Two mobile operating systems currently comprise nearly 100% of the global smartphone market. Building one is hard, and for most phone makers, there are
The problem is not building a new operating system, but app support. We have and used to have a lot alternatives, but because mainstream apps are missing no one wants to use them. They don't have enough users, so developers won't develop for them, egg-chicken problem. Everyone tried to solve this by android compatibility layer, but android apps will always run better on android...
I use microG since years at this point, and while most things are working, I always find some quirks, and some random apps not behaving as they should. I'm fine with that, but a non-tech guy would freak out from that. And it's not even a completely different os, only an alternative implementation of GMS aka Play Services.
See previous and current examples, all of them was/is a good or at least usable as an os, but if you can't use your bank's app or whatever app you need in your daily life, you won't switch to it. Even M$ couldn't solve this problem, why Mr. Pei could solve it.
Edit: Obviously in the article they don't speak about an actual OS, but one more Android skin... So Mr. Pei is not planning to solve this, they are just redefining the meaning of words, Android skins are called "OS"s in entrepreneur speak nowadays.
Mozilla had the right idea with FirefoxOS - doing everything as HTML. All the apps people give a shit about are glorified browser windows anyway. What Mozilla fucked up was aiming for a low-end market... with an OS defined by constant power / performance overhead.
Every application in any modern phone could be transpiled HTML5 and you would not know the difference. If it's not "one o' them Genshin Impacts" then it probably already is.
“If you think about the tech stack for what an OS is, I don’t think we need to work on the lower parts of the stack — drivers and how hardware connects to software and the kernel,” Pei added. “I don’t think we need to work on that, but we should work on innovating the user experience, because operating systems haven’t really changed for 40 years.
Yeah, they speaking about another freaking skin...
Migrating from Android to a possible "NothingOS" has the same energy as migrating from Twitter to BlueSky:
You're escaping from one techbro's grubby hands just to fall into another's.
He's gonna do the same shady shit eventually. It's the way all these companies operate, if don't go FOSS you ain't escaping anythin'
He's already done shady shit, like build his company as a remote work startup during the pandemic, then pulled the rug out from under his employees by mandating full time in office, adding that people who try to work from home are childish.
If you think about the tech stack for what an OS is, I don’t think we need to work on the lower parts of the stack [...] but we should work on innovating the user experience, because operating systems haven’t really changed for 40 years. [...] these devices have so much information on us [...] but they don’t leverage any of that information to make the experience any better.
Now this, I wholeheartedly agree.
I feel there's a looooot of room to try new, innovative, potentially even wacky user UIs and experiences, Sailfish OS and Nintype/Minuum being my favorite examples of a existing concept being done with completely fresh UX, HOWEVER
Not having to work on the lower part of the stack??? What???
If we're talking about x86, sure I guess, throw a Linux kernel on it and build whatever on top but ARM, specially on phones, looks like a compete hellscape! (to a outsider like me)
Ask any Linux phone Dev how easy it was to get the hardware going. Their heads will likely start spinning.
Could you build a new, modern mobile entirely from components with Mainline Linux support? And did this support come from the manufacturer or was it hundreds of hours of painful reverse engineering from the community?
New phone without having to work on the lower parts of the stack??? Dunno bout that man.
This is the same guy who started a remote company then tried to pull a "return to office" move on his employees without much notice. I highly doubt they could build their own OS given their poor management record