The title is a little misleading, it led me to believe that it was about Android app developers, not custom distro developers.
That being said, I'm a bit thorn on this. From the developer side, Samsung seem a bit too prickly, specially when Android is touted as an "open" platform. On Samsung's side, it has no obligation whatsoever to provide all the data and source when asked.
When you play in big business level, except from marketing statements, there's no such thing as open. In this context, open was touted by Google as an advantage over iOS just to get mind share. Give them time and they'll tear this notion apart. Samsung already started.
Obligation, no. But they seem to have "promised" devices and support for the CM dev team (I may have misunderstood that though.). They got the devices, but not the support. So i think it's fair enough of that guy to call them out on it. Don't promise things you won't keep. At least not to people that you want to keep around. (Whyever they would care about CM...)
In fact, as far as I understood, they do have an obligation, because the (GPL licensed) kernel requires them to release source code for some parts of the system
They've released the stuff their obliged to release under the GPL, that's not the problem. However they also promised to release more, and now apparently have no intention on following up on that promise.
I just hope that Motorola+Google will be a rolemodel in the future in this regard. The Google Nexus devices have been very good to go for modding. Now Motorola needs to do that as well. They could even make us of the modding scene to provide regular updates instead of what is currently the case with every manufacturer for non-Nexus devices.
I wonder if this might end up as a trend towards the 'leaders' on the Android scene moving away from the Samsung stable to another manufacturer, and if so will that lead to the community following? Samsung is doing fantastically well at the moment so it doesn't make sense for them to start picking up a bad reputation for not playing ball.
He said he planned to switch to the Xperia T, but I guess he will want to leave something at least workable for the S3 before that.
EDIT: not sure about the current state of things, seems that this is a translation from an interview that happened several months ago, which explains why the Xperia T wasn't out yet.
Yeah, I've kept a close eye on this since I own an S3 and run CM and he's definitely still working on it from what I've seen. That might be related to Samsung saying they'd be more open going forward though
That's partly it. TI is letting the OMAP5 out into the embedded market, where it has a lot more strength, through SOM partners like Phytec. This could look to some as a possible future subversion tactic.
The only hope for OMAP in the open-source/AOSP world is if maybe Amazon picks it up. Or, as a longshot, maybe someone making automotive systems (although Freescale is doing better there).
If Qualcomm or Broadcom happens to want it, um, I wouldn't hold out for any new releases.
I tend to agree with this. Samsung made a big event of declaring their support for developers of Exynos devices, but have stayed quiet and slow in showing that they meant it. The Mali GPU is one area where access is still locked down to a binary blob, and it doesn't look as if that will change.
The GPU is accessible under license to ARM partners according to ARM, though I've not seen it, and I think it's not Samsung's to give away unilaterally. There's competitive advantage issues involved no doubt, including the fact that Imagination GPU design, just down the M1 motorway from ARM, and used in the OMAP5, are locked down as well.
The Exynos 5 dual core is a great device, but as in so much to do with ARM licenses, the peripherals are the work of the licensee. Texas Instruments use dual Cortex-m4's along with A15's in the OMAP5 which is a very capable approach when combined with good peripherals.
I can see why Samsung would be cautious, but I do think they've been misleading in promising more than they could deliver. So far it just looks like a fork of Android Jellybean.
I understand his frustration. I can only imagine he spent many hours fighting to get his distro to work with his S3 and he now realized he has to give up. It's never a good feeling!
Before blaming OEMs, look to Google for leadership, or lack of it, in supporting Android as an open system by encouraging open drivers and documented peripherals.
Google has wielded the "Anti-fragmentation Agreement" as a whip to keep Android OEMs from straying into temptations like Aliyun. Surely a bit of love for AOSP-based Android distributions that grow the system integration talent pool are within the scope of not being evil.
This article highlights one of my frustrations with Android as a platform. As a longtime Linux user, I couldn't wait to get an Android smartphone. For some reason, I thought I would enjoy the same freedoms on a familiar and secure platform. But I shortly discovered that Google reinvented the wheel (poorly) for everything, manufacturers went to great lengths to lock down devices, and carriers blocked security and version updates even when the manufacturer had released them in other markets. The Android I had to endure for the last two years looked nothing like the Linux I used on my desktop and servers; it was more reminiscent of Windows 95. So, in an ironic twist, I replaced it with a Lumia 920 running Windows Phone 8 this past December. It isn't Linux either, but it excels as a smartphone and the UI is fantastic (I much prefer live tiles over icons littering a desktop). When I use it, I get the sense that Microsoft paid close attention to the things users hate about Android and iOS. If they can resist feature creep and retain their focus on simplicity and security, they might have a real contender. It's too bad Google didn't exert more control over the user experience in Android and allowed manufacturers and carriers to create such a mess that exists today.
Heck, you can even install launchers that clone the Windows Phone interface entirely.
I'm kind of surprised Microsoft hasn't released an Android UI library for Modern apps on Android, as I think it could push more Android users into choosing Windows Phone when its time to purchase their next device.
They still make it damn impossible to run the windows phone simulator on Mac (even with a virtual machine, can't run the game part), so no iOS devs bother with the platform.
You're making some argument from fairness here. I'm talking about MS's lackluster ability to entice mobile developers.
It may not be fair the people who came out with a C# only phone 3 years after the iPhone can't convince the vast majority of mobile developers to swap, or even develop for it too, but they can't because they made their tools require onerous effort to get up an running.
It most certainly IS a double standard. You have a different standard trying to take away from the incumbent. The incumbent MOBILE development platform is a mac, running OS X, and this is because mobile developers want to release on iOS (as there is a ton of money there and it was "the first"), and Android is supported great on mac as well.
To do Windows phone development, buying a virtualizing tool such as VMWare isn't good enough: The simulator only half works for windows phone 7, it can't use any of the 3D APIs.
So in the age of macbook airs (with small SSDs) and developers who replaced their optical drives with SSDs in MBPs (A HUGE pile of people were doing this in 2010-2011), they're saying "Give us a huge space to install windows and the tools, and go through complicated basecamp setups to make our shit work".
When told to do all this work, THEN having to port our games and apps (which can often be made cross-platform to Android with far less work), most of us just skipped the platform.
I have been at 9 events where Microsoft developer relations people have done WinPhone demos to iOS devs. Events with >50 people, all iOS devs of some sort.
Woah woah. Hang on. I used to work in this area so I want to comment. First, the mobile market has of late always been a mess. The whole Apple-introduced smartphone era is basically about who owns the customer: the carrier or the device manufacturer. Clear winner? Not yet.
Google's whole play on the Android thing was to ensure Apple didn't triumph too far, and they've succeeded. It was never about putting Linux on phones. Look, Nokia's died. (Though last time I mentioned this someone on HN got upset.) The other device manufacturers (including Smart TVs now) are essentially presently dependent on Google (though Samsung is trying very hard not to be), and our mobiles are all online without us buying Apple devices.
Second, the whole lockdown thing is dual-faced. First, the carriers in the US and other markets that subsidise phones (advertising: $0!) with long term lock-in contracts don't want people hacking the phone. Therefore there's a requirement for them to make it a bit hard to open up. On the other hand, often a content licensing requirements from content partners force them to include this stuff, typically Hollywood studios. (I have it on good information that Microsoft paid a hundred million dollars up front to Samsung to convince them to introduce their new DRM solution in the Galaxy series. So the DRM thing is basically a moot point, if you are paid that much!) And finally, the devices do need some secure storage for cryptographic purposes, for example storing passwords.
Net result: the consumer is shafted. Thanks, industry. I hope FirefoxOS can get some traction.
Here's what I believe 'should' happen:
- Carriers realize their value-add is the consumer relationship, actually provide decent billing interfaces to all and sundry, replace PayPal and the credit card companies. (Problem: They can't. Their internal billing systems are generally a mess, custom hacks layered on to black boxes they outsourced to AMDOCS)
- Device manufacturers should embrace openness, encouraging people to load whatever they want with an OPTIONAL vector for validating cryptographic trust of a codebase. The manufacturers should actually invest in facilitating this. In most of the world, people are super keen to muck with phones, there is a community that will fix bugs and improve things, and those who sell the devices have time to customize.
- AOSP people should join FirefoxOS and work for a more open world instead of working on Google's pretend-open platform (step 1: upload everything to google, step 2: see step 1) for free. Work on there a year or so, then pimp yourself out to a device manufacturer. Guaranteed cash, feel good about your work, help break the unhealthy Googopoly.
Carriers seem to be the problem here. That CM dev is moving to Sony, but I'm having the same crap over there. My Xperia U is still running Gingerbread even though Sony released ICS for that device 4 months ago (and the bootloader is locked by my carrier). Why carriers get to choose what OS runs on my phone is beyond me. My ISP doesn't get to decide which OS I run on my computers, I don't see why this would apply for cell phone carriers.
Carriers are totally the problem. Yesterday, Vizio announced two phones running stock Android (Jelly Bean) at CES. They look gorgeous, and The Verge staff, which has had hands-on time with them, says they're amazing. They would use these phones as their daily drivers.
But these phones aren't being released in America - they're being released in China. Why? Because the American carriers won't play ball (because the phones run stock Android), so they're forced to take their product overseas. In China, you can just buy a phone, put a SIM card in it, and you're off to the races.
Think about that for a second. A rare American hardware upstart success story is forced to sell its first phones in China because American carriers won't accept them.
I agree except that when you say "in China", what you actually mean is "in most of the world". You see, when we have less hegemonic mobile network operators (due basically to their inability to hunt down and nail people who reneg on long term phone contracts), we have a more heterogeneous system. In the end, everyone benefits. What we need is less authority, more internet.
Hopefully GSM will die soon and we can all use some kind of public, tor-rerouted, bandwidth-metered cryptophones over wifi. I'm sure the NSA will love that.
Surely you mean "CDMA will die soon". That's the American technology behind the American idea of "this phone belongs to this carrier". The carrier decoupling that you see in Europe and elsewhere was enabled by GSM-family technologies (GSM, UMTS, LTE) and their use of SIM cards.
I believe there is a separate baseband firmware (drives radio chip) to the system firmware (Android; drives UI and operating system). My impression is that, in normal cases, the former is loaded in the factory and is never altered, and Android applications essentially view the network through a cellular-protocol-agnostic API.
So what's the key difference in the operating environments of mobile carriers in China (and much of the rest of the world), where choice of handset seems to be fairly decoupled from access to the network?
For Europe, I think it was having lots of relatively small countries. That led to lots of national phone companies which had a monopoly in their country. To facilitate international calls, they had to cooperate. When wireless networks started, none of them could build a network covering significant parts of Europe. Again, they (eventually; there were some early networks that were incompatible with each other) had to cooperate.
Next/somewhat in parallel, the EU forced national phone companies to be privatized and allow competition.
The net result is GSM, where companies share their networks, and subscriptions are tied to SIM cards.
For China, I guess they just made the right choice, looking at the USA and Europe as options.
1. Chinese carriers are all government-run, basically. There's basically two: China Unicom and China Mobile. Almost everyone used to use the latter, but Unicom is getting better recently (last few years). It takes a lot of effort to follow though, so I'm not 100% on all that (I'm back to China for Chinese new year in a couple of weeks, living in Thailand now!)
2. Nearly everyone uses prepay.
3. Most people aren't living in debt.
4. There is a massive market for cheap and/or stolen phones.
5. A not insignificant number of phones, particularly at the luxury end, are 're-routed' from the Hong Kong 'export' market back in to the mainland.
6. Chinese save a lot of money and don't tend to live beyond their means as much as Americans.
7. The entire mobile infrastructure is generally less than ten years old.
8. The Chinese produce their own infrastructure hardware (chiefly Huawei).
9. The Chinese government wisely sees infrastructure as a political and security tool, and so probably subsidises it in many ways. (This also goes for roads, electricity, etc.)
10. Coverage is far better than the US. You almost never have no coverage, even waaaay out in the hills. And in 10 or so years, I've never had a dropped call.
Prepay doesn't seem like something that would drive a market difference in the US. The carrier prepay options are much more limited here.
Similarly, I wonder how the the age or creator of the mobile infrastructure (7 & 8) figures into offering of mobile devices independent of network access by carriers - maybe in the homogeneity of the mobile network?
The reason they get to choose is because you let them. Stop buying subsidized phones that come with multi-year contracts. They use the device to bait you into the contract, once they've got you they have no reason to try to improve the device in any way.
You seem not to have read the linked article, which is about one of the CM developers dumping work on one of the GS3 (there are a bunch of different phones labeled with the same name) devices because of unsolvable problems and poor support from Samsung.
Seriously: Buy. A. Nexus. Phone. Almost everything else is going to bite you.
Sure, there are issues with certain GS3 models. That doesn't mean that, as a consumer, you can't get one that works with CM. Plus, I have no special loyalty to Samsung or the Galaxy brand. It could be an LG, or HTC, or Sony, for all I care.
And getting a Nexus isn't an option for me, not anymore. I have a Galaxy Nexus on Verizon (a great phone) at the moment, and none of the other carriers have as good coverage where I live. They definitely don't have 4G, which Verizon does. Plus, I have grandfathered unlimited data, which I'm never giving up.
I'd love to get another Nexus, but that's not an option for me. After the fiasco with the Galaxy Nexus on Verizon, I don't see Google releasing another Verizon Nexus, not for a long time.
So I'll probably buy a phone that's 6-8 months old, has been unlocked/rooted, is quite popular, and was a premium model when it was released. Then I'll install CM on it and go on my merry way.
Actually I have a not-quite-a-nexus-anymore "Verizon Galaxy Nexus" (toro) in my pocket, so I feel your pain. If anything this supports my point, doesn't it? I tried to follow the "Just Buy A Nexus" advice and still got burned because Verizon/Google couldn't support the device on AOSP.
Though to be fair, the CM 10.1 builds on toro work just fine. I sweat a little at every update though.
Interestingly, I have a GS3 as well and have no problems with the things the developer is mentioning here.
The article is 3 months old. In that time, CM10 hit stable release for all US carriers (and international, IIRC) and is honestly rock-solid. All those "unsolvable" issues were solved. I'd recommend it without hesitation to every GS3 owner.
10.1 is buggy/crash-prone, admittedly, but the team's only been working on it for a few weeks. Bluetooth and Wifi tethering work solidly, some of the newer Google Apps (like Gallery and Photosphere) are a little hacky. Touch to focus works great.
Every issue Codeworkx cites as his frustrations with the S3 have been resolved by the great folks at the CM/TeamHacksung. The article is a non-issue.
I'm willing to bet that you have a US market "GS3". That's a Qualcom Snapdragon S4 phone, not Exynos (seriously: it's an entirely different set of electronics in the same box). It works much better largely because it shares the same SoC as the Nexus 4.
And yet, in comparison with it's cousin the N4, it's not nearly as well supported or trouble free. Buy. A. Nexus. Phone.
Nexus 4 has one major flaw -- you can't tell people to buy a phone that is not on the market. :) I can wait until my n900 breaks down (which won't happen any time soon, that thing is built like a rock), but N4 would be so much better if it was, you know, there.
They have the same amount of RAM if you're comparing the American S3 and I'm assuming you are since you're implying the model of S3 in your mind is dual core (American S3 is dual core).
Also, the N4 had an extra 48 lines of pixels that are used for the navigation bar, so effective screen real estate is the same on both devices.
You can enjoy all the freedom you want. Just root the phone (and install custom ROM, if you want). Carrier and manufacturer lock is then irrelevant. The only problem is locked bootloaders, but you can avoid that also - buy a phone that hasn't got a locked bootloader.
Nice to hear that you like WP8 UI better, but it's got nothing to do with freedom or linux.
"Just rooting the phone" is sometimes much easier said than done, even for technically savvy people. I tried to root my LG Thrill (an absolutely horrid phone, so bad LG abandoned it 6 months after release instead of fixing its very serious problems) but I found that my Windows 7 PC would not run the utility required to root the phone. It turns out that it only reliably runs in Windows XP. I'm not about to go buy a license for a 10-year-old OS just to root a phone.
While in the beginning you say you wanted more freedom like in the Linux ecosystem, everything you said afterwards contradicts what you said, and it seems you wanted the exact opposite - you wanted the company behind it to be in full control - and not really yourself to be in control of it.
If you look at the Linux "ecosystem", it's a lot more fragmented than Android, and every vendors does whatever they want to their own distro, and can completely change the UI, and they don't even have the same app extension across distros. So yeah, it seems to me like you wanted the opposite of all of that all along and not really the freedom/flexibility of such an ecosystem (that by default comes with many people doing different things in the ecosystem).
Actually I kind of agree. They should of developed an easier way to bring HWA to the UI from the beginning instead of waiting until 2012 to address the issue. Google also messed up the pro audio stack in Android and phones still cant compete with Apple/MS when it comes to low-latency audio. "Google reinvented the wheel (poorly)" is unfortunately a very true statement for many of their products.
Ubuntu phones are DOA. Whatever they promise us now will not be what potential customers expect from a smartphone in six months, let alone a year+. The mobile market changes at such an insane rate. They have an incredible amount of ground to cover, deals to make, and markets to enter if they expect to get any ROI at all.
Is this a linkbait title or do I misundertand things? The article is about how Samsung does not actively support Cyanogenmod. Translating that to "screwing over Android developers" is like saying that Nintendo screws over Linux developers by not making it easy to install Linux on a Wii.
Except Samsung promised they would work with developers, whereas Nintendo did no such thing.
Samsung is becoming very big and powerful, and seems to be acting very different than they have in the past, just like other big companies such as Apple, Microsoft, Google.
Yes, with heavily customized UIs. Rather distinct from the standard Android experience. Personally, I find it annoying, and it doesn't add that much value for me. I liken it to the bloatware installed on Windows machines in the past.
77 comments
[ 3.8 ms ] story [ 150 ms ] threadLinks to the other parts are contained in each post.
That being said, I'm a bit thorn on this. From the developer side, Samsung seem a bit too prickly, specially when Android is touted as an "open" platform. On Samsung's side, it has no obligation whatsoever to provide all the data and source when asked.
EDIT: not sure about the current state of things, seems that this is a translation from an interview that happened several months ago, which explains why the Xperia T wasn't out yet.
Um, isn't TI exiting the phone market?
The only hope for OMAP in the open-source/AOSP world is if maybe Amazon picks it up. Or, as a longshot, maybe someone making automotive systems (although Freescale is doing better there).
If Qualcomm or Broadcom happens to want it, um, I wouldn't hold out for any new releases.
The GPU is accessible under license to ARM partners according to ARM, though I've not seen it, and I think it's not Samsung's to give away unilaterally. There's competitive advantage issues involved no doubt, including the fact that Imagination GPU design, just down the M1 motorway from ARM, and used in the OMAP5, are locked down as well.
The Exynos 5 dual core is a great device, but as in so much to do with ARM licenses, the peripherals are the work of the licensee. Texas Instruments use dual Cortex-m4's along with A15's in the OMAP5 which is a very capable approach when combined with good peripherals.
I can see why Samsung would be cautious, but I do think they've been misleading in promising more than they could deliver. So far it just looks like a fork of Android Jellybean.
Google has wielded the "Anti-fragmentation Agreement" as a whip to keep Android OEMs from straying into temptations like Aliyun. Surely a bit of love for AOSP-based Android distributions that grow the system integration talent pool are within the scope of not being evil.
I'm kind of surprised Microsoft hasn't released an Android UI library for Modern apps on Android, as I think it could push more Android users into choosing Windows Phone when its time to purchase their next device.
They still make it damn impossible to run the windows phone simulator on Mac (even with a virtual machine, can't run the game part), so no iOS devs bother with the platform.
Microsoft DOES go around to iOS devs.
It is ok for Apple to force me to buy a Mac, but not for Microsoft to do the same?
At least Microsoft does not force me to buy a specific hardware for their operating system.
Talk about double standards...
It may not be fair the people who came out with a C# only phone 3 years after the iPhone can't convince the vast majority of mobile developers to swap, or even develop for it too, but they can't because they made their tools require onerous effort to get up an running.
It most certainly IS a double standard. You have a different standard trying to take away from the incumbent. The incumbent MOBILE development platform is a mac, running OS X, and this is because mobile developers want to release on iOS (as there is a ton of money there and it was "the first"), and Android is supported great on mac as well.
To do Windows phone development, buying a virtualizing tool such as VMWare isn't good enough: The simulator only half works for windows phone 7, it can't use any of the 3D APIs.
So in the age of macbook airs (with small SSDs) and developers who replaced their optical drives with SSDs in MBPs (A HUGE pile of people were doing this in 2010-2011), they're saying "Give us a huge space to install windows and the tools, and go through complicated basecamp setups to make our shit work".
When told to do all this work, THEN having to port our games and apps (which can often be made cross-platform to Android with far less work), most of us just skipped the platform.
So I am still waiting for a Mac hardware free version of Mac OS X.
"Normal" iOS dev rates pay for a macbook air in a day of work. Easy to get rates pay for a macbook air in 2-3 days of work.
Google's whole play on the Android thing was to ensure Apple didn't triumph too far, and they've succeeded. It was never about putting Linux on phones. Look, Nokia's died. (Though last time I mentioned this someone on HN got upset.) The other device manufacturers (including Smart TVs now) are essentially presently dependent on Google (though Samsung is trying very hard not to be), and our mobiles are all online without us buying Apple devices.
Second, the whole lockdown thing is dual-faced. First, the carriers in the US and other markets that subsidise phones (advertising: $0!) with long term lock-in contracts don't want people hacking the phone. Therefore there's a requirement for them to make it a bit hard to open up. On the other hand, often a content licensing requirements from content partners force them to include this stuff, typically Hollywood studios. (I have it on good information that Microsoft paid a hundred million dollars up front to Samsung to convince them to introduce their new DRM solution in the Galaxy series. So the DRM thing is basically a moot point, if you are paid that much!) And finally, the devices do need some secure storage for cryptographic purposes, for example storing passwords.
Net result: the consumer is shafted. Thanks, industry. I hope FirefoxOS can get some traction.
Here's what I believe 'should' happen: - Carriers realize their value-add is the consumer relationship, actually provide decent billing interfaces to all and sundry, replace PayPal and the credit card companies. (Problem: They can't. Their internal billing systems are generally a mess, custom hacks layered on to black boxes they outsourced to AMDOCS) - Device manufacturers should embrace openness, encouraging people to load whatever they want with an OPTIONAL vector for validating cryptographic trust of a codebase. The manufacturers should actually invest in facilitating this. In most of the world, people are super keen to muck with phones, there is a community that will fix bugs and improve things, and those who sell the devices have time to customize. - AOSP people should join FirefoxOS and work for a more open world instead of working on Google's pretend-open platform (step 1: upload everything to google, step 2: see step 1) for free. Work on there a year or so, then pimp yourself out to a device manufacturer. Guaranteed cash, feel good about your work, help break the unhealthy Googopoly.
But these phones aren't being released in America - they're being released in China. Why? Because the American carriers won't play ball (because the phones run stock Android), so they're forced to take their product overseas. In China, you can just buy a phone, put a SIM card in it, and you're off to the races.
Think about that for a second. A rare American hardware upstart success story is forced to sell its first phones in China because American carriers won't accept them.
Hopefully GSM will die soon and we can all use some kind of public, tor-rerouted, bandwidth-metered cryptophones over wifi. I'm sure the NSA will love that.
There's not much sense in fighting for something that is largely irrelevant to most people.
Next/somewhat in parallel, the EU forced national phone companies to be privatized and allow competition.
The net result is GSM, where companies share their networks, and subscriptions are tied to SIM cards.
For China, I guess they just made the right choice, looking at the USA and Europe as options.
1. Chinese carriers are all government-run, basically. There's basically two: China Unicom and China Mobile. Almost everyone used to use the latter, but Unicom is getting better recently (last few years). It takes a lot of effort to follow though, so I'm not 100% on all that (I'm back to China for Chinese new year in a couple of weeks, living in Thailand now!)
2. Nearly everyone uses prepay.
3. Most people aren't living in debt.
4. There is a massive market for cheap and/or stolen phones.
5. A not insignificant number of phones, particularly at the luxury end, are 're-routed' from the Hong Kong 'export' market back in to the mainland.
6. Chinese save a lot of money and don't tend to live beyond their means as much as Americans.
7. The entire mobile infrastructure is generally less than ten years old.
8. The Chinese produce their own infrastructure hardware (chiefly Huawei).
9. The Chinese government wisely sees infrastructure as a political and security tool, and so probably subsidises it in many ways. (This also goes for roads, electricity, etc.)
10. Coverage is far better than the US. You almost never have no coverage, even waaaay out in the hills. And in 10 or so years, I've never had a dropped call.
Similarly, I wonder how the the age or creator of the mobile infrastructure (7 & 8) figures into offering of mobile devices independent of network access by carriers - maybe in the homogeneity of the mobile network?
Seriously: Buy. A. Nexus. Phone. Almost everything else is going to bite you.
And getting a Nexus isn't an option for me, not anymore. I have a Galaxy Nexus on Verizon (a great phone) at the moment, and none of the other carriers have as good coverage where I live. They definitely don't have 4G, which Verizon does. Plus, I have grandfathered unlimited data, which I'm never giving up.
I'd love to get another Nexus, but that's not an option for me. After the fiasco with the Galaxy Nexus on Verizon, I don't see Google releasing another Verizon Nexus, not for a long time.
So I'll probably buy a phone that's 6-8 months old, has been unlocked/rooted, is quite popular, and was a premium model when it was released. Then I'll install CM on it and go on my merry way.
Plus most of CM gripes with the S3 are related to Exynos SoC and not the Qualcomm SoC.
Though to be fair, the CM 10.1 builds on toro work just fine. I sweat a little at every update though.
The article is 3 months old. In that time, CM10 hit stable release for all US carriers (and international, IIRC) and is honestly rock-solid. All those "unsolvable" issues were solved. I'd recommend it without hesitation to every GS3 owner.
10.1 is buggy/crash-prone, admittedly, but the team's only been working on it for a few weeks. Bluetooth and Wifi tethering work solidly, some of the newer Google Apps (like Gallery and Photosphere) are a little hacky. Touch to focus works great.
Every issue Codeworkx cites as his frustrations with the S3 have been resolved by the great folks at the CM/TeamHacksung. The article is a non-issue.
And yet, in comparison with it's cousin the N4, it's not nearly as well supported or trouble free. Buy. A. Nexus. Phone.
I have an N4 and I am very happy with it, the extra RAM and the quadcore processor make a big difference.
Also, the N4 had an extra 48 lines of pixels that are used for the navigation bar, so effective screen real estate is the same on both devices.
Nice to hear that you like WP8 UI better, but it's got nothing to do with freedom or linux.
If you look at the Linux "ecosystem", it's a lot more fragmented than Android, and every vendors does whatever they want to their own distro, and can completely change the UI, and they don't even have the same app extension across distros. So yeah, it seems to me like you wanted the opposite of all of that all along and not really the freedom/flexibility of such an ecosystem (that by default comes with many people doing different things in the ecosystem).
Samsung is becoming very big and powerful, and seems to be acting very different than they have in the past, just like other big companies such as Apple, Microsoft, Google.
damn you all to hell.