http://twitter.com/#!/segphault/status/90546480737296384

What Linux developed by a company today isn't "controlled by a single vendor". What does that even mean? CyanogenMod and countless other AOSP-based roms seem to contradict that implication anyway.

Not sure what the criticism is regarding the hardware vendors. That'd be like criticizing RHEL if a distributor sold machines and locked it down to only allow RHEL to boot. Does that reflect on the project itself? No.

I suppose you can criticize them for not using GPL3 (as it would prevent the hardware lockdown) but you'd be criticizing a loooot of people in that case.

(A defensive reply to a "witty" bit of snark earns downvotes and silence. Shocking.)

>derStandard.at: How does a company like Google balance the need between wanting secrecy on new developments and doing it open source?

Chris DiBona: That varies between project. If you look at Chrome, every commit is instantly online. Android has a punctuated release model based on its marketing needs.

derStandard.at: Why is the chosen model for Android and Chrome so fundamentally different?

Chris DiBona: It's a different kind of market. If you look at Android we have lots of partners. We have chipset partners, we have handset partners, we have carrier partners. They all want to use Android and they all want to have something special about themselves. So they want to use Android for that specialness. What that means is that one handset vendor probably doesn't want to interact too much with the other handset vendors because they are competitors. And Android gets caught in the middle of all of this. And the bigger question then becomes how you architect software that it's still useful around that kind of model.

He doesn't really say why Android development needs to be secret. So much for openess.

I thought this was a pretty direct response:

What that means is that one handset vendor probably doesn't want to interact too much with the other handset vendors because they are competitors. And Android gets caught in the middle of all of this.

ie - the handset vendors want to keep things secret from each other (and they could gain a lot of information if they saw the commits as they happened on Android). No surprise there, right?

>and they could gain a lot of information if they saw the commits as they happened on Android

These two statements contradict each other. All handset makers are using the same Android code with some changes. The Android 2.3 in Sony Ericsson is same as the one in Samsung or HTC. Or is Google developing separate Android OSes with each hardware vendor? My question is what's stopping Google from releasing the base Android OS? And what's stopping them from following a open development model like chromium?

Sure MS/Apple/HP/RIM can gain some insight if they saw the commits but that's another matter.

Ignoring Honeycomb, they do that - they just delay it. I thought the delay is what you were talking about.

Or is Google developing separate Android OSes with each hardware vendor?

It's not an either/or question. Each vendor has different drivers (for things like different kinds of screens, 3D cameras etc), and if other vendors saw these it would give away their capabilities early.

I believe that each vendor doesn't get to see the whole tree until there is a complete release.

And what's stopping them from following a open development model like chromium?

Chromium doesn't have tight hardware integration.

A better example is something like the graphics drivers in the Linux kernel. Nvidia & ATI have historically been very reluctant to release the code for them and have tried to avoid it in various ways.

Handset vendors have the same motivations as the video card vendors, and so Google has to accommodate them somewhat to get them to use Android.

It's kind of exciting and also kind of scary. I'm not a fan of the Android development stack at all, and this comes from someone who's been championing Android over iOS for the past several years. I think overwhelmingly it's probably a good thing, I just won't be dabbling too much.

"A platform controlled by a single vendor and locked down by hardware vendors != the Linux desktop "dream come true"

"For smartphones, sure. For desktops this is not going to happen for a long time if ever."

As much as Android is now both the dominant client Java and the dominant Linux userland, there are two things that have to happen before it is the Linux "dream come true."

First, it should become really open, without a Google controlled two-tier structure for OEMs. Not all open source, but open to all comers: Anyone who makes an Android device that passes compatibly tests should be able to ship the Android Market on it. In this, Microsoft is more even-handed with OEMs. Any OEM can license Windows, and they get the same thing HP and Dell get.

Second, Google needs to make this openness and evenhandedness happen for tablet makers, which are more like desktop PC makers than handset makers. Tablets are not tied to mobile networks, nor to mobile network operators' retail channels. There is no mobile operator dictating what the software content of a tablet must be in order for it to be accepted by their channel, so there is no excuse for the process to be less than transparent and open to every product that passes compatibility tests.

The difference between Microsoft and Google is that if you don't license Windows, Microsoft gives you nothing. Big, fat 0. If you don't sign a deal with Google, you get 95% of Android.

Your demands perfectly illustrate human nature: no matter how good you have, not matter how much you have, there's always more to want and demand.

With Android Google is being generous to extreme, unprecedented levels. They give away, for free, with source code, technology that would require millions to redevelop from scratch, but that's not good enough for you. And the same time, you make no demands on Apple or Microsoft to be even 1% as generous as Google is.

The restrictions that Google has on Android have very good rationale: to prevent explosion of incompatibilities, something that has plagued uncoordinated desktop Linux efforts for years and has no signs of being resolved. rpm vs. apt vs. yam. Gnome vs. Qt. etc.

And finally, no matter how many of your demands Google will meet, people will always come up with more. You have already demonstrated that you'll find a fault with a very big and generous gift of Android. I can easily imagine that after Google lifts the restrictions that bother you so much (even though the likelihood that you're personally affected by them is 0) someone will consider that 30% royalty in Android Market is too high.

Do you have any evidence to support this statement? Agreements of that nature are not normally made public, and the circumstances surrounding the google maps/GPS on android thing, means this is very likely not the case.

>The restrictions that Google has on Android have very good rationale.

Im so happy that google is choosing to restrict my freedom to use my devices for my own good, thank you google.

>You have already demonstrated that you'll find a fault with a very big and generous gift of Android. I can easily imagine that after Google lifts the restrictions that bother you so much (even though the likelihood that you're personally affected by them is 0)

Android is not a gift, its intent was never to charitably donate something to the community at large, its intent was to make google money. Their entire platform is based off of the open-source work of others, it did not spring forth from google fully formed, expecting reciprocation is not too much to ask.

Google is not 'restricting your freedom'. You can fork android and do whatever you want.

> Android is not a gift, its intent was never to charitably donate something to the community at large, its intent was to make google money

Intent is irrelevant. Pre Google: 0 commercially viable Linux based phones. Post Google: dozens.

> Their entire platform is based off of the open-source work of others, it did not spring forth from google fully formed, expecting reciprocation is not too much to ask.

They do reciprocate. The source for android is yours to hack.

that is a bit of a hobson's choice, sure (some of) the code is available (pre 3.0), but without a device that is open and able to run it, it may as well be closed.

>Intent is irrelevant.

Intent is everything, would you be right to suspect the motivations of a mining company conducting a review of environmental impacts, even if the results appear sound on the surface?

>They do reciprocate. The source for android is yours to hack.

Except for of course, gingerbread.

I'm not sure what you define as "a device that is open", but surely the many unlocked handsets apply?

If not, then http://beagleboard.org/project/android/ perhaps?

If not, then maybe the Nokia 810: http://www.linuxfordevices.com/c/a/Linux-For-Devices-Article...

Am I missing your point somehow?

google doesn't make devices

> would you be right to suspect the motivations of a mining company conducting a review of environmental impacts

how is this a good analogy? they write software. its open source. you don't have to trust them.

> Except for of course, gingerbread.

this statement is actually false.

I also know from experience in the mobile application business that Android Market is the easiest platform vendor or carrier-controlled app market to deal with.

My comments relate mainly to how the tablet business differs from the handset business. The same practices that might be necessary for Android handsets don't apply to tablets, and evidently don't work as well, either.

I think tablets will be a very big part of computing, and Android tablets should be a big part of the tablet business. But, so far, Google hasn't allowed many low-cost OEMs to ship Android-based products with Android Market and Google's other proprietary applications. Had Google gone with, say, Archos on a $350 tablet instead of Motorola for a high-end product, the tablet roll-out for Android might be very different now. More generally, Google should not have tried to pick a winner, or a winning approach to tablet product formulation.

That openness to all comers makes Android Market vibrant and diverse. Maybe Google should try that approach to tablet OEMs, too.

On the other hand, Android will be the closest thing to "linux everywhere (but the desktop)" . It could power TV sets, washing machines, car navigators, watches etc. Looking forward to using it on my next Roomba.

Minecraft is written in Java.

> When will ... even more specialized medical/industrial/scientific software be available for android?

Presumably when the people buying the software begin to express an interest in moving to android. I'm not saying that it will happen, just that the software will be ported over, or new software written, when people start wanting it.

I don't think that anyone seriously believes that android is going to replace OS X and Windows on desktops, just that, as time goes on, people will be spending more and more time on tablets and smartphones and less on traditional desktops and laptops. I find it a bit sad, mostly because I refuse to believe that I'll ever be able to type as quickly on a tablet as I do on a keyboard.

Now, what you are talking about - computationally-intensive tasks that traditionally require workstations - Android does not perform nearly as well. But the people who really need that kind of processing power and can't satisfy it with a tablet, game console, or other home-oriented system are only a small segment of the overall market.