This seems to run a VM rather than natively supporting linux syscalls. So termux will be more performant with its natively compiled binaries than this solution.
I've run Linux command line tools and GUI tools using Termux and a random X11 app. . It already works, just not with stuff requiring root permissions (no wireshark etc.). Apps like https://play.google.com/store/apps/details?id=tech.ula do it through VNC instead.
What I found to be the most limiting factor isn't the OS, but the tiny screen and the lack of good input methods.
Funnily enough, Samsung ran trials where their Dex mode provided a full Linux GUI when plugged into a monitor/USB dock, but that got killed off. Now Dex is back to being an Android desktop environment again. On the plus side, Dex now also works on tablets without being docked, so I guess that was worth it for me.
I don't think that'd be very useful on an ARM based android device. But they could release an ARM version of steam eventually.
Android devices with decent x86 emulation being released would be the ideal way to have that setup. Hopefully Qualcomm doesn't gate that feature to windows laptops.
If you want to run Windows games on Android, Winlator already does a lot of work for you (packaging Wine+Box86/Box64). My phone isn't exactly a powerhouse so the most impressive game I've been able to play with has been Oblivion so far.
What doesn't help is that the ARM GPUs in most phones don't work like the GPUs on desktop. Their I/O model is different, so you can't just take desktop shaders, transpile them, and run them on mobile without losing a boatload of performance. Maybe the desktop-ish chips will work for that, but I wouldn't hold my breath at the moment.
Well there is a steam client for Android already. They could extend that to an app store and sell mobile games there too, maybe even getting developers of pc games to port their games to mobile platforms?
I would advise against this. I have enough with google monopoly in the browser and mobile areas to add another one in the desktop market with this, plus android is not a privacy/safe environment to use. I will probably run android apps if necessary in my linux box, not the oposite.
Of course you use applications on some systems also according to the trust that you have towards said systems. There are still applications that could be nice to have running on Android - it is a possibility.
Plus: you know, you could use Android on an airgapped device... Many of mine are.
Apart from development, one of the main reasons someone would want to have linux on android is to work with data on your mobile like sms, camera images or interface with apps using popular or custom linux apps.
Given the way Google has been restricting the utility of termux under the guise of security for the past few Android releases, I think they will keep the access to main os and it's data as minimal as possible.
It is great if it is, but I don't think it will be a substitute for termux and its api any time soon.
It's still better than nothing and a good thing for development purpose.
This is bringing the ChromeOS version of containers which are completely isolated from the outside OS - they have their own mounted storaage. You can mount in some folders to share data, but that's it. You DO have sudo though.
The example of running android studio is a great one if that becomes easier than it currently is. (apparently you can do it via termux, if you replace the java runtime in the download)
Develop/deploy on the same device would make getting into android dev much easier. Also, you have your dev tools with you wherever you are, in case you run into a bug or something while testing your app.
The last time I looked at it (several years ago), it had become a subscription-locked money grab.
It was listed as a free app in the Play Store, but when you ran it, you could not do anything without buying a subscription. The reviews for it said it was full of dark patterns and shady stuff.
Some developer blogs described how good and capable it was, but it seems that they were written before the money grab changes.
Despite all of the rumors of its demise its continued to have a large number of daily commits to it the entire year so far. It’s clearly not “dead” I think medium to long term it still becomes the uber unified cross device operating system for them.
Chrome is taking a dependency on Androids underlying networking stack for now and picking up the ability to essentially run it in a VM and Android itself is also picking up the ability to run Fuchsia in a VM.
Fuchsia is picking up the ability to have full Linux compatibility as well via their Starnix efforts.
All of the pieces are starting to come together to make that more and more seamless until presumably at some point they are ready to start calling them the same thing fundamentally.
> Will be a disaster to make so many machines obsolete.
Not at all. Most Chromeboxes/books can be unlocked to install whichever OS the user wishes. I have like a half dozen of them happily running Debian, Manjaro and Alpine after unlocking and reflashing them with the firmware (Coreboot) at https://mrchromebox.tech/
They're really good quality hardware, and once given a proper OS they become very reliable mini PCs.
That would be a killer feature for me; it's already close with termux etc, but that's not good enough for some things. If it could run Docker as well... Can it on chromeos? Never used Chromeos.
I think in ChromeOS you can run Linux (sort of like WSL2 in Windows). I don't think one can run Linux apps directly in ChromeOS. One can run Android apps directly in ChromeOS though.
Yep, I was looking at the Liberty and also the tablet; I don't really care about the pricing (which is pretty mental, but for the good cause), however the specs are, well, not great for a linux machine that can actually do something. Especially the memory. It would be my daily driver (plugin to a monitor or xreal glasses) if it had 12 gb, but 4gb... that's not really a thing. I use my s24 for mostly everything and would love to move to something like this as I don't really like samsung (or android) that much, but let's up the memory eh.
Like everywhere, memory is still what holds things back; I would pay $3k for the Liberty if it had 12-24gb, but he.
I agree with all of it, but the reality is that I need more memory and I understand that not many people (well, the author of the article you sent for one) are looking for this convergence. But I am. Especially with AI , it is becoming less and less a requirement for me to carry a laptop: it will be a combination of talking and typing to do whatever I need to do. With native linux I can already do that, without, it quite a pain for on phones. Or at least too much work for something that only I am going to use. But we will get there; guess the next one will be better.
In Linux it's trivial to have long-running processes. In Android it's the opposite.
So I wonder if I create a Python script which runs 24/7 and collects the sensor data, if it will get killed at will by the OS or if it just runs the 24/7 like it would on Linux.
So more recent versions of Android have fixed the issue of killing background processes (Android 12), or Termux has found a programmatic way around it? Good to know.
My normal usecases are ssh sessions that run for weeks (months?) at a time and I run backups that takes up to an hour with rsync and have never had any issues. Not saying that there aren't usecases that will run into issues, I've just not run into it. Can't see how a python script running 24/7 collecting sensor data would have any issues.
And this has worked fine for me in every single android version since forever (long before v.12 and every version after) (though there have been some teething issues right after a new version is released, most notably permissions on storage. But last versions have been without any hassle that I can remember). There apparently are issues with more than 32 phantom processes or excessive CPU and some options to turn it off but it is not something I've ever had to do.
Maybe a problem with people trying to run a full featured linux distro with GUI support etc? Cool but never had a need. Termux environment is rich and powerful as it is.
Termux can definitely run for a long time even on Android 12 I think. I tried to put a web server to a Termux running on Chromecast (Android 12), and it is still running after months.
It's a trap. We can already run almost all Linux apps via other solutions like termux, qpython and others. They want to kill all of it, but they can't just do it without a pr hit.
So they just "pretend" to bring in crappy Linux support while removing what people actually use. You want terminal apps on that Linux? (99% of what I use) You need to connect a USB keyboard. Oh, you want X11 apps? They need to ve rebuilt with our "library" (that will probably require significant redevelopment). And so on.
No, I don't trust Google neither other monopolies.
A "crippled" (in capabilities) version of Termux is on GooglePlay - that which targets the latest SDK instead of the last before forbidding execution of binaries outside the APK.
Android dropped a lot of native APIs, though, so it may take a while for the Play Store version to be feature complete again (and even then they will still need to continuously convince their AI reviewers that they do in fact need all of those permissions).
Google decided that it's not acceptable to execute downloaded binaries, which is a fair requirement for most non-malicious apps, but a program for Termux. Termux can work around it, but I'm fully expecting them to get kicked out of the Play Store not long after their work is complete for circumventing these protections.
> they refuse to acknowledge the way is via Java user space
I don't think that has much to do with it. Google's problem with Termux is that it downloads and executes native code. Also, you're phrasing that like refusing to acknowledge a megacorporation's version of the truth is a bad thing.
You're not wrong, if Google adds proper input methods support to Wayland (like on ChromeOS) apps will have to adapt. Or stick to the main toolkits (GTK and qt6) where Google provides IME modules.
IME on Wayland sucks right now outside of ChromeOS.
I just put Ubuntu 24.10 onto a box yesterday where Wayland is the new default and so far I must say everything has been smooth sailing including a few games on Steam that I tried so far.
>No, I don't trust Google neither other monopolies.
If it's any conciliation neither does the US government. We are in for a ma-bel shakeup of the market. You know, the one that let internet running to your house happen.
Is this really still true? This isn’t meant to be a defence of Google by any means, but to me it seems Microsoft has adopted a lot of pervasive practices in recent years. To the point where they would actively add things like Minecraft installers in my personal windows machine with whatever tiny version of windows comes with a prebuilt PC used like a game console.
I guess it’s not privacy invasive to add a Minecraft commercial directly into my OS, but they seem to be doing more and more of that stuff as well through their “telemetry”. Though I suppose much of it is targeting enterprise rather than our personal privacy.
MS are doing shitty stuff but they are hardly the monopoly they were in the 90s. You can ignore and avoid Microsoft completely if you want these days. The same is harder for Google - most people at least will want to watch something on YouTube at some point.
This is going to depend on the market. If you consider enterprise I'd argue Microsoft are very much still a monopoly. A client of mine is in the process of removing their existing telephony solution, which has many more features than teams, for teams, because teams is bundled with the necessary Microsoft 365 to get all the Azure controls needed for regulatory compliance.
Most enterprise organisations killed off their different chat clients and phones and switched to teams in my area of the world. Not necessarily tech companies, but everyone else didn’t want the additional cost when Teams came “for free”.
Teams phone system is a significant additional licensing expense, it's definitely not free.
It's also not standards compliant, despite speaking a malformed version of SIP, so you can't just plug it into standard SIP Infrastructure and have it send and receive invites for calls, or send and receive SIP simple messages for texting.
Crowdstrike outage has shown that isn't even remotely true in comparison - Microsofts failure took out a lot of critical infrastructure. Killing Google search wouldn't come close.
Microsoft made contracts with its clients to automatically create M$ accts, for example, my .edu email acct was self hosted years ago, then magically I had a tightly integrated (oauth at first bla bla) M$ acct. Since I didnt ask for it, and I dont want it, I couldnt care less of I run afoul of them. Heck, please revoke it, atleast that would break their contract. AFICT my credentials for journals etc still rest with my .edu (its a problem that credentials are necessary at all). Skynet on the other hand is hell bent on using its monopoly to force me to make an acct:
>Prior to its breakup in 1984, Bell System's legal monopoly over telephony in the United States allowed the company to impose strict rules on how consumers could access their network. Customers were prohibited from connecting equipment not made or sold by Bell to the network. The same set-up was operative in nearly all countries, where the telephone companies were nationally owned. In many households, telephones were hard-wired to wall terminals before connectors like RJ11 and BS 6312 became standardized.
You make it sound a little like this isn't timing and ma bell would have developed securer edge nodes, but everything on their network sophisticated enough to attack was powned by bored children.
Sure, phreakers could attack the central office equipment, and get free calls, disrupt the network some, but they couldn't make some other customer's phone explode, or melt into slag, or ring 1,000,000 numbers at once, from 5,000 miles away.
I would imagine the break up of AT&T and the monopoly that formed around phone connections certainly inspired Americans working on the early algorithms of the internet. We still use telephone numbers today, even though mobile internet has made them fully obsolete
You would imagine wrong. The internet had settled just about all the major protocol decision by 1984 it was all about the hardware you could connect to the network.
I'm wondering if any of the antitrust cases will be pursued as vigorously in the next administration. Neither presidential candidate seems to be a fan of Lina Khan. She has done excellent work in bringing antitrust enforcement back into attention.
Corporations, however, hate antitrust law with a passion and unfortunately they now decide a lot of policy via their financial muscle thanks to Citizens United.
I see this kind of sentiment on HN a lot and it's weird to me. Like, what's the issue with discussing on a hacker forum ways that Google is making Android worse for hackers? Especially considering the alternative is iOS and it's much worse in that regard.
it's not paranoia when they're out to get you. these companies have made it clear time and time again, control and profits are the only things they care about - not the effort that you put in to produce something useful to others. they will break it at every opportunity to make a quick buck.
I'm inclined to suggest walking directly into their trap, gradually developing all the missing apps of a mobile-linux phone and then introducing a wine-like layer that emulates their trap on Linux mobile..
No, there underneath is a migration plan being implemented...legacy Android will run sand boxed Linux apps to match Fuchsia offerings....first OEM Fuchsia device is 2028.
Obligatory joke that the year of the linux desktop may finally come as the android desktop.
I wonder how usable/peformant such as setup would be. But conceivably it could incentivise more native linux app development along the lines of KDE/Qt etc. if there is larger distribution potential.
I think we're seeing another big step on the path to the deprecation of ChromeOS.
Back in June this year, they were "embracing portions of the Android stack, like the Android Linux kernel and Android frameworks, as part of the foundation of ChromeOS"[0]. Now they're enabling a Linux VM on Android, which is the inverse of what they did on ChromeOS under the ArcVM moniker, i.e. running Android in a VM[1].
I don’t actually read it this way for what it’s worth. I agree with you that they are unifying in someways for sure but I don’t read it as the death of their desktop OS for even a moment.
The thing is: Android is fundamentally Linux. Why should we have to run Linux in a VM (on underpowered hardware). Why not just open up Android, and bring it closer to its roots?
They could start by getting rid of the need to root phones, to have full access to them: entering developer mode could give you access to something like sudo.
Security. Many people want to run critical applications on their handhelds, and the policy is to err on the side of safety.
(In "better safe than sorry", that is a different kind of sorry - I spoke only yesterday with a user that got locked out of the device completely because of a small accident.)
More like: Many people want to run their applications on other people's phones, without those other people having any say about what that applications do or don't.
It is security, but security of the app publishers against the user.
I guess you mean: a non rooted Android, conversely, inhibits the removal of bloatware and spyware. Yes, a proper solution covering all needs is needed.
(Personally, I do not use said «critical applications» because I do not trust container, application and providers. So...)
This - enterprise app creators (more concretely, their security auditors) outright demand that the users are locked out of their devices and that the device confirms and enforces that before their app is installed. The amount of these apps and these processes is increasing and has now made it into governmental rules as well (e.g. new digital IDs).
They'll never go that route because Google wants full control on your device. But yeah fundamentally, there's no technical reason why Android couldn't be a normal Linux distribution.
We are back to square one and what I said originally yes, they don't want to do it because they want full control on your device.
There's no technical reason why Android can't just be another Linux distribution. There's no technical reason why you could not boot into an OS from a usb key on your phone like a normal computer. There's no technical reason why the play services have more rights than the actual owner of the phone. It's all about who controls your device.
Android follows a "look, don't touch" model and will continue to be this way as long as possible on Google's side.
It's really a distro isn't it? Linux kernel packaged with a set of userspace software (mostly in java) and bunch of non-java supporting bits making it work.
A custom Linux kernel, compiled with clang, where Rust is used without drama, and most drivers live on their own process talking with the Linux kernel via Binder IPC, microkernel style.
Which means most Android drivers are actually written in a mix of C++, Java and Rust, with a standard ABI, HAL.
> The code change, which hasn’t been merged yet, removes the entire Ferrochrome launcher app from AOSP. Google’s reason for removing this app is that it doesn’t plan to ship it or maintain its code. It seems that Google is shifting towards using the Linux-based Debian distro instead of Chrome OS as its testbed for AVF development.
Well, not really - not more -, because we have mobile devices that are born with Android and installing a Desktop Linux on them would be a hassle of doubtful results, and we have Desktop Linux applications that are missed in the Android software availability...
Both ways are sought: running Linux applications on Android and running Android applications on Linux and elsewhere.
It sounds like this would make Android devices more useful for developers, which is a good thing. Although I am pretty happy with building what I need via web apps these days.
The one thing I am missing on Android is that Chrome on Android does not yet support the File System Access API. So you can't make web applications which let the user store their data locally:
I'm seeing a lot of speculation about a shadowy hidden agenda in these comments, but based on my time at Google I'd say it's very possible there is no overall strategy here. The only prereq is that someone wanted to do this and was able to get buy in from their director. Would it make sense for Google to not maintain two separate OSes that both strive to support the other OS's software? Sure, maybe it would. But that hasn't stopped Google before from having multiple products that do the same thing for no good reason. And there are likely a lot of folks who have built little fiefdoms (or at least domain expertise) within ChromeOS who want to see that investment maintained.
And remember, there's also fuchsia, supposedly a future replacement OS but who knows what that's going to turn into if anything.
To folks who think this is a way to "kill Linux support" on Android, I don't know what to tell you. Almost nobody uses those features and they're not a threat to Google; I can't imagine anyone is thinking "how can we invest developer time to get rid of this?" and being taken seriously.
...at the moment - as soon as we can build full mobile applications through something different than the SDK/NDK, development could surge, and the software supply with mobile environments in mind could also significantly pass through this alternative way.
I actually really appreciate this, and if it’s like the ChromeOS version I’d be really happy with it. Ironically it’s all I’d need to make the iPad a laptop replacement for me (which is why I don’t ever see it happening in iOS. It might not even be technically possible there without partitioning the limited memory due to needing a full VM, which isn’t an issue on Android).
It’s a nice solution. A sandboxed container in which you can run whatever you like in isolation from the host device with proper window forwarding. Combined with a bit more polishing on Android desktop mode this absolutely becomes a desktop replacement for most people, especially in corporate environments, so there’s a potentially large chunk of the Windows market that could be shaved off here.
I work on longer contracts and would absolutely buy a Pixel-per-job if it gets me a full desktop and phone in one small package that I can keep isolated from other work.
Old comment of mine in another thread:
Our phones might be our next Desktops/Laptops/main personal computing device: 1. local first (you do go to the bathroom/gas station with your phone) 2. portable 3. reduce ewaste, money spent
The lack of convenience in the form of larger screen might be mitigated using smart glasses, projector(unihertz tank 3 has built in), or just connect to an external monitor
Snapdragon 8 gen 3 performs like cpu from 2020 and midtier gpu from 2016 AVF might ship with android 15 as Mainline module (One need kernel 5.10+)
Windows on arm is progressing (qcom snapdragon elite chipset on microsoft surface devices), qcom gunyah hypervisor, MediaTek's GenieZone hypervisor, winlator etc..
I hope that will allow for more "usability" for ipads. (yes, for ipad + keyboard you might as well just have macbook air, but maybe at least let's have that option)
The only reason why iPads aren't just as usable as laptops is entirely on Apple's decision to limit it to iPadOS/iOS
If they'd just let you transition to regular macOS after connecting a screen/mode/keyboard, it'd be just as powerful as a MacBook. They already use the same CPUs and have just as much RAM
I haven't seen any updates than those first screenshots, almost making me wonder if this is real. I do a search every few months to see if someone created a tool that allows anyone to run Windows/Linux on Android. Nope. Because I really want to do something really meaningful on my Galaxy Tab other than just watching YouTube or web browsing.
If you're talking about kdrag0n's work he had it in his pateron and I think Mishaal Rahman had tried it. Hopefuly he'll pop up in this thread and clarify what happened to it
I’m not really keen on the terminal being a web view. Yes, there are lots of decently performing web terminals, etc. But to me it seems like an engineering failure to have to render a terminal like that.
199 comments
[ 2.7 ms ] story [ 240 ms ] threadWhat I found to be the most limiting factor isn't the OS, but the tiny screen and the lack of good input methods.
Funnily enough, Samsung ran trials where their Dex mode provided a full Linux GUI when plugged into a monitor/USB dock, but that got killed off. Now Dex is back to being an Android desktop environment again. On the plus side, Dex now also works on tablets without being docked, so I guess that was worth it for me.
Hopefully the next Pixel Tablet includes an optional wired keyboard.
iPad Pro had VMs in 2022, sadly removed by Apple.
https://apps.apple.com/gb/app/ish-shell/id1436902243
I don’t think you can run anything requiring a GUI
Android devices with decent x86 emulation being released would be the ideal way to have that setup. Hopefully Qualcomm doesn't gate that feature to windows laptops.
What doesn't help is that the ARM GPUs in most phones don't work like the GPUs on desktop. Their I/O model is different, so you can't just take desktop shaders, transpile them, and run them on mobile without losing a boatload of performance. Maybe the desktop-ish chips will work for that, but I wouldn't hold my breath at the moment.
Plus: you know, you could use Android on an airgapped device... Many of mine are.
Given the way Google has been restricting the utility of termux under the guise of security for the past few Android releases, I think they will keep the access to main os and it's data as minimal as possible.
It is great if it is, but I don't think it will be a substitute for termux and its api any time soon.
It's still better than nothing and a good thing for development purpose.
Edit: typo
... in the Linux virtual machine.
I bet it's going to be the same here.
Develop/deploy on the same device would make getting into android dev much easier. Also, you have your dev tools with you wherever you are, in case you run into a bug or something while testing your app.
https://www.android-ide.com/
It was listed as a free app in the Play Store, but when you ran it, you could not do anything without buying a subscription. The reviews for it said it was full of dark patterns and shady stuff.
Some developer blogs described how good and capable it was, but it seems that they were written before the money grab changes.
Given the disruption of LLM everywhere, and in Google specifically, I wouldn't be surprised if the headcount was reduced in these teams as well.
Chrome is taking a dependency on Androids underlying networking stack for now and picking up the ability to essentially run it in a VM and Android itself is also picking up the ability to run Fuchsia in a VM.
Fuchsia is picking up the ability to have full Linux compatibility as well via their Starnix efforts.
All of the pieces are starting to come together to make that more and more seamless until presumably at some point they are ready to start calling them the same thing fundamentally.
Sources
https://www.androidauthority.com/microfuchsia-on-android-345...
https://fuchsia.dev/fuchsia-src/concepts/components/v2/starn...
https://blog.chromium.org/2024/06/building-faster-smarter-ch...
I hope though, that they "backport" this new core to existing devices too. Will be a disaster to make so many machines obsolete.
And to ChromeOS Flex too. Definitely.
Not at all. Most Chromeboxes/books can be unlocked to install whichever OS the user wishes. I have like a half dozen of them happily running Debian, Manjaro and Alpine after unlocking and reflashing them with the firmware (Coreboot) at https://mrchromebox.tech/ They're really good quality hardware, and once given a proper OS they become very reliable mini PCs.
Yes, Android apps in general work fine, apart from those relying on the DRM.
Like everywhere, memory is still what holds things back; I would pay $3k for the Liberty if it had 12-24gb, but he.
Even with such low amount of RAM you can use it as desktop though: https://puri.sm/posts/my-first-year-of-librem-5-convergence/
See also: https://puri.sm/posts/the-danger-of-focusing-on-specs/
So I wonder if I create a Python script which runs 24/7 and collects the sensor data, if it will get killed at will by the OS or if it just runs the 24/7 like it would on Linux.
https://wiki.termux.com/wiki/FAQ#Will_Termux_work_on_Android...
And this has worked fine for me in every single android version since forever (long before v.12 and every version after) (though there have been some teething issues right after a new version is released, most notably permissions on storage. But last versions have been without any hassle that I can remember). There apparently are issues with more than 32 phantom processes or excessive CPU and some options to turn it off but it is not something I've ever had to do.
https://github.com/agnostic-apollo/Android-Docs/blob/master/...
https://github.com/agnostic-apollo/Android-Docs/blob/master/...
Maybe a problem with people trying to run a full featured linux distro with GUI support etc? Cool but never had a need. Termux environment is rich and powerful as it is.
So they just "pretend" to bring in crappy Linux support while removing what people actually use. You want terminal apps on that Linux? (99% of what I use) You need to connect a USB keyboard. Oh, you want X11 apps? They need to ve rebuilt with our "library" (that will probably require significant redevelopment). And so on.
No, I don't trust Google neither other monopolies.
See https://github.com/termux-play-store/
By the way, I'm using the Play store version and don't have any issues but all I do for now is read man pages and use ssh.
Android dropped a lot of native APIs, though, so it may take a while for the Play Store version to be feature complete again (and even then they will still need to continuously convince their AI reviewers that they do in fact need all of those permissions).
Google decided that it's not acceptable to execute downloaded binaries, which is a fair requirement for most non-malicious apps, but a program for Termux. Termux can work around it, but I'm fully expecting them to get kicked out of the Play Store not long after their work is complete for circumventing these protections.
I don't think that has much to do with it. Google's problem with Termux is that it downloads and executes native code. Also, you're phrasing that like refusing to acknowledge a megacorporation's version of the truth is a bad thing.
IME on Wayland sucks right now outside of ChromeOS.
If it's any conciliation neither does the US government. We are in for a ma-bel shakeup of the market. You know, the one that let internet running to your house happen.
I guess it’s not privacy invasive to add a Minecraft commercial directly into my OS, but they seem to be doing more and more of that stuff as well through their “telemetry”. Though I suppose much of it is targeting enterprise rather than our personal privacy.
It's also not standards compliant, despite speaking a malformed version of SIP, so you can't just plug it into standard SIP Infrastructure and have it send and receive invites for calls, or send and receive SIP simple messages for texting.
I've never looked at the cost, but I was told switching to teams is being done to save money. So perhaps not as much money as hoped...
- LinkedIn
- GitHub
- O365 dominance is not extreme, but present.
https://news.ycombinator.com/item?id=41773781
>Prior to its breakup in 1984, Bell System's legal monopoly over telephony in the United States allowed the company to impose strict rules on how consumers could access their network. Customers were prohibited from connecting equipment not made or sold by Bell to the network. The same set-up was operative in nearly all countries, where the telephone companies were nationally owned. In many households, telephones were hard-wired to wall terminals before connectors like RJ11 and BS 6312 became standardized.
Corporations, however, hate antitrust law with a passion and unfortunately they now decide a lot of policy via their financial muscle thanks to Citizens United.
Gogle can do whatever they want to qpython and termux, 99% of the population will not care at all
Doesn't termux already require all its packages to be compiled against android NDKs?
with Rooting.
"Docker on Android "
https://gist.github.com/FreddieOliveira/efe850df7ff3951cb62d...
I wonder how usable/peformant such as setup would be. But conceivably it could incentivise more native linux app development along the lines of KDE/Qt etc. if there is larger distribution potential.
Back in June this year, they were "embracing portions of the Android stack, like the Android Linux kernel and Android frameworks, as part of the foundation of ChromeOS"[0]. Now they're enabling a Linux VM on Android, which is the inverse of what they did on ChromeOS under the ArcVM moniker, i.e. running Android in a VM[1].
[0]:https://chromeos.dev/en/posts/building-a-faster-smarter-chro...
[1]:https://9to5google.com/2019/05/24/chrome-os-android-apps-vm/
They could start by getting rid of the need to root phones, to have full access to them: entering developer mode could give you access to something like sudo.
Security. Many people want to run critical applications on their handhelds, and the policy is to err on the side of safety.
(In "better safe than sorry", that is a different kind of sorry - I spoke only yesterday with a user that got locked out of the device completely because of a small accident.)
(Personally, I do not use said «critical applications» because I do not trust container, application and providers. So...)
All the more reason to find ways to break those restrictions.
Surely running a digital ID app on something like e/os is possible? If not there should be a lawsuit.
There's no technical reason why Android can't just be another Linux distribution. There's no technical reason why you could not boot into an OS from a usb key on your phone like a normal computer. There's no technical reason why the play services have more rights than the actual owner of the phone. It's all about who controls your device.
Android follows a "look, don't touch" model and will continue to be this way as long as possible on Google's side.
I thought that Android is mostly Java (that happens to run on a Linux kernel but might be anything as far as the specs are concerned).
Which means most Android drivers are actually written in a mix of C++, Java and Rust, with a standard ABI, HAL.
https://www.androidauthority.com/chrome-os-android-launcher-...
> The code change, which hasn’t been merged yet, removes the entire Ferrochrome launcher app from AOSP. Google’s reason for removing this app is that it doesn’t plan to ship it or maintain its code. It seems that Google is shifting towards using the Linux-based Debian distro instead of Chrome OS as its testbed for AVF development.
Well, not really - not more -, because we have mobile devices that are born with Android and installing a Desktop Linux on them would be a hassle of doubtful results, and we have Desktop Linux applications that are missed in the Android software availability...
Both ways are sought: running Linux applications on Android and running Android applications on Linux and elsewhere.
The one thing I am missing on Android is that Chrome on Android does not yet support the File System Access API. So you can't make web applications which let the user store their data locally:
https://caniuse.com/mdn-api_window_showopenfilepicker
It looks like they are working on it:
https://issues.chromium.org/issues/40101963
The last commit related to this feature seems to be from from 2 days ago.
I have my fingers crossed that it will be available soon.
And remember, there's also fuchsia, supposedly a future replacement OS but who knows what that's going to turn into if anything.
To folks who think this is a way to "kill Linux support" on Android, I don't know what to tell you. Almost nobody uses those features and they're not a threat to Google; I can't imagine anyone is thinking "how can we invest developer time to get rid of this?" and being taken seriously.
...at the moment - as soon as we can build full mobile applications through something different than the SDK/NDK, development could surge, and the software supply with mobile environments in mind could also significantly pass through this alternative way.
The platform that captures developers in the long term is on the top.
It’s a nice solution. A sandboxed container in which you can run whatever you like in isolation from the host device with proper window forwarding. Combined with a bit more polishing on Android desktop mode this absolutely becomes a desktop replacement for most people, especially in corporate environments, so there’s a potentially large chunk of the Windows market that could be shaved off here.
I work on longer contracts and would absolutely buy a Pixel-per-job if it gets me a full desktop and phone in one small package that I can keep isolated from other work.
https://news.ycombinator.com/item?id=30322035
https://news.ycombinator.com/item?id=30328692
Old comment of mine in another thread: Our phones might be our next Desktops/Laptops/main personal computing device: 1. local first (you do go to the bathroom/gas station with your phone) 2. portable 3. reduce ewaste, money spent
The lack of convenience in the form of larger screen might be mitigated using smart glasses, projector(unihertz tank 3 has built in), or just connect to an external monitor
Snapdragon 8 gen 3 performs like cpu from 2020 and midtier gpu from 2016 AVF might ship with android 15 as Mainline module (One need kernel 5.10+)
Windows on arm is progressing (qcom snapdragon elite chipset on microsoft surface devices), qcom gunyah hypervisor, MediaTek's GenieZone hypervisor, winlator etc..
If they'd just let you transition to regular macOS after connecting a screen/mode/keyboard, it'd be just as powerful as a MacBook. They already use the same CPUs and have just as much RAM
https://github.com/termux/termux-x11
It seems to work very well...
He's just been really busy with university and creating OrbStack that he never got time to make a tutorial on how to replicate his work.
However, other people have figured it out: https://github.com/wasdwasd0105/limbo_tensor
This is already reality with GNU/Linux phones: https://puri.sm/posts/my-first-year-of-librem-5-convergence/