Google is between a rock and a hard place. They release Honeycomb source, then a bunch of makers will rush out with phones (this move is clearly about phones and not tablets) running an OS they don't want on phones yet. And if they do, the headline would be "Steve Jobs vindicated: Google Android is fragmented as hell". Of course, ignoring that Apple did the same exact forking and then merging move with iOS.
Personally I would err on the side of openness and let the shitty manufacturers run wild, but at least this way the blowback is just a bunch of philosophical arguments on nerd sites and not actual customers buying bad Android products.
One could argue (and I do) that it is the life they chose, they want their cake and to eat it too. But absolute control over the user experience is pretty much incompatible with 'open'[1]. As someone who was privileged to here some of the debate about this pre-1.0 launch I can tell you that the things that have come to pass are not 'surprises.'
An alternative way to deal with this is to execute well and have you're stuff be the 'best' by virtue of the fact that you have a bunch of great focussed folks working on it and delivering new updates, but that is a hard strategy to execute successfully. There are a lot of folks 'outside' vs 'inside' when you are truly open.
In one such internal discussion, when the question of lots of people outside of Google trying lots of different UI's and form factors, it was noted (with a well known cliche) 'that is a good problem to have.' I guess not so much :-)
[1] Submitted as exemplar the KDE vs Gnome "discussion."
Isn't honeycomb just a codename for the latest version of android though? Isn't that like saying: Ubuntu LATEST will not be open sourced until...Ubuntu OLD is still open ? In my mind, that makes it a whole lot more closed than open.
Unlike say, Ubuntu or Google Chrome, Android has always been developed behind closed doors and then opened, usually a few weeks after devices ship with it preloaded.
That is a difference on the spectrum of openess, and on that measure Chrome or Ubuntu has always beaten Android. The only difference for Honeycomb is the time between device and code reveal.
It's a bit of a philosophical question as to whether changing that 2 week delay to a 2 month delay moves you from "open" to "closed".
I think it's salient to point out, too, that that shift on the spectrum is tiny compared to the spectrum-distance between Google and any other player in the industry.
When I hear "open source", I think "developed in the open". While projects that occasionally dump source can be technically called "open source", they don't look at all like the open source projects I like to work or or with.
Honeycomb originally started as a fork for tablets only. It looked like the plan was to have the 2.x branch for phones and the 3.x branch for tablets. I think recently something happened and they are now going to standardize on the 3.x branch for everything. This leaves google with its pants down as 3.x is barely ready for tablets, let alone for phones. So now they have quite a bit of work to do and don't want to give away the goods to smaller manufacturers (who will simply make shit phones and hurt the brand) just yet. I think these smaller partners can survive waiting a couple months for 3.0 while google puts in all this work to make it acceptable. Heck, its free expert level coding for them. I'd hate to see what Sylvania or A-Open or whomever would do with a half cooked 3.0 source.
Regardless, you may not say Ubuntu is closed source, to follow your analogy, which is what the author writes about Android. You may say that Android is not completely open source (which you could before Honeycomb was conceived) or that the latest version of Android is closed-source for now. There's no question the article is inaccurate and it's clear to me the author is trying to deliberately create an exaggerated conflict to get more page views. All in all a pretty poor article I think.
I would not call that "open". Doing "open source" does not necessarily imply "being open".
My main problem with Android is that you do not see any of the development that lead to Honeycomb. Even if you are interested, there is no way to see where the project is going without being on the inside. Android has a completely closed development cycle until _after_ the release of the first devices. I cannot consider that 'open' in a general sense.
I think the point that people want to make is that all competitors of Android are even more closed down.
Yeah, it's certainly not 'open source' in the traditional sense of an open community.
However, it's "the best we've got", and all things considered, it's pretty good: Google could have kept many things much more tightly controlled and still been somewhat 'open' (see Windows vs Mac).
Steve Jobs harping about it though is sort of nonsensical: great, pick your nit, but to really make me stand up and care, go them one better, which is something he will never, ever do.
> tightly controlled and still been somewhat 'open' (see Windows vs Mac).
Huge chunks of OS X are open source. It's not necessarily equivalent to throw it into the same box as Windows (there are no "open" forks of a Windows based OS, but there are of Darwin, and FreeBSD has benefited from Apple as well.)
> Isn't Darwin open source in the same way Android is?
Not really. Android is a complete system, with a kernel, libraries and the whole GUI on top of that. Darwin is just the kernel, and I think they've sort of phased out doing much with the 'open source' aspect of it.
"honeycomb will not be open sourced until they decide it's ready" That's what I was thinking until the article speculates how smaller tablet providers will lose out.
If it was my platform, I know, I'd like to be first to market with well built tablets rather than a bunch of mediocre tablet attempts, which could reflect badly on the platform itself.
"If it was my platform, I know, I'd like to be first to market with well built tablets rather than a bunch of mediocre tablet attempts, which could reflect badly on the platform itself."
And since most of those tablets used pre-Honeycomb Android, it's actually counterproductive to hold back Honeycomb if the goal is to raise the quality of the worst tablets.
Well, it is the register, which I am surprised would be popular on a site like this. Regardless, google is pushing back the release date. Its their prerogative to do so. Not sure why this matters to 99.9999% of the population who will never look at the source or install a custom ROM, etc.
Not to mention there are several versions of android. There's the public version which is a little behind. There's the cutting edge version that members of the OHA get and if they follow their contract they get google-branded apps (gmail, youtube, maps, market, etc).
Dunno, seems much ado about nothing. They didn't give market access or gmail branded apps to the previous garbage tablets released so far. This is just google trying to stop this wave of horrible half-assed tablets from destroying the android tablet market. It may be too late, but leaving dozens of manufacturers selling non-capacitive touch screen based tablets with seriously underpowered specs on buggy software that was never meant to run on a non-phone is hurting the brand.
Expecting the "market" to work itself out is having a little too much faith in markets and consumers imho. Brand damage is real and is helping sink Android into the tech ghetto.
To me, 'Open' is not the same as just 'Open Source'. I'm not really that interested in whether Android is open source or not.
What I am interested in is whether the device lets me do what I want - that is, it allows me to download and install whatever software I choose to, in what manner I want. Android lets me do this. That is why Android is 'Open'.
Whilst I understand for most consumers this is not an issue, and I have no qualms with people who go along with that, for me personally as a programmer, I will not buy any other smartphone platform if it does not afford me that same liberty.
How is Android open under your definition? My phone has a locked-down bootloader that will only load an OS that has been signed by Motorola, I can't get root, I can't flash it, and I can't upgrade to the latest version of Android. Right now I'm feeling pretty closed.
Google's position on this has always been that they're open to the carriers/manufacturers doing whatever they want with the software, and when they do something bad, well, people do bad things sometimes.
I find it amusing that when Google is open to carriers/manufacturers doing whatever they want, people (usually non-android users, though you're an exception here) complain about how they're not open, and when Google tries to take more control of the development process to protect the user experience, people (again, usually same non-android users) complain about the same thing.
That said, I think Google could definitely do things better. I, for one, wish Google would exercise their "Google Experience" branding to enforce unlocked boot loaders and root access. On the issue of open sourcing Honeycomb, I could really care less. On the spectrum of closed to open, they're way farther on the open side than anyone else in the industry so I don't find it hypocritical at all for them to claim to be open.
(NB: I also own a Motorola device with a locked boot loader. I was very surprised and delighted by how much the devs were still able to do, but I definitely wish I could get some custom kernels installed.)
To a large extent, the two go hand-in-hand. A closed-source platform may give you a lot of freedom to customize things, but what it doesn't give you, you can't have. On an open platform, you can do anything.
It's the difference between "most of the doors are unlocked here" vs "if you want a new door, here's a blueprint and a chainsaw."
I don't understand these "Google Android is not open" arguments. Can you go to http://source.apple.com/ and download iOS 1.0? No. Can you go to http://source.blackberry.com/ and download their OS? No. Sure, Honeycomb isn't coming out right this very moment, but at least there's something there.
By any definition Google is at least several orders of magnitude more "open" than Apple. Apple releases no source (other than what is required of them, and only after a few weeks or months). Google releases all source (sure, after a few weeks or months).
That said, I'm sure you could argue over who gets to claim WebKit as their own these days since I've seen a few people refer to it as "Google's WebKit" since they surpassed Apple in checkins at some point (or had for a month at least, it's not something I care to follow that closely).
The fact that Google does more work on it would tend to weaken the point. Also, the fact that Google is using Apple's software tends to undermine the Team A vs Team B regime that's the premise of most of these discussions.
That would be my point. Practically all post-2007 phones and tablets depend on Apple's open-source work. It's easily the most important application on mobile devices. Had Apple decided to roll their own proprietary browser instead of extending an open-source project, they would have a far bigger lead over the competition. And they did it without a lot of crowing about how moral it made them.
That doesn't make his statement any less true. The link you provided only covers open source software that Apple included in their OS, and doesn't come close to including all iOS source code.
Even taking the (flawed) premise at face value, there is still a mile of difference between Apple's walled garden for apps and other content, versus the Android ecosystem.
You're playing on two different meanings of the term "open". Because Android might not qualify as OSS, they create the implication that it's just as tightly controlled as Apple's walled garden. But that's not even close to true.
The point of the OP's argument seems to be something like "Apple isn't open, and now this proves that Android isn't open either; hence, Apple and Android are the same in this respect". This argument is fallacious.
"On Thursday, the company said that as its select partners release the first tablets based on Android "Honeycomb" – the latest version of its mobile operating system – it will not open source the Honeycomb code."
And "But there was no indication that the code wouldn't be promptly open sourced as the first devices were released. What's more, Google did not make a public announcement that it will keep the source closed. "
I'd be surprised if Google has manged to rid every last piece of GPL'd software from the OS. Otherwise, they're going to have to release the source.
> I'd be surprised if Google has manged to rid every last piece of GPL'd software from the OS. Otherwise, they're going to have to release the source.
They have to release source to the GPL parts (WebKit, the kernel and stuff that links directly to it, like drivers and possibly other stuff they use). From their glibc-like Bionic and up (including the userland, Dalvik), everything is BSD-style and they have no obligation whatsoever to release anything. Not even to manufacturers.
But the reality is surely that Google and its partners don't want smaller name manufacturers eating into their tablet sales. Or perhaps they don't want larger names nabbing pieces of code for their own tablet OSes.
Why would they be worried about this? Isn't the purpose of Android* to increase global pageviews and sell more (local) ads? Don't 2nd tier OEMs and competing platforms further these goals?
While I am not sure if you can take google at face value in terms of their reasons, the author's conclusions seem wildly unsourced. Unless someone would like to expand on his reasoning?
One example: RIM wants to ship their Playbook with support for Android apps. They probably need the Android source code in order to do this. They've annouced that Playbook will only support Gingerbread Android apps. With the Honeycomb source code they might be able to support apps that work better on tablets.
It seems to me however as an Android watcher, that when Google developed Honeycomb in parallel to Gingerbread, they created an internal fork of Android to specifically support tablets.
What needs to be done and it seems what they intend to do is heal the fork, so they have a single codebase again. The real issue is whether they manage to do that in time for Ice Cream.
I agree that the author's points are purely speculative, but the situation kind of demands speculation, since the official answer was very unsatisfying.
The tablet market somehow seems more competitive than mobile is currently, and third parties like Amazon could rebrand/white label Honeycomb for complete rival tablet platforms, forgoing the 'with Google' label, services, ad platform, etc. Tablets are general computing platforms and not locked down like the carrier dominated mobile market; Google would have a tougher time controlling this kind of thing than they do with mobile phones.
Seems like that's what was done with the Nook, no? While I can see your point, I guess I'm just not convinced that google sees projects like that as hurting them. Surely a traditional vendor would seek lock in and revile such a use, but isn't the whole point of android open source that google's platform is their web applications/ad platform and that's where they seek to stymie competition?
I think the key phrase is "and its partners". Google can't increase global pageviews without partners and they can't step on their partners' toes too much.
Aren't these the same partners that compete in the 2.x smartphone space? In a world where damn decent android phones like the LG optimus are pushing the edge of the $100 barrier while "high end" phones continue to come out and stay in the $500 range it doesn't seem to me like they're too upset. Perhaps a notable partner or two did a lot more work on honeycomb than anyone did on previous versions?
I think I'd lend more credence to the theory that google wants to make sure that the early releases are all of particularly high quality so that's the reputation google tablets get. That would seem to encourage a short term embargo. They got beat up a lot on the reputation of the 2nd/3rd tier chinese oem 2.x tablets. Then again, the worst android phone I've personally used was a samsung.
"While we’re excited to offer these new features to Android tablets, we have more work to do before we can deliver them to other device types including phones. Until then, we’ve decided not to release Honeycomb to open source."
The until then part strikes me as pretty important. It sounds like Google is just delaying the release of the source until they reach a certain milestone. This is a very reasonable position to take. I think the author of the piece is jumping to conclusions.
I was planning on purchasing an Android tablet, but no sale until the source code is out there. I own an Android phone and have been happy with it, but I'm not interested in giving up my freedom just to have my hands on the latest toy. I am extremely disappointed in this decision. I wasn't under any particular illusions about Google, but I actually believed they were comitted to Android as an open source project. If they get their act together and get their source code out the door soon and don't repeat this nonsense for subsequent Android releases, they will keep me as a customer of devices based on their operating system. Fool me once, shame on you, fool me twice, shame on me.
It's a bit premature to say that Steve Jobs is vindicated in so far as Android is concerned from a journalistic standpoint and claims of "vindication" seem to be a bit more editorial than factual. However, timing is everything when it comes to linkbait.
Open sourced means "source released with binaries".
It does not mean "All code available with all checkins at all times"
Because community projects pretty much have to function that way does not mean that's the only way to do open source.
--An iOS developer and ex-embedded Linux/Linux kernel driver developer who's a little POed that everyone thinks only their way of doing open source is "True to Freedom" or whatever
Personally I would err on the side of openness and let the shitty manufacturers run wild, but at least this way the blowback is just a bunch of philosophical arguments on nerd sites and not actual customers buying bad Android products.