Other than that you’ll get low volume low end devices at high end or higher prices due to lack of economy of scale.
What most consumers actually want and I’ll put in that list also many many people that say they want a “Linux Phone” but in reality will go back to their iPhone/Android after a week is quite different than what a tiny niche is willing and capable of dealing with.
I used Android from day one till the LG Nexus and I switched to an iPhone after having to spend yet another weekend compiling a kernel to update my phone.
Most people want a device that “just works”.
Overall I would be good money on the fact that the percentage of people that say they want to be able to hack their phone freely and will actually do so is very slim because again Android gives you 90% if not more of that freedom already and some devices even can boot other OSes.
Sure things like the blobs on the baseband are still closed but that will always be the case no one realistically is expecting a high end 100% open source phone, and if you are keep dreaming.
And yes while having a phone that just works and one that is open isn’t mutually exclusive in principal but it is in practice especially once you account for economics.
Like it or not securing an open system is harder, and companies don’t like investing time and money developing features hardly anyone will use.
But what I mean is most of the system isn't really open like that. Configuration, application data and stuff like that is mostly managed by the system and you can't access or change that data without major hacks. Termux feels like a second-class citizen because most of the ecosystem isn't built with something like it in mind (in contrast to the "CLI-by-default" experience on other Linux systems). There's also a few (non-embedded) system components like the backup system (which, if I'm not mistaken, is provided by the Google Services Framework) which you can't easily replace.
Of course, when it comes to security this is not something you should give to anyone who doesn't know to protect themselves using it. And I assume it wouldn't be a financial success either, I just guessed this is what people mean when they say they don't like Android because it's too restrictive.
Unlike on the desktop, most important phone apps aren't so large and complicated that they couldn't be ported to or reimplemented on a new platform with a realistic amount of effort. You could offer a significantly better developer experience than either of the dominant platforms today, which would be essential to supporting the apps users expect to find available on any mobile platform today but also potentially attracting some unique and better apps over the longer term.
From the user's side, they'd be genuinely in control of their own device. There could be real security, stability and privacy benefits as a result, and you could do away with a lot of the things that annoy users of current mobile platforms.
As ever, the problem is how to bootstrap a two-sided market. It would probably have to be extremely easy for developers to port their existing Android apps. You might also have to convince one of the major phone manufacturers who can make good hardware at competitive prices to support your platform as an option, or possibly make it easy to install it as a replacement on existing phones. But with the right promotional strategy even these things don't seem totally out of the question. It's a huge potential market, on a scale where one or more well-capitalised big players in the industry could potentially take an interest.
As much as FOSS crowd loves to hate Microsoft, better get the facts right.
https://www.cbc.ca/news/business/stephen-elop-to-get-25-5m-f...
How many major phones can you buy with AOSP installed by default today?
How much development gets done on AOSP that isn't at least heavily influenced by the direction Google takes?
Does AOSP provide comprehensive privacy and security options for users but still freely connect with other devices and services using open standards?
Obviously if you're going to be Linux-based and if we're assuming some degree of compatibility with Android APIs to make porting apps easy then there is going to be common ground with AOSP, but I don't think that makes it the only or necessarily the best option.
Heck how many people do you think install SailFish on their phones? You can buy pretty decent Android phones with AOSP and even SailFishOS builds yet people aren’t taking advantage of that in any particular manner.
If almost no one uses it, even amongst those who claim they want that freedom why will build a business model around it?
I want a mobile OS that gives me complete control over my personal data.
The issue is that whilst people want it there is never a good enough “reason” for that other than I want it, Maemo doesn’t give you more freedom than what AOSP does already at least not on the software level, you can degoogle and Android phone completely and do w/e you want with it. However people don’t seem to be doing that, and those who do often do that for academic purposes rather than to have their own personalized daily driver.
Even more so often the wishes of many people when it comes to customization doesn’t even require a rooted phone and a customized AOSP ROM, if you look at what people customize on their desktop which is often limited to their desktop environment and their workflow can be done on stock Android using alternative launchers and other apps.
Wanting for an Android competitor for the sake of having competition is fine, but it won’t look much different or it won’t be much of a competitor.
The other problem is that you have to choose between hardware quality and customization, because hardware is locked down and there isn't enough economic air left in the room for a third player to be able to invest enough to compete.
As for not looking much different from Android, the key difference would be where the control lies, and what the mindset is around control. You should have a robust code signing and containerization/permissions system, but those should be in the hands of users. With Android every single app installs with permission to portscan every network you connect it to. You will have an app store, but the user will come first, not the ad network.
I liked my Nokia N900 and kept up for a while with attempts to keep its Maemo going after Nokia abandoned it. However, a decade later Maemo has bitrotted and its dev community has dwindled away. Nowadays the Phosh interface (i.e. the Librem phone, or Mobian running on the PinePhone) is seen as the most promising Free Software choice in the long term. Sailfish OS is also actively maintained, but its UI layer is closed source.
App distribution was also better on both platforms. Maemo used apt under the hood, and Preware[1] was phenomenal. These days the two vendors that own 99.4% of the mobile OS market[2] don't want to let you install apps unless they can get a 30% cut.
[1] https://webos-internals.org/wiki/Application:Preware
[2] https://www.statista.com/statistics/266572/market-share-held...
They both allow free apps.
[1]: https://fxtec.com/
[2]: https://browser.geekbench.com/v5/cpu/compare/4573446?baselin...
I have an Pixel2 and was thinking I'd get an iPhone next, but I'm really turned off by Apple's behaviour lately. I also don't want to send more money to Google.
I'm willing to accept that it may not be as polished as the big players, but how would you say it compares to other high-end Android phones?
And if you've tried running linux on it is it usable with reliable call quality?
Taking a serious look