Android's Success a Threat To Free Software? 416
Glyn Moody writes "Two years after its launch, Google's Linux-based Android platform is finally making its presence felt in the world of smartphones. Around 20,000 apps have been written for it. Although well behind the iPhone's tally, that's significantly more than just a few months ago. But there's a problem: few of these Android apps are free software. Instead, we seem to be witnessing the birth of a new hybrid stack — open source underneath, and proprietary on top. If, as many believe, mobile phones will become the main computing platform for most of the world, that could be a big problem for the health of the free software ecosystem. So what, if anything, should the community be doing about it?"
How is this different from the status quo? (Score:2, Informative)
There is a disconnect between open source proponents and the way open source software is actually used.
The reason that Linux/Apache/MySQL/PHP is so successful is that it eliminates a cost and provides a standardized platform that is easy to maintain and replicate.
There are billions of dollars' worth of proprietary software running on top of that stack.
Part of the reason for the complete and utter failure of Affero GPL is that it gives an implicit right of audit and can result in what the UK calls an "Anton Pillar" order which could literally result in a team of bailiffs seizing and searching your servers in case you modified some AGPL software.
The point is, that while there is public consensus on the use of open source for infrastructure, there is no similar enthusiasm for viral obligations nor is there any interest in opening up the value-add/secret sauce on the top of the stack.
Maemo (Score:5, Informative)
Vote with your wallets. Maemo [wikipedia.org], the most open internet-tablet/smartphone platform currently on the market (assuming OpenMoko is dead). Not perfectly open, but a lot better than the Android.
From the 770 [wikipedia.org] in 2005, to the N800 [wikipedia.org] and N810 [wikipedia.org] in 2007 to the latest release of the N900 [wikipedia.org] this year.
There's even third-party clone [armdevices.net] which the platform needs to become truely mainstream.
Re:Maemo (Score:2, Informative)
You realize the GOLD PLATED version of the N95 or whatever is cheaper than the N900 right?
Re:Not New: Apple's stack is hybrid too (Score:5, Informative)
Apple does not just exploit open source, they also contribute bleeding-edge, high-quality code for GCC (LLVM), although they would legally not be required to do so by the BSD license.
OT: Palm has by far the most apps (Score:1, Informative)
The old PalmOS has by far the most "apps" and they don't have to be approved by anyone:
http://www.freewarepalm.com/ [freewarepalm.com]
http://www.handango.com/ [handango.com]
http://www.pocketgear.com/ [pocketgear.com]
http://www.mobihand.com/ [mobihand.com]
http://www.pdastreet.com/ [pdastreet.com]
and also: http://sf.net/ [sf.net]
I never understand why everyone is so amazed by the iPhone's "Apps". Handheld apps have been around for over 10 years.
Re:Okay, I'll be the one to say it... (Score:5, Informative)
Or support the N900 instead of the Android. It's not a totally open stack, but it's much more so than Android, and the apps also tend to be direct ports of Linux OSS. And the whole thing is less locked down to begin with.
Re:The obvious answer (Score:2, Informative)
You do realize that it's perfectly ok for free (as in freedom) software to cost money [gnu.org], yes?
Actually, summary is wrong. (Score:3, Informative)
Wrong.
I went and clicked the links (I know, I am new here) and if you look at the actual data in AndroidLib (http://www.androlib.com/appstatsfreepaid.aspx), you will see that 60% of the apps are free apps.
Re:Actually, summary is wrong. (Score:3, Informative)
Re:The obvious answer (Score:3, Informative)
It is perfectly OK for free software to cost money, but since the requirement is to allow redistribution of the source, anything even remotely popular will become a race to the bottom.
If I pay $X for a chunk of software that includes source, I can immediately resell the same software for $X/2. Even if I only find one person who wants to buy it, that is a good deal for me.
That is why open source apps usually try to charge for support, customization or maintenance, rather than just licensing the core code.
So sure, go ahead and sell free software, just don't expect to make any money doing so.
Re:Well, let's see (Score:2, Informative)
The point of free software is freedom [gnu.org].
The fact that free software is generally of higher quality is a bonus, one that the "open source" movement focused on. The guy who created the "Open Source Definition" has said it's important to focus on freedom [debian.org], but unfortunately many still think that talking about people's freedom to use, share, and modify software is just too radical.
Re:Okay, I'll be the one to say it... (Score:3, Informative)
Re:What an Oddly Backwards Opinion Piece (Score:3, Informative)
If the software comes with the hardware, why does it need to be proprietary? The answer is usually so that the company can drop support for that model in future and encourage you to buy their latest one. If you made it a requirement that the software be Free for all hardware that you buy, then your lab would have the choice, when the hardware goes out of support, to pay someone else (or get a PhD student) to maintain the software, or to buy new hardware if that's a better choice. If a company is unwilling to give you the code to the software on their (expensive) hardware, then you should be suspicious of their motives.
Free Software does not mean 'developed by the community' it means that it comes with the grant of a set of rights outlined by the FSF's four freedoms. There's no reason why Free Software can't be developed by a company and there's no reason why that company can't make money from it. If your business is selling hardware, then making the controllers Free Software may even save you a bit of money by getting outside patches to fix bugs or add features.
Re:I do see a problem here, but it isn't Android (Score:3, Informative)
I'd go a step further and say the problem is that the low level driver interfaces are generally what's closed in these things. You CAN'T write a replacement OS because you can't get the information to talk to the hardware. Sure, the Linux kernel Android runs on is FOSS. But the drivers that make it possible to talk to the hardware aren't. It's the NVidia/TiVo model. And I have yet to see a real product, outside of the OpenMoko project (which appears to be dead), that doesn't suffer from the same problem. Including the much-discussed N900. Personally, I'd be happy to pay the asking price for the N900 for a phone that added 2 things. A decent amount of RAM (no, 256M isn't "decent" and swap/flash don't count) 1G minimum, I'd prefer 2G. And open driver interfaces. Either full open source for the drivers or a full set of documentation for every bit of hardware in that device. That includes the Wifi, cell radio, bluetooth, everything. If they want to write a proprietary GUI or apps, that's fine with me. It's the base OS that I believe should be open. Not just because it would be nice, but because then I can support the device into the future should I choose to. Even if the OEM decides it's not worth it.
Look at what's happening now with the next version of Android. The various devs are having a hard time getting it running on older phones like the G1. Not because it doesn't have enough resources, but because they don't have drivers for the radio, camera, etc. and the OEM hasn't seen fit to release any. I'm sure they would rather have us all throw out our older phones and buy new ones from them. And they probably want their driver devs working on the new stuff. That wouldn't bother me if I could at least port the drivers to a new kernel or other stack, but when we can't do it, and they won't do it, that just leaves users stuck.
Re:Maemo (Score:5, Informative)
Re:Maemo (Score:1, Informative)
Android by default runs code in a java-like VM with limited capabilities. Maemo runs binary executables directly within a linux kernel. Getting a console in android requires effort. In Maemo, you can just launch a console.
Re:Okay, I'll be the one to say it... (Score:5, Informative)
Here is a copy paste of an earlier post I made....
Here are some prices from one of the cheaper web stores in Finland. Please note that these have taxes included and probably the "europeans are idiots" bonus (1 dollar = 1 euro)
Based on these it would seem that most top of the line phones actually cost around 500 - 600 euro (that is probably 500$-600$ in US) and even correlates pretty nicely with release schedule. Don't get the price on the Sony Ericsson, though it isn't actually out yet I think.
BTW: People were able to get it as cheap as $442 from Dell [maemo.org] a while back. Don't know what is the cheapest now (nor would I buy anything from Dell :)
Re:Okay, I'll be the one to say it... (Score:3, Informative)
Unless of course you want to do something that Google/T-Mobile don't want you to do...
I think you meant, "Unless of course you want to do something that you agreed not to do when you signed up for service..."
Re:We've gone long enough without real progress... (Score:3, Informative)
In my 15 years or so with Linux and FOSS I have encountered many who think like you do, and unfortunately I think you are misguided. What you fight so heavily against is actually what you most desperately need. There's a saying "The stone that the builder refuse, will be the head cornerstone." It is true in this instance.
You skirted the issue that the Kernel itself REQUIRES this level of oversight or it would never work at all. It's all well and good to claim otherwise but the proof is right in front of everyone. The kernel has made massive leaps and bounds in the same time that the surrounding software has just slinked along. It too could get a massive boost, but we deny it.
Here's the part you are also missing: a base system, a foundation, with full oversight and steering, and vision, and goals, is not a threat to anything. It lays a solid foundation on which to build from in any way anyone sees fit. It wouldn't stop anyone from making a better widget or completely overhauling it for a specific need or want. But instead of just a kernel as the core, it would be the whole base system of some apps and config tools, and UI. From there anyone can still do their own thing. But some consistency and cohesiveness would go a long way in making it all better for everyone involved. Usability and look/feel would have a starting point. App interaction, updating, etc. would too. Some basic APIs and framework. How this can be scary or unwanted is beyond me.
Re:Okay, I'll be the one to say it... (Score:3, Informative)
I do not see that big of a difference between Opera and Firefox as far as speed. I have Opera on my system but I use Firefox all the time. I do not like Opera's UI as much as Firefox's and the Plugins for FireFox make it for me. Yes some are junk but that is always going to happen with plug ins or apps.
I don't miss that feature in Thunderbird but it would be nice. The new version of Thunderbird is a big step.
Gimp vs Paint.NET? Gimp is a lot more powerful than Paint.net. My wife is a big Gimp User and she also has Photoshop Elements and Paint.net, if you want a simple Paint program than yes Paint.net is probably better. If you want to really do a lot of heavy graphics work the GIMP is a lot better. If you need the super heavy duty graphics then CS4 wins.
But that it the point. None of those programs are JUNK. They are all very good tools and very cheap.
Paint.net is also available for free and you can get the Source. Is it FOSS? I have not checked to see if it really FOSS but hey it is free as in Beer and I can get the Source so it is pretty close and is also not junk.
The point is that not all FOSS is junk and not all Closed source programs are gems.
Re:Okay, I'll be the one to say it... (Score:3, Informative)
[quote]Firefox has its shortcomings too. Use of XUL makes the interface somewhat clumsy and slow. Quality of addons varies. I agree it is a lot better browser than IE, but if we compare to for example Opera you know all of it's features will be built with same consistent quality and work fast since they're integrated directly to the browser. There are of course opinions, but the UI responsivess is really easy to see and makes browsing a lot nicer.[/quote]
It's also XUL that enables the extreme extensibility of the UI by extensions. Yes, the are trade-offs, but I prefer Firefox's advantages over Opera's.
Re:Okay, I'll be the one to say it... (Score:3, Informative)
But, just like the desktop computing environment before, the commercial developers will be followed by OSS developers who just have an itch to scratch that no existing app handles...
Don't you mean that the commercial developers will follow the OSS developers? Because that's how it actually happened. Software was generally free of charge and often passed around until some companies decided to start making money from it, going all the way back to ancient versions of Unix and other hackish OSes like ITS, in the 1960s.
I'd tell you who to go talk to to verify such things, but they'd probably just tell you to get off their lawn...
Re:Uh...build your own free app? (Score:2, Informative)
Actually, the issue is that the FSF still hasn't solved the "how can developers eat and pay rent" problem. Free means free (it is, if you will, a subset of "free") and, well, that doesn't pay my bills.
Then again, I have gotten some money for offering "service and support", not nearly enough to pay the rent, but slightly better than nothing. I mainly use a "pay me for support" model to give annoying users "invoice therapy": If you want to be treated like a customer, you must first become a customer by paying me. You would be amazed at the number of idiots out there who think Free means "Implement the features I want or answer my email without getting paid for your time and work".
Re:Okay, I'll be the one to say it... (Score:1, Informative)
I'm the exact opposite - I don't like Firefox's UI. It feels very heavy and crude, but its been like that since the early Mozilla days. I found Opera's earlier UIs a bit off-putting, but I really like v10's. Its slick, out of your way, and very clean.