expected it would be the second form factor to ship, after a more "traditional" blackberry wedge sort of thing (Sooner). Post-iPhone it seemed silly to ship such a device first (you can imagine the reaction), so we skipped it and shifted focus to Dream.
WAP made sense for its time: limited computing power, limited bandwidth, super low data caps.
I agree (more or less) with the fact that someone would have come out with something similar. I strongly disagree that it would have changed the world.
Apple had the "courage" to go all-in on the form factor with the retail footprint to sell it, and of course the built-in fan base and long history of innovation to make it interesting even in its fairly primitive form.
Any other manufacturer would have either been too small to be so influential, or would have hedged their bets with a dozen alternatives, and their "iPhone" would have just been another dusty device in the corner of an AT&T store that no one knew how to use.
Maybe small manufacturers, but if Nokia or Motorola had done it first (with good software) maybe we'd be a in different world.
Even if they had something great, the US focus of tech media would have made it hard to see, and the US/EU frequency differences would have made it hard to use in the US unless Nokia made a US version, but lack of distribution makes that less likely.
> or would have hedged their bets with a dozen alternatives
Nokia or Motorola wouldn't have bet their phone business on a single UI/hardware combination like Apple did, and so I don't believe it would have had the technological impact the iPhone did.
The removal of the keyboard, in particular, was far from universally perceived as a good idea at the time.
I also think it would have been quite a while before anybody would have had the guts to not provide a slot for memory cards or to not have removable batteries.
https://en.wikipedia.org/wiki/LG_Prada
> It was first announced on 12 December 2006.
> It is the first mobile phone with a capacitive touchscreen.
> The LG Prada was announced shortly <<before>> Apple CEO Steve Jobs announced the iPhone on January 9, 2007.
> After the release of the iPhone the head of the LG Mobile Handset R&D Center was quoted saying he believed Apple had stolen the idea from the KE850 after it was announced as part of the iF Design Award.
iPhone just did what Apple's good at, taking existing technology and packaging it nicely, but it would have happened with or without them.
>Chris DeSalvo’s reaction to the iPhone was immediate and visceral. “As a consumer I was blown away. I wanted one immediately. But as a Google engineer, I thought ‘We’re going to have to start over.’”
“What we had suddenly looked just so . . . nineties,” DeSalvo said. “It’s just one of those things that are obvious when you see it.”
https://www.theatlantic.com/technology/archive/2013/12/the-d...
Android was nearly ready to launch as a Blackberry clone when Jobs demoed the iPhone.
What is that method you are referring to?
I remember the BlackBerry Storm, the first touch screen phone by RIM launching at the same time as the iPhone 3G. No Wifi, slow janky scrolling, no apps (well, sure, if you don't mind downloading some random .jar that have to be recompiled for the custom fork they ran on that device). Felt like a rushed beta.
Meanwhile the iPhone just worked. Smooth scrolling, fast browser especially on Wifi. You could get apps from the AppStore, no friction.
https://www.androidauthority.com/lg-prada-1080646/
>The LG Prada was the first mobile phone with a capacitive touchscreen, but there are reasons why you don't remember it.
In my first post on this thread, I said Apple improved the packaging of smartphones (ie. form, UI, ecosystem).
I was at one of those TI partner companies during this time and we were contracted to assist with middleware for Android Sooner. I even got to test with one. Then everything got put on hold.
The carriers had complete control over everything up until Apple came along. Even Android enjoys the fruits of Apple's initial demands.
The main obstacle is SoC drivers that are kept closed.
How exactly was it limited? My only issue with older Windows Mobile devices was non-capacitive touch, and HTC HD2 (still on Windows Mobile 6.5) did not even have that problem.
Good luck running a desktop OS optimized for keyboard, mouse and large screens on a mobile phone with a small screen and only touch as input.
Oh. WinMobile was like that. And the original Android was like that. People went over to iPhone in droves, and Google had to scramble to change Android to mimic iOS.
Now, it is a tightly locked os with some limited bit open. Just enough to pretend to be open source but not enough to let anyone build a real functional phone from it or being able to rebuild the os of one own phone.
Chet Haase and Romain Guy both were treasures in Java community and now in Android community.
Thanks Romain Guy and Chet Haase for everything you guys do.
How so? Anywhere I can read about their contributions during Java time? Thanks.
That is to say, it's hard to even wrap my head around how Google has been able to leverage Android to influence the development of mobile as a platform, or the information they've been able to collect from 3 billion Android devices. So it's hard for me to guess how much money Android helped Google to earn.
But it takes ten seconds of back of the envelope math to determine that Android makes money for Google. Google pays $8-12 billion per year to be the default iOS search engine in 2019[2]. I think it's safe to say that Android is worth at least that much to Google.
The risk remains that Apple creates its own search engine and cuts Google off from even being allowed to be the default, which was the risk that Google originally tried to fight with Android vs. Microsoft, but that hasn't materialized yet.