Why the Google Android Phone Isn't Taking Off

Hugh Pickens writes "Farhad Manjoo writes in Slate that while the iPhone commands nearly 14 percent of smartphone sales and BlackBerry about 21 percent, Android has only 3 percent. And even though Android is far friendlier to developers, it has failed to attract anywhere near the number of apps now clogging the iPhone. Manjoo writes that Google went wrong by giving handset manufacturers and carriers too much control over the design and marketing of Android phones so there is no idealized 'Google phone' — instead, Android devices get names like the T-Mobile G1 or the myTouch 3G, and each is marketed separately and comes with its own distinct capabilities and shortcomings. 'Outside handset manufacturers lack ambition — -none of them even seems to be trying to match the capabilities of the iPhone, let alone to knock us down with features that far surpass those of Apple's device,' writes Manjoo. 'A smart handset manufacturer could build a top-of-the-line Android device that outshines Apple's phone in at least a few areas — better battery life, a much better Web browser, a brighter or bigger screen, faster or more functional controls... something that might help Android inspire gadget lust. But so far, that's not happening.' John Gruber echoes this advice and adds this advice to Android manufacturers: 'If Apple is BMW, you can be Porsche.'"

Comments

"It's the Network"

[Stupendoussteve]

The G1 and myTouch are nice, unfortunately they're on T-Mobile, which is nice but not nice everywhere. If T-Mobile worked in my area I would certainly try them out, at least.

Re:"It's the Network"

[Darkness404]

But do you think Verizon would even -allow- a phone with Android to run? I mean, I've compared the same dumb phone (I think it was the Razr) across AT&T Sprint and Verizon, the AT&T and Sprint phones were pretty good but the Verizon phone was pretty much neutered to the point where they can't do anything beyond changing the background, changing it from ring to vibrate and using the camera. Verizon is -terrible- when it comes to phones, they might have the "network" but when all the phones are total crap, the network is useless. I think it even went as far as Verizon rejecting any phone with wi-fi.

Re:"It's the Network"

[Bodero]

Verizon Android Phones Are Officially Coming.

There exists a pretty strong misunderstanding that Verizon "locks down" their phones. They did, yes. But in the past year, they've stopped disabling GPS on their phones (including the Omnia, Storm and Tour), said that all future Blackberries will have Wifi, and launched their Open Development Initiative to get data devices (among other things) on their network.

Oh, and their next generation network (which is launching 2+ years before AT&T's) is LTE, based off the GSM standard.

But I don't blame you, they've definitely had restrictive tendencies in the past.

Re:Citation needed

[maxume]

Well, you don't have to pay Apple money to develop for Android, and you don't have to get Apple's permission to distribute your app to users. Those are probably pluses even if you experiencing full-on reality distortion.

Apple may still be providing a more attractive program though, simply by bothering to market their phones.

Re:Citation needed

[jelizondo]

As I mentioned in my previous comment, the "friendlier" to developers is really that it is open source, that you can use open source tools (Eclipse, Ant, etc.) for development, but that's all folks! as the good bunny said.

Documentation sucks, most of developers are outside the U.S. (from my experience mainly India and Pakistan) where they can't get a developer's phone, the emulator is fine except that it can't emulate making calls or receiving them, etc. etc.

It is "friendlier" in a sense but like the Apple of old days (i.e. early 80's), when you could get real support, real docs and real machines to develop apps.

Someone needs to get behind developers' support right away!

Re:re Lack of apps, developers?

[cduffy]

I'm curious -- have you done any development for Android, or are you armchair'ing this one?

They've been _extremely_ careful about what is declared a public API, to the point of holding back features on account thereof. That's one of the major reasons RFCOMM support (to pick something dear to my heart) has been unavailable for developers in every version released so far -- they're unwilling to declare the API stable until they have something they know they'll be able to maintain through newer versions of BlueZ and security audits/updates, and their compiler flags any attempts at using anything which isn't a public API (they've also released updates which break attempts to get around these measures and build software using version-specific, unreleased APIs).

Personally, I expect Android to take off on a larger scale when the fleet of phones expected to release late this year (from numerous manufacturers) make it out their respective doors.

Re:Show some evidence

[bobetov]

What part of "don't have to pay Apple money to develop for Android" and "don't have to get Apple's permission to distribute" did you not understand?

Android is a platform that give much more, and more meaningful, freedoms to app developers.

I'll add another big one - on the Android platform, replacing core apps with your own version is *encouraged*, and in fact *designed into the platform*. Unlike Apple's recent filing about "altering the core experience" re: Google Voice. Apple could create an iPhone-themed app suite for the G1 tomorrow, host it on their own servers, and no one could say otherwise. That's a pretty fundamental difference.

Say what you will about the iPhone as a sexy beast, etc, but as a developer platform and ecosystem, the only thing Android is missing is higher handset sales.

Too Slow!

[codepunk]

Developer Friendly? Not, I spent some time a few months ago trying to port one of my games to the G1. The game requires some fairly heavy physics, it runs
blistering fast on the the iphone. The G1 however just is not up to the task, face it the IPhone is just a much better performing device. When it comes
to squeezing performance out of these tiny devices get java out of my way, I need to be able to program against the metal.

Re:Show some evidence

[sammyF70]

Factors that are friendly to developers :

• the developer fee is $25 ONCE
• there is no review of your application and when you publish it, it's on the Android Market right away
• Applications are written in Java, which means you can develop on any platform you like in a language that's not OSX specific
• the SDK integrates nicely in Eclipse, which means you can have a ~nice~ (debatable I know) IDE for free
• the documentation is (contrary to what other said) quite good and the #android-dev IRC channel on freenode provides great help

Sadly, Android suffers from

• an incredibly bad Market, with one of the worst search engine ever written (a regular joke on #android is whether Google should ask Microsoft or Yahoo for a better search engine), and virtually no useful feedback mechanism for developers. This results in developers pushing fake updates just to make users aware that their applications exist, and comments like "sucks. crashes" littering the market (with no way to know which version actually crashed, nor to ask the user who posted the comment why/how/where the application crashed
• 24 hours refund even for 99c applications who have demo versions available and virtually no copyright protection. When you first get a ORDER/REFUND" cycle of less than 4mn on one of your applications, you get angry. The angriness just transforms into fatalistic depression after the tenth. Similarly, with games, when refunds is asked after 23 hours, you can't help but ask yourself what exactly, in terms of gameplay, users expected from a 99c game. (this is especially infuriating when the game has high replay value, but can be finished the first time around in a few hours)
• Non-homogeneous hardware. You're supposed to write your applications so that they can run on vritually any display resolution, with or without trackball, hardware keyboard, and whatnot. Add to that the fact that the T-Mobile G1 is quite underpowered for graphically intensive applications and that that's probably what you're using to develop and the range of stuff you can do shrinks greatly.
• Android applications are writen in Java, with the shortcomings this brings with it (anybody want to write a game and see how the GC kills the framerate by processing stuff that has nothing to do with your application? it's *really* fun)

• So .. all in all, yes, it's far friendlier to developers, but it's also a highly frustrating platform to develop for.

Re:Show some evidence

[Mr2001]

It's really, really easy to get started with iPhone development. You pay Apple a small fee and get access to piles and piles of sample code, great documentation, a mature API, and you even get 2 support incidents in which you get to interact with a real Apple developer for your money.

With Android, you get access to piles and piles of code without having to pay anyone. Want to know how the default music player works? Just look at the source code. Good luck finding the source to the iPhone's music player.

If you want to interact with a real Android developer, just post on the newsgroups or find them on IRC. Again, it's free.

I haven't done iPhone development, so I can't comment on the relative quality of Android's documentation or API, but they seem fine to me.

Re:they could still do it if they wanted

[Dun Malg]

The difference is the iPhone had -everything- usable

Except copy/paste, 3G, video recording, and several other features that came standard on other phones of the same period. But yeah, every one of the features they deigned to include worked quite well.