It's unbelievable that they were not able to capitalize on all those Android devices that are still running very old versions of Android. Given that Google and the device manufacturers basically disowned those devices, I was expecting (or rather hoping) that someone will fill in that gap/need. I guess its more complicated than it seems, but I do wonder how.
I think the problem with supporting "all those Android devices" is the long tail Android has [1] once you get past supporting Samsung/Motorolla/HTC devices.
How exactly would one make money off abandoned android devices? This users would pay?
They seemed to be considering partnering with advertisers to integrate ads into the base OS. Question is, who would voluntarily sign up for such a thing?
I would. I have a GalaxyTab 10 (the abandoned Google I/O edition) still running Honeycomb that I would happily pay $10 to easily update it to something vaguely more modern. It's been useless for years, not because the hardware is bad, but the software is useless. I still can't figure out how to get something more modern on it. It's been sitting in a drawer (plugged in, oddly) collecting dust forever.
Well that's the problem right there - you'd pay a whole 10$ and from the tone of your post, you consider that generous. The amount of people who even know or care about Android version on their devices, intersected with the amount of people who'd pay reasonable rates, is tiny.
What's so special about Nexuses in this regard? I find the duration of support still pitifully short on them even if it's slightly better than what some other phones manage. Whether or not people care about their version, everyone should if only because of security.
It's not as bleak as that. There are a lot of people with manufacturer-abandoned phones that would probably be willing to shell out a modest sum to get less ancient firmware. As with everything, the economics are a question of quantity and effort. A billion Android devices enter the marketplace each year.
I think the real problem is that it is just turning out to be difficult to make a silky smooth upgrade process, the kind that is consumer-friendly enough to create a mass market. It all fundamentally starts with "hack the device" and most people can't. Maybe it's just a really hard problem because it varies from device to device and base OS version to base OS version.
You might have specific usage patterns where shows but I would say that there's anything wrong with Honeycomb but, rather, that the major problem is that the hardware itself doesn't really stretch enough to run newer Androids or newer websites/applications. I'd just keep the less resource-hungry Honeycomb installed until the hardware can no longer do modern content.
A big chunk of apps require at least 4.0. Or maybe they don't but if you have an older version you will have to use an older version of the app with less features.
You can always find ways to monetise an operating system. Some possibilities include:
* add products/referral links to os search results (like ubuntu lenses)
* add ads to the base operating system (like kindle's sponsored homescreen)
* sell spots for default search providers (like bing for siri)
And you target old phones by providing them with an upgrade path to cyanogen mod based on a more modern android version. Many devices are technically fully capable of running modern android versions but have been abandoned by their manufacturers.
Have you tried rooting and re-imaging an android phone with a custom rom? I'm an experienced software engineer and it took me a painful while to make it happen. I supposed they could make it easier, but the amount of effort that would go into maintaining install kits for every phone out there would be crazy.
Sony are great - key, unlock bootloader, flash twrp, install. Not that much harder than installing linux on modern laptop - enter bios, disable secureboot, curse 4 times until you find the second secure boot setting you also need to disable, exit bios, curse till you find the boot menu hotkey for this exact laptop, boot, install
I actually just installed CyanogenMod on my Galaxy Nexus last week, and it was easy as pie. Is there something I'm missing? Or are newer Samsung phones more fettered?
Note that this still does you comparatively little good if you don't have CyanogenMod to put on it. Mind, I didn't say "no good", I said "little good".
It's the CyanogenMod dev team at Cyanogen Inc who've been impacted by the layoffs. There's still the independent effort, but it's lost a major employer and investor.
Good to know. I'll keep that in mind for my next phone.
Even if CyanogenMod suffers, there's plenty of other open-source Android versions out there. Unfortunately finding a clear, unbiased comparison of them is even harder than it is for different Linux distros. :(
I had to go through the goldcard process to flash my first Android phone, it took a while to get the right microSD card and you also had to install the Android SDK to use their flashing utilities, which wasn't the easiest thing to get up and running at the time.
I got there in the end, and I'm pretty sure most newer phones can be flashed in far easier ways.
I agree it's a pain - I've installed Cyanogenmod on a few Nexus devices (which are supposed to be the easiest ones) and it wasn't simple.
However is it intrinsically hard? Could you imagine a software tool you would run which would ask a few questions, automate as much as possible over the OTG port, and guide the user for the rest?
That's the kind of tool Cyanogen Inc could write. Of course monetizing it is another thing.
I think it is just a hard problem. Many devices with many Android versions and many provider customizations. And, oh yeah, the carriers are often going out of their way to make rooting as impossible as they can. Screw up badly enough and you can easily brick a phone.
They had 136 people. Solving the hard problems is what makes companies successful and what proprietary companies are good at, or so I'm always told. All the information you need exists on the Cyanogenmod wiki pages, it's a matter of grinding that into code.
I've been wanting to do something like that for a long time, get away from the recoveries and politics of the different factions. Make something a user can just pick up an old device, run an install process, and have a fast new OS.
Google of course doesn't make it easy, mostly due to carrier and manufacturer demands. But also due to the lack of a defined ABI between the user interactive parts and the libraries for talking to hardware.
There's really no reason an operating environment couldn't be compiled for ARM and run on the majority of devices if this specification was enforced. I'm looking to gonk and a new Google initiative as a possible solution.
It may be intrinsically hard. The one time I installed Cyanogenmod, I rooted the phone thanks to a misconfiguration resulting in local privilege escalation. Now, if only the phone manufacturer or the carrier cared to close this vuln with an OTA... Cyanogen could find themselves in a security arms race, where they are the attackers. I don't even know if that wouldn't make it illegal to sell their kits in some countries.
I remember a time when apparently you could just navigate to get.cm on your phone and it would download an .apk specific to your phone which would perform the rooting/unlocking and install of cyanogenmod. But that effort seems to have failed. http://www.xda-developers.com/how-to-install-cyanogenmod/
You say that as if there aren't countless pitfalls that can stop the process, e.g. phone not rooted, installing the tools to do it can be complicated, differences across operating systems, USB permission issues with the android tooling, tooling just not working for whatever reason, phone not behaving the way its supposed to once you are in the bootloader, recovery image doesn't install, the ROM download is bad, etc etc etc
These are not random examples - they ALL happened with the phone I am now typing on.
Its also not something a business could be built on top of, which is the original topic of this thread. Users expect single click installs.
Rooting and unlocking the bootloader are different. Root is having root on the OS. The Bootloader is what loads the various partitions. If your phone's bootloader is locked obviously flashing is a lot harder.
The mobile phone market is massive. Even if you can capture a small subset of the users you can make money. Not everyone needs a one step install process.
Problem is that on some phones, for instance my S4, if you don't root first, you can't install an alternative recovery ROM. Without an alternative recovery rom, you can't install a new OS that is not signed, so you can't install CM.
It costs too much in engineer time to maintain just the hardware support to do that. This will continue to be the case and also block GNU/Linux on these devices until the manufacturers start to send drivers and patches to Linux mainline before selling their devices. Unfortunately the hardware world moves so fast that the devices will be obsolete before they are supported by Linux mainline.
The amount of work you need to bring an Android device to a modern version is very, very big.
Does a new kernel version work? (Android will require it). Will the device drivers work on a new kernel version? Will driver blobs be compatible?
Chipset manufacturers will definitely not send you a new version of a driver if you're not the mobile manufacturer (and sometimes, not even if you are). You're on your own, updating the driver, or writing a shim that acts as a proxy between the driver, kernel, Android, etc. It's just not feasible.
CyanogenMod is open-source, though. They can sell a value-added version with extra apps and stuff, but unless they come up with something really amazing, most people will still just use the free version.
It would be recommended to charge for Cyanogenmod instead of giving away the operating system for free.
Further it would be recommended from a revenue point of view to support the Android devices with the biggest user base. Samsung, Huawei, Xiaomi. These devices should be super easy to upgrade.
> Layoffs reportedly came after a long executive retreat for the company's leaders and were conducted with no advanced notice. Employees who were not let go were told not to show up to work today. Those who did show up were the unlucky ones: they had generic human resources meetings rather ominously added to their calendars last night. So, everyone who arrived at Cyanogen Inc. in Seattle this morning did so to lose their job (aside from those conducting the layoffs). That's a bit grim.
It's like startups are looking to one-up each other in terms of "crappy ways to treat employees you're letting go."
I am too but I'm not sure a drawn out layoff process benefits anyone. When looking from either side of the table, I think I'd prefer short, sharp and with good severance.
I disagree. I once worked for a services company that basically had a "6 month notice" policy, meaning if they let someone know July 23th 2016, he still has a job until January 31 2017. This gives people the opportunity to start looking for work elsewhere, without even having to reveal that they "got fired". And it's just not as much of a psychological shock for the employees. Plus, if the firing was not performance related, but due to the company not being hired for enough projects, this was something that could change after 4+ months if the company had a lucky streak, and it sometimes did happen.
What do you find to be wrong about this situation, and what kind of protection are you looking for? How would it help in this scenario - force Cyanogen to keep the employees on, and go bankrupt, or lay off most employees and probably still go bankrupt (but with some chance of survival)?
Well. To fire someone from one day to the next is nearly impossible here in Germany. You always have some time (a minimum being defined by the law) of time between the day they tell you of their decision and your last working day. This time is there for your protection to enable you to seek another job.
That was what I was talking about. If I can get fired from one day to the next a company would at least have to pay me a hefty premium as I would be not better of then say a freelancer.
Personally I will not buy a phone that does not support CynagonMod (or visa-versa). Not sure on the long-term impact of this. Yes and I would gladly pay to get Cyanogen on my phone (with root, but they don't so I had to replace it with CynagonMod).
As someone who's dealt with this for years and years, let me tell you CyanogenMod is crap.
1) They don't actually support any devices. All devices, except the ones they sell (like that Chinese one they struggled for over one year to release, the OnePlus) are community supported.
2) Their own stuff (everything they put on top of AOSP) is of a very very bad quality, code-wise and UX-wise.
3) They are highly irresponsible. They have a Jira bug tracker with triagers that don't know anything about coding or Android at all. They have a Gerrit code review system where they antagonise all external contributors, nitpicking on everything, while Cyanogen, Inc. employees push everything they want into the tree without any code review, breaking stuff in the process (something that happens very often).
Ouch, well I'm new to android, feel free to recommend any other android OS that doesn't depend on the manufacturer to update them. (A lot don't do this).
Root is interesting. I want it to be able to more easily develop on the device, mount loop filesystems, etc. But I think it would be acceptable to have a container or chroot with a restricted root user and possibly shared storage accessible through a bind mount.
Maybe installing images could be supported through a self-signed key and fastboot.
Is there anything device-wide root is needed for on modern Android?
I hope the OS team starts up as a new firm. Cyanogen absorbed many players in the ROM community and there's literally no alternative unless you want to use a Chinese ROM. No offence to Xiaomi - but their quality was lacking. And they largely relied on CMs work.
This is a massive blow to the Android ecosystem and one that would move them closer to that of iOS. AOSP doesn't stand up to proprietary Android.
Cyanogen had a great team and a very promising business model . Where they screwed up badly is relationships and partnerships with OEMs and vendors.
The massive fiasco with Oneplus (which by the way is going strong) or the short-lived partnership with Micromax-Yu which was terminated for reasons unknown. I think they got greedy too fast without building loyal customers. At some level they believed that customers would have no real choice for older devices .
What instead happened was that Xiaomi and Oneplus completely destroyed incentives to hold on to older phones for any reasonable amount of time. And Cyanogen had already burned those bridges.
I am willing to bet CM will either build their own phone (the cyanogen "Nexus") or get acquired by Google or someone - who anyway are looking for some kind of replacement for the Android One/Silver program.
I bet all those who are being let go are partner services.
Quite a shame really, although my main phone is an iPhone I did like how their spin of Android made it a bit easier to de-google the OS. Sorry to anyone out of a job due to this.
They are an arrogant company with zero suspicion of the limits of their own knowledge. Before they started spouting off about their "We're putting a bullet through Google's head" nonsense I tried engaging with their management. About an interesting phone concept that I have / had. After they said that, it was like, what's the point. So I gave up trying to get their attention and went another route.
Long story short, I'm working with a Chinese company on my phone product. I have a prototype sitting on my desk that's been produced on a skeleton budget. It's better than anything I've seen Cyanogen ship.
If any ex-Cyanogen engineers are interested in a chat, hit up the email in my profile. :)
79 comments
[ 3.3 ms ] story [ 142 ms ] threadWorking for a bigger company means you are less in control of your destiny.
[1] http://thenextweb.com/insider/2015/08/05/this-is-what-androi...
Samsung is bad enough. Take a look here [2]. Click "Tablets" and then choose "Galaxy Tabs" as the product.
I'm sure a lot of the devices are similar at a hardware level, but still, the sheer volume of them is ridiculous.
[2] http://www.samsung.com/us/support/downloads
They seemed to be considering partnering with advertisers to integrate ads into the base OS. Question is, who would voluntarily sign up for such a thing?
I would. I have a GalaxyTab 10 (the abandoned Google I/O edition) still running Honeycomb that I would happily pay $10 to easily update it to something vaguely more modern. It's been useless for years, not because the hardware is bad, but the software is useless. I still can't figure out how to get something more modern on it. It's been sitting in a drawer (plugged in, oddly) collecting dust forever.
* Care about the Android version on their devices
* Would pay reasonable rates
* Stick with an older phone / don't buy Nexuses
I think the real problem is that it is just turning out to be difficult to make a silky smooth upgrade process, the kind that is consumer-friendly enough to create a mass market. It all fundamentally starts with "hack the device" and most people can't. Maybe it's just a really hard problem because it varies from device to device and base OS version to base OS version.
You might have specific usage patterns where shows but I would say that there's anything wrong with Honeycomb but, rather, that the major problem is that the hardware itself doesn't really stretch enough to run newer Androids or newer websites/applications. I'd just keep the less resource-hungry Honeycomb installed until the hardware can no longer do modern content.
* add products/referral links to os search results (like ubuntu lenses)
* add ads to the base operating system (like kindle's sponsored homescreen)
* sell spots for default search providers (like bing for siri)
And you target old phones by providing them with an upgrade path to cyanogen mod based on a more modern android version. Many devices are technically fully capable of running modern android versions but have been abandoned by their manufacturers.
That's news to pretty much every *ix vendor except Apple, Red Hat, and maybe Canonical.
It's actually massively soured me on Android (as well as several vendors, yes, specifically you, Samsung) as a result.
The Samsung hardware, fetters and chains excepted, is tolerable. The software is crap.
https://en.m.wikipedia.org/wiki/Samsung_Knox
Yes, per Wikipedia:
"All Nexus devices feature an unlockable bootloader[3] to allow further development and end-user modification.[4]"
https://en.m.wikipedia.org/wiki/Google_Nexus
http://android-developers.blogspot.com/2010/12/its-not-rooti...
http://www.engadget.com/2011/11/03/galaxy-nexus-gets-rooted-...
Note that this still does you comparatively little good if you don't have CyanogenMod to put on it. Mind, I didn't say "no good", I said "little good".
It's the CyanogenMod dev team at Cyanogen Inc who've been impacted by the layoffs. There's still the independent effort, but it's lost a major employer and investor.
Even if CyanogenMod suffers, there's plenty of other open-source Android versions out there. Unfortunately finding a clear, unbiased comparison of them is even harder than it is for different Linux distros. :(
I got there in the end, and I'm pretty sure most newer phones can be flashed in far easier ways.
However is it intrinsically hard? Could you imagine a software tool you would run which would ask a few questions, automate as much as possible over the OTG port, and guide the user for the rest?
That's the kind of tool Cyanogen Inc could write. Of course monetizing it is another thing.
http://forum.xda-developers.com/showthread.php?t=2391493
I've been wanting to do something like that for a long time, get away from the recoveries and politics of the different factions. Make something a user can just pick up an old device, run an install process, and have a fast new OS.
Google of course doesn't make it easy, mostly due to carrier and manufacturer demands. But also due to the lack of a defined ABI between the user interactive parts and the libraries for talking to hardware.
There's really no reason an operating environment couldn't be compiled for ARM and run on the majority of devices if this specification was enforced. I'm looking to gonk and a new Google initiative as a possible solution.
These are not random examples - they ALL happened with the phone I am now typing on.
Its also not something a business could be built on top of, which is the original topic of this thread. Users expect single click installs.
The mobile phone market is massive. Even if you can capture a small subset of the users you can make money. Not everyone needs a one step install process.
Does a new kernel version work? (Android will require it). Will the device drivers work on a new kernel version? Will driver blobs be compatible?
Chipset manufacturers will definitely not send you a new version of a driver if you're not the mobile manufacturer (and sometimes, not even if you are). You're on your own, updating the driver, or writing a shim that acts as a proxy between the driver, kernel, Android, etc. It's just not feasible.
It would be recommended to charge for Cyanogenmod instead of giving away the operating system for free.
Further it would be recommended from a revenue point of view to support the Android devices with the biggest user base. Samsung, Huawei, Xiaomi. These devices should be super easy to upgrade.
Smartphone market share http://www.idc.com/prodserv/smartphone-market-share.jsp
It's like startups are looking to one-up each other in terms of "crappy ways to treat employees you're letting go."
That was what I was talking about. If I can get fired from one day to the next a company would at least have to pay me a hefty premium as I would be not better of then say a freelancer.
1) They don't actually support any devices. All devices, except the ones they sell (like that Chinese one they struggled for over one year to release, the OnePlus) are community supported.
2) Their own stuff (everything they put on top of AOSP) is of a very very bad quality, code-wise and UX-wise.
3) They are highly irresponsible. They have a Jira bug tracker with triagers that don't know anything about coding or Android at all. They have a Gerrit code review system where they antagonise all external contributors, nitpicking on everything, while Cyanogen, Inc. employees push everything they want into the tree without any code review, breaking stuff in the process (something that happens very often).
Maybe installing images could be supported through a self-signed key and fastboot.
Is there anything device-wide root is needed for on modern Android?
I really wish they had not chose "Cyanogen" as the company name.
This is a massive blow to the Android ecosystem and one that would move them closer to that of iOS. AOSP doesn't stand up to proprietary Android.
Cyanogen was the best lifeline for old hardware.
The massive fiasco with Oneplus (which by the way is going strong) or the short-lived partnership with Micromax-Yu which was terminated for reasons unknown. I think they got greedy too fast without building loyal customers. At some level they believed that customers would have no real choice for older devices .
What instead happened was that Xiaomi and Oneplus completely destroyed incentives to hold on to older phones for any reasonable amount of time. And Cyanogen had already burned those bridges.
I am willing to bet CM will either build their own phone (the cyanogen "Nexus") or get acquired by Google or someone - who anyway are looking for some kind of replacement for the Android One/Silver program.
I bet all those who are being let go are partner services.
Here is a good article: http://www.xda-developers.com/corporate-explained-whos-cyano...
Long story short, I'm working with a Chinese company on my phone product. I have a prototype sitting on my desk that's been produced on a skeleton budget. It's better than anything I've seen Cyanogen ship.
If any ex-Cyanogen engineers are interested in a chat, hit up the email in my profile. :)