I spent years working on the Ubuntu phone. It was a fully convergent device: by default it would present a phone UI, but plugin an a keyboard and screen and it presents a full desktop UI. The UI was designed by a team of brilliant professional designers and it was a technical success.
Why isn't it around any more?
(1) The carriers. Many people get a free or heavily subsidized phone from their carrier, who makes money from the data charges. Carriers were just not interested in phones from smaller manufacturers, and it's not entirely clear that there is no backflow of cash involved at higher levels. There is also the issue of controlling access to user metadata, always a profitable exercise.
(2) The manufacturers: why would they put a third-party OS on their phone when they control the entire stack from the silicon to the end-user's metadata through their in-house Android fork? They can sell hardware at a negative margin if they can profit from selling access to metadata, or possibly data.
(3) The apps. Developers want to target the biggest, most lucrative markets so they target Android and iPhone which already have a critical mass. You could provide a container to run Android apps on Ubuntu, but then you end up with no differentiating factor except your Android apps look like crap in desktop mode. Running in a container can also restrict the kind of data collection many apps rely on for monetization.
Another problem with apps is that many web services check to see whether the device is iPhone or Android and respond accordingly. Like the browser checks of yore ("you need at least IE 6 to use this site, try updating to a newer browser!") this gives a less satisfactory experience to end users.
The same shit is happening today, except it's Chrome instead of IE6. We have an entire generation of webdevs who learned fuck-all from the previous round of the browser wars.
I do feel if canonical were to attempt another crowd-funding attempt at an Ubuntu phone, they might well be successful this time. That phone was a terrific concept. I feel the market has changed since then, with carriers accepting less control over our phones, and the fad of installing every random app has somewhat distilled itself to a smaller number that is widely used.
> Carriers were just not interested in phones from smaller manufacturers, and it's not entirely clear that there is no backflow of cash involved at higher levels.
I really doubt it’s something sinister in enterprise sized logistics. Mainly because the big guys don’t need to be al 90’ies Microsoft about it in the modern world of leaned up Enterprise. The service the big global brands come with do save the carriers a lot of money, but it’s done perfectly legal.
First there is the logistics. As a smaller carrier, you won’t really have to worry about doing BI on sales projections and what not, because Apple, Samsung and so on will simply tell you how many phones you can expect to sell, which is basically how many they’ve already produced for your region. They’ll handle everything from warehousing to shipping to your local stores pretty much without you having to do anything but sign the extremely “take it or leave it” sort of B2B contract, that also includes the big companies not charging you for products that don’t sell. What this translates to is that you can basically run your company without a logistics department or any BI related to the actual phone hardware. This saves you money both on manpower but also on not having too much inventory.
Then there the support side of things. By carrying big well known brands with a very low degree of user freedom for modifications you cut the non-carrier related support you need to offer to almost nothing. You could probably offer good money to get carriers to sell, even a well known brand like the fairphone, and they’d still turn you down because it would cost them too much money to do so.
I think Apple is already on its way for a convergent device with the Apple silicone. They've already unveiled an iPad with an M1 chip and in the near future we would get iPhones with powerful chips which is able to act as a fully functional MacBook once the phone is connected to an external monitor/TV, of course with a bluetooth keyboard and a mouse.
I bought an ubuntu phone. The phone was nice, but it lacked apps, and the apps were pretty much all web based - so the map app wouldn't even load without an internet connection - pretty useless in a foreign country. And the version that was released to the public didn't have the option to plug in a keyboard and screen for a full desktop.
I don't understand the desire for a Linux phone as so described in the article as a daily driver.
The Android platform represents an enormous amount of work that encompasses a more secure base kernel, an unparalleled selection of applications designed for mobile usage and written in a memory-safe language, fantastic sandboxing and user privacy features leagues ahead of any desktop operating system, and great diversity in the hardware market.
While it's apparent that Android has significant downsides - de facto proprietary drivers, an environment of mostly closed-source applications, some Big G integration, and a general lack of long-term updates occasionally countering the work put into security (thanks, chip manufacturers: https://news.ycombinator.com/item?id=26593274), these are problems that can be solved. It doesn't make sense to throw the baby out with the bathwater.
Closed-source applications are more a moral problem than a security one thanks to Android's sandboxing, and a vibrant ecosystem of high-quality free-and-open-source Android applications (https://f-droid.org - Amaze, Notally, QKSMS, and Tasks are some of my favorites) makes them all but optional. Google's presence in the AOSP codebase is rather limited (mostly around notifications) and taken out altogether by custom ROMs like Lineage and GrapheneOS. Proprietary drivers remain as Android's biggest problem, perhaps alongside the manufacturers that ship locked-down phones with bloatware (looking at you, Samsung).
Why go the desktop-Linux-to-mobile route when you could fork Android, write drivers for the Pinephone / Librem 5, punch a hole to the base system for privileged applications, and have the best of both worlds with an order of magnitude less effort?
I'd be shocked if it took off. The FOSS development paradigm seems to be allergic to good UI design, and this is far more important for people on mobile than desktop.
It's not like the competitors are any better, in the desktop space at least.
The problem in my eyes is plateauing. Once a project moves past the first couple of iterations, it rarely interests the original authors enough to make incremental improvements. And since there is no financial incentive to continue, the project usually cools down to eventually be replaced by the following iteration. This means that you rarely get mature software.
1) The GNU utilities have pretty good CLI UI design and there are plenty of us who are perfectly happy with just that and a decent WM like FVWM even on a phone
2) It's not like closed mobile stuff has a great UI either. My favorite example is Apple's camera app. Did you know you can do manual focus with it? It's absolutely impossible to discover but I think it involves a "reverse three finger pinch." I've even had it explained twice to me (and used it) and forgotten how to do it after. I can never remember how when I need it. This garbage makes the TAR UI look like it was designed by an artist.
> The FOSS development paradigm seems to be allergic to good UI design
I'm having a hard time understanding what it is about the development paradigm that's allergic to good UI design. In good faith I'll assume you meant that "FOSS projects seem to lack good UI design". This I'll agree with, if we're talking about GUI design (others have pointed out that the CLI interfaces are usually great). I really don't think the lacking GUI design is necessarily inherent in FOSS, but I too wonder why more people with design skills aren't attracted to working on FOSS stuff.
Modern UI on the phones makes discovering features really hard. One often ends up with searching obscure forums just to know very useful feature like IPhone trick with holding the space bar to move cursor.
I can understand that on 4” inch phones there were no space for extra buttons, but with 6” screens this became ridiculous. So much for “good” design.
It's a case of developer UI. Has nothing to being "allergic" to "good UI design".
Being able to design something friendly to the average user, yet powerful enough to be friendly to power users, is a very difficult task to accomplish.
Developers understand the latter quite well, I think, as they themselves are power user-types. The former... well, that requires having dedicated designers who know how the average user thinks, as well as perhaps guinea pigs in the form of the average user. And that might require a lot more funding, perhaps.
Openmoko failed for so much more than financial reasons. It's been a while since I've thought of that fiasco, but my memories:
* The leadership was terrible. They had no clue what it took to make a mass market product. They just assumed that they would sell hardware and a community would provide a working operating system and apps magically.
* The hardware was buggy. There was one issue that if you let the battery drain fully, you could not get the phone to recharge it and had to use an external charger. Another issue was that the GPS receiver was accidentally surrounded by metal, so barely functioned. There were all kinds of problems with the radios in the early days. Oh, and that touchscreen -- I guess it was typical of pre-capacative touchscreens, but it was hard to use without a stylus and impossible to hit widgets near the edge of the screen.
* The hardware was massively underpowered (compared to competitors) by the time the Freerunner actually shipped. Weak CPU, little RAM, 2G cellular radio in an era when 3G had become standard, so like 5kbps max data transfer.
*Because of the failure of Openmoko leadership, the community fragmented a hundred ways. This meant that there were a dozen or more "distributions" of an OS for the phone, and none could do more than one or two of the things a typical user wanted in a phone at that time. Then there was finally a big bloated distribution (SHR if memory serves) that packed in enough libraries to make a more or less "usable" device, but doing so maxed out the phone's meager storage and RAM, making it nearly impossible to do anything "smart" with it.
* Once again, in absence of strong UX leadership, the community resorted to dumping X11 apps without modification on the tiny screen. Think impossible to read fonts and dialog boxes that ran off the screen with no scroll capability. The vast majority of devs seemed to only use it by hooking it up to a computer via USB networking and SSHing into it.
As a technical user, I could live with this. Kinda. Sorta. Using it was an exercise in masochism. I was embarassed ... no ... humiliated when a nontechnical person compared their iPhone with the OpenMoko that I had talked up so much (before receiving it).
I had planned to destroy the phone in some fantastic fashion (e.g., melting it with a laser) as soon as I got a real phone. But by the time I could afford an Android, I was so done with it that I just dropped it in the trash (after wiping it, of course).
> They just assumed that they would sell hardware and a community would provide a working operating system and apps magically.
No, not really. They had an exact opposite problem - they made several iterations of the default operating system, starting almost from scratch at each iteration, which burned quite a lot of energy and willpower of the community, which in turn focused their efforts on alternative distros like SHR or QtMoko.
> had to use an external charger.
Fortunately, you could use a standard Nokia BL-5C battery with the Freerunner. You didn't even need to have an external charger, just a charged spare battery would suffice. Also, IIRC this was an issue only with the first batch (so a small minority of produced phones).
> Another issue was that the GPS receiver was accidentally surrounded by metal, so barely functioned
Not really. The GPS problem was because of microSD clock interference. You could solder a resistor on microSD slot pins or use a software workaround that clocked the reader down enough to not interfere.
> it was hard to use without a stylus
I have programmed quite a lot on that touchscreen with OSK without using a stylus. It worked fine, but yeah, it would be much better if the screen weren't recessed (N900 did that well, that touchscreen was excellent).
> so like 5kbps max data transfer.
More like 100kbps.
> Then there was finally a big bloated distribution (SHR if memory serves) that packed in enough libraries to make a more or less "usable" device, but doing so maxed out the phone's meager storage and RAM, making it nearly impossible to do anything "smart" with it.
Uhm, no? SHR was a light OpenEmbedded-based distro with E17-based window manager. It was one of the snappiest and most reliable distros for that device, I used it for a few years as a power user and was very happy with it. And maxing storage wouldn't be an issue anyway since you could boot from an SD card.
Eventually I've switched to a N900 because of the Freerunner's slowness. If Glamo wasn't so slow I guess I would use it for a few years more before switching. I still have it and it still works, although I don't really use it anymore.
I've also had a Freerunner, in theory the ideal pocket computer with Linux on it.
Practically, you could barely have a phone call with all the echo going on and navigation-wise it took _ages_ to find its GPS fix (no AGPS iirc) - the shielded antenna, mentioned down-thread, probably didn't help.
And yeah, whatever you had on screen was probably way too small to be read or interacted with :(
I think i got rid of it on eBay after a few weeks...
GPS receiver being surrounded by metal without some sort of antenna just seems like some sort of crazy bad design skill that even software engineers may know is awful… so, how on earth would this thing have ever shipped at all? Weird to call that an accident?
I enjoy the Pinephone, but it certainly has its issues. I’d kill for better performance; it’s very close to usability. I also have some issues with the hardware (it seems my battery has stopped working entirely, recently.)
On the other hand, there is one confounding software issue that I think makes it the hardest as a full replacement for a modern smartphone: push notifications. An equivalent for APNS or FCM that allows for relatively low power consumption, low latency push notifications would need to be devised, financed and actually adopted. It’s nice to imagine a world where apps on phones are like how apps on desktop are and just keep, for example, active WebSocket connections at all times, but this doesn’t appear to be something practical with current technology.
Maybe federated, E2EE push notifications could come into existence? A sort of ActivityPub of pushes... It’s a pipe dream, but one can hope.
This is kind of anecdotal to Meego, WebOS, FirefoxOS and all the other forms of them that happened in the past.
I don't think that people will use the phone solely because it's Linux. They'll use it if it can solve their tasks at hand.
Personally I think that Linux will never take off if we won't create financial incentives to make and deploy apps, additionally to fix the mess that's user rights/permission management in GTK and QT.
Currently there are no working native alternatives on Linux, libhandy is still a joke for simple tasks like a fading sidebar (or even swipe gestures!) and QT can't be used for anything serious due to their license.
My hope for servo, FirefoxOS and WebOS was that there will be some day an alternative to Electron that's focussed on permissions and sandboxing, and that allows to be externally configured and is more modular on the environment (VM) level. Basically like a settings app on Android that can toggle GPS, toggle Networking access, toggle Camera access etc.
Layouting-wise CSS has won the masses. Everyone that tries to reinvent it for their own opinionated views has failed and given up (including me who spent over 6 years developing an isomorphic App Engine full-time). It's time to let go. The web has won and there's no point in denying it.
React has won most of the developer crowd because of React Native and convenience of staying in the same language and more importantly, being able to reuse the same code architecture.
I think in order for Linux to succeed there has to be a compile pipeline (and runtime?) that allows to deploy those apps easily. If that's not possible (due to whatever reasons) the platform will never really take off.
Privacy and Security and Openness is just no argument for the average user that doesn't care and just wants to get their tasks done.
What is this idea that Qt can't be used? You can license with the community license under LGPL. If Qt can't be used with an open source license, then nothing LGPL can use open source commercially, but the point of LGPL is to not scare away businesses with viral copy-left code. Yes, you have to deliver your object files and a Makefile, et c. Also, you have to open source your Qt library edits, but who needs to do that for business reasons? Are Qt library edits the industrial secret of your business? Probably not. Since when is letting other people run the linking stage on your code tantamount to giving them the keys to your business?
I think you’re right. I don’t understand why more UI toolkits aren’t adopting CSS for layout, especially the newer CSS grid. It’s so much simpler than alternatives like TKinter.
I get the memory/performance issues with Electron. Those can be solved. I’m able to open large XML files in VSCode that simply crush other apps like Notepad and XMLSpy.
> I think in order for Linux to succeed there has to be a compile pipeline (and runtime?) that allows to deploy those apps easily. If that's not possible (due to whatever reasons) the platform will never really take off.
I agree. An important difference to the situation a few years ago is the availability of actually viable cross-platform development frameworks like React Native, Flutter, Capacitor,...
If all I need to do in order to support a new niche OS is configuring one additional build target, I might just go for it. Otherwise, certainly not.
And yet I am using a Linux Phone since 2014 as a daily driver :) And no, it is not for everyone, just like the Linux desktop. It would need a company like Nokia to put its weight behind it to bring it that far.
But if a smaller company like Jolla can do it, it can be sustainable. If there are users and enough money coming in, all it needs to do is exist. And maybe some day there will be a company like Nokia putting its weight behind it and it can gain marketshare in big numbers. But even without it, it is a viable platform, just with some drawbacks.
So yes, for the near future I keep on using my Sony Xperia with Sailfish OS. I also keep an eye on the Pinephone and Librem 5, but today they are not ready to be a daily driver for me.
Stories about "Linux as the underdog" have confused me for several years. The truth is, Linux has won, and it's OK to recognize that.
The majority of phones sold are "Linux phones" -- it's the kernel for Android!
The top selling laptops for years provide "Linux on the desktop" -- Chromebooks run ChromeOS, which is Linux!
Nearly all host servers, at nearly every cloud provider, run Linux. The vast majority of cloud VMs run Linux.
The vast majority of set-top boxes, IP cameras, IoT controllers, and other embedded devices are Linux based.
Linux is even powering an autonomous helicopter on MARS, that has been wildly successful in its mission.
The narrative of "poor underrated Linux gets no love" was plausible until maybe around the mid-2000s. But it's long past time to drop this idea as a community.
Now, I think it IS fair to say that several forms of userspace ecosystems, based on the Linux kernel, have failed to get traction over the years. Maybe we can say "Ubuntu Touch" failed to get adoption, but that has little to do with some characteristic of Linux itself.
Google is absolutely itching to replace the last GPL component of Android – Linux – with their own software, Fuchsia. Google will then control the whole stack, with no obligation to anyone. This has been coming for a long time; IIUC, Android started with many components from the larger GNU/Linux ecosystem, but today only Linux itself remains.
Linux is pervasive because it's free (gratis) and flexible, but customers aren't necessarily seeing the core benefits of the Linux philosophy in terms of liberty and privacy. These examples of Linux succeeding are where some sort of application stack is very non-libre and often non-gratis, but maybe that's just one example of how to commercialize Linux. It's a generalized and well supported base for commercial products and services that is less restrictive and costly than other OSs, which is why we saw it replace server Unix distributions and even Windows Server in some cases. Many companies are embracing open source and extensibility more and more, though, which is definitely a good thing.
I'm interested to see a stronger focus on privacy and liberty in consumer Linux products like smart devices and IoT. Somehow iOS/MacOS, a markedly closed system, is the only one that really seems to be going in that direction right now.
A kernel that isn't exposed to userspace, POSIX and Linux syscalls are not public APIs on Android, which is composed of Java userspace and a well defined set of NDK APIs.
Any application that tries to use private APIs might be killed by Android Sandbox infrastructure.
As such Google can replace it by something else, e.g. Fuchsia, and app developers won't even notice that the kernel has changed (ART is being ported to Fuchsia).
Then on IoT space, the competition from BSD licensed POSIX clones is heating up, most OEMs don't want any GPL tainted OS on their devices, just in case.
The kernel isn't useful alone. I am 90% happy with my macOS. Its kernel is not as good as Linux, but it runs all my unix apps. I can use free, mature software solutions for most of my needs; E.g., I use mpv and a WiFi-connected HDD to listen to music. (With a much more pleasant playlist creation UI thanks to ugrep and fzf.)
Being good enough is enough. Linux on the desktop could be better, but for more than a decade it has been good enough. If linux phones became as usable as linux on the desktop, that would be good enough.
> But let's start by making a short list of main reasons of failure per effort, starting with the companies involved:
> ...
> Nokia (Maemo/Meego): Change in corporate strategy (new CEO),
> ...
Nokia Linux based phones didn't fail simply because of a new CEO, but because of a partnership with Microsoft that brought a CEO and a deal with Windows Mobile on Nokia phones in exchange for phasing out of Linux and Symbian devices. If it wasn't for Microsoft, Linux on Nokia phones would have succeeded.
Remember this every time you work with WSL thinking that Microsoft loves Linux and the Open Source model; they don't. They simply aim at controlling it, this time also from the inside (see Linux Foundation Platinum membership).
If you have Linux desktop, having a GNU/Linux phone is nice too. Not having to search for apps for every stupid little thing on some cesspool of an appstore would be great too, if you can just write a little script or whatever to scratch your itch, and be able to trust it wholly.
Anyone who thinks Linux can gain a momentum on mobile should look at Samsung and Huawei efforts to get off Google. It seems almost impossible even if you already have a foothold in mobile industry.
I want a phone that's Android in the hand but when I plug it into a usb-c hub, it projects full a fat linux distro on my external monitor(s).
Though I love Linux, it's not because it's Linux - I just want to live the one device life. If IPhones could run MacOS on external monitors, I'd buy an iPhone.
Alternatively, a web-based OS (with desktop linux on a hub) would be awesome if all my apps were available as web apps - but they are not so I'd still need a mobile OS and a desktop OS for the usbc hub life.
However it seemed to not been interesting to the users because noone really used it and they cancelled the project.
Their phones (the Galaxy series) still do support projecting a desktop mode when you plug the phone into a USB-C monitor. They now run "just" Android apps though.
On the topic of Linux phones, I have expressed my desire for more powerful hardware in the past. However, when I think about it, starting with modest specs means that things would have to be made to work properly on those. That means eventually when we move to more powerful hardware, the performance would be that much more better. Atleast that is how I imagine the silver lining to be.
So how would a Linux phone possibly gain momentum?
I have no idea how it solves any job to be done for almost anyone. Does it have the apps I want, including for my bank? Does it have a great camera? Does it have fantastic battery life? Right now, the answer to all of that is definitely not, and it becomes a chicken and an egg problem, where you don’t have the investment to build all that.
And don’t give me the obvious and trite answer of “privacy”. Because we have been bombarded with ”privacy” for years, and yet Google and Facebook keep raking in the cash and the users. I see no evidence the vast majority of people care at all. (And even then: prove to a user, or me for that matter, that a cobbled together Linux stack is in fact more secure than iOS. Or, explain how Linux magically protects me from all the ad trackers on the web.)
Smartphones = bunch of untrusted apps = security requirements
Is there an app store + readily available sandboxing solution that allows people to install some random guy's game on their phone without getting their credit card stolen?
I've tried and tinkered with all but the newest pinephone release. I sucks vs a similarly priced android phone but its working way more than I ever expected it to work this early on.
Hopefully we can get some solid modern hardware support in a linux handheld eventually. It would be nice if Qualcomm could open up a few things and make it easier for open source phone hobbyists to get things going.
The Pine64 community has been steadily growing and at least gives me confidence in what they can do with the older hardware the pinephone is working with.
Anyone know of any good Linux phones. Mine is going to be dead soon and for a while I liked Pixel because it didn't have all the Samsung or Motorola or anytime else bloatware, but with what Google has been doing I'd prefer a full Linux phone.
I've looked a Librem but $1500 is kind of hard to justify when I have all my other expenses that come with a young family.
[+] [-] bregma|4 years ago|reply
Why isn't it around any more?
(1) The carriers. Many people get a free or heavily subsidized phone from their carrier, who makes money from the data charges. Carriers were just not interested in phones from smaller manufacturers, and it's not entirely clear that there is no backflow of cash involved at higher levels. There is also the issue of controlling access to user metadata, always a profitable exercise.
(2) The manufacturers: why would they put a third-party OS on their phone when they control the entire stack from the silicon to the end-user's metadata through their in-house Android fork? They can sell hardware at a negative margin if they can profit from selling access to metadata, or possibly data.
(3) The apps. Developers want to target the biggest, most lucrative markets so they target Android and iPhone which already have a critical mass. You could provide a container to run Android apps on Ubuntu, but then you end up with no differentiating factor except your Android apps look like crap in desktop mode. Running in a container can also restrict the kind of data collection many apps rely on for monetization.
Another problem with apps is that many web services check to see whether the device is iPhone or Android and respond accordingly. Like the browser checks of yore ("you need at least IE 6 to use this site, try updating to a newer browser!") this gives a less satisfactory experience to end users.
[+] [-] Forbo|4 years ago|reply
The same shit is happening today, except it's Chrome instead of IE6. We have an entire generation of webdevs who learned fuck-all from the previous round of the browser wars.
[+] [-] dm319|4 years ago|reply
[+] [-] moksly|4 years ago|reply
I really doubt it’s something sinister in enterprise sized logistics. Mainly because the big guys don’t need to be al 90’ies Microsoft about it in the modern world of leaned up Enterprise. The service the big global brands come with do save the carriers a lot of money, but it’s done perfectly legal.
First there is the logistics. As a smaller carrier, you won’t really have to worry about doing BI on sales projections and what not, because Apple, Samsung and so on will simply tell you how many phones you can expect to sell, which is basically how many they’ve already produced for your region. They’ll handle everything from warehousing to shipping to your local stores pretty much without you having to do anything but sign the extremely “take it or leave it” sort of B2B contract, that also includes the big companies not charging you for products that don’t sell. What this translates to is that you can basically run your company without a logistics department or any BI related to the actual phone hardware. This saves you money both on manpower but also on not having too much inventory.
Then there the support side of things. By carrying big well known brands with a very low degree of user freedom for modifications you cut the non-carrier related support you need to offer to almost nothing. You could probably offer good money to get carriers to sell, even a well known brand like the fairphone, and they’d still turn you down because it would cost them too much money to do so.
[+] [-] chrisseaton|4 years ago|reply
Is this really the case anymore? Why would you buy a phone from a third party like that?
Most people I know buy their phone direct from Apple or an Android device from Amazon retail.
[+] [-] sajithdilshan|4 years ago|reply
[+] [-] collyw|4 years ago|reply
[+] [-] zepto|4 years ago|reply
‘It runs Linux’ is not sufficient differentiation for people to be interested outside of hobbyists and highly specified niches.
If we can change that, then a low end disruption strategy could work.
Unfortunately the work on Linux desktops and phones mostly seems to be about making an iOS-alike experience.
[+] [-] pabs3|4 years ago|reply
[+] [-] j-james|4 years ago|reply
The Android platform represents an enormous amount of work that encompasses a more secure base kernel, an unparalleled selection of applications designed for mobile usage and written in a memory-safe language, fantastic sandboxing and user privacy features leagues ahead of any desktop operating system, and great diversity in the hardware market.
While it's apparent that Android has significant downsides - de facto proprietary drivers, an environment of mostly closed-source applications, some Big G integration, and a general lack of long-term updates occasionally countering the work put into security (thanks, chip manufacturers: https://news.ycombinator.com/item?id=26593274), these are problems that can be solved. It doesn't make sense to throw the baby out with the bathwater.
Closed-source applications are more a moral problem than a security one thanks to Android's sandboxing, and a vibrant ecosystem of high-quality free-and-open-source Android applications (https://f-droid.org - Amaze, Notally, QKSMS, and Tasks are some of my favorites) makes them all but optional. Google's presence in the AOSP codebase is rather limited (mostly around notifications) and taken out altogether by custom ROMs like Lineage and GrapheneOS. Proprietary drivers remain as Android's biggest problem, perhaps alongside the manufacturers that ship locked-down phones with bloatware (looking at you, Samsung).
Why go the desktop-Linux-to-mobile route when you could fork Android, write drivers for the Pinephone / Librem 5, punch a hole to the base system for privileged applications, and have the best of both worlds with an order of magnitude less effort?
[+] [-] AussieWog93|4 years ago|reply
[+] [-] strictfp|4 years ago|reply
The problem in my eyes is plateauing. Once a project moves past the first couple of iterations, it rarely interests the original authors enough to make incremental improvements. And since there is no financial incentive to continue, the project usually cools down to eventually be replaced by the following iteration. This means that you rarely get mature software.
[+] [-] swiley|4 years ago|reply
1) The GNU utilities have pretty good CLI UI design and there are plenty of us who are perfectly happy with just that and a decent WM like FVWM even on a phone
2) It's not like closed mobile stuff has a great UI either. My favorite example is Apple's camera app. Did you know you can do manual focus with it? It's absolutely impossible to discover but I think it involves a "reverse three finger pinch." I've even had it explained twice to me (and used it) and forgotten how to do it after. I can never remember how when I need it. This garbage makes the TAR UI look like it was designed by an artist.
[+] [-] gspr|4 years ago|reply
I'm having a hard time understanding what it is about the development paradigm that's allergic to good UI design. In good faith I'll assume you meant that "FOSS projects seem to lack good UI design". This I'll agree with, if we're talking about GUI design (others have pointed out that the CLI interfaces are usually great). I really don't think the lacking GUI design is necessarily inherent in FOSS, but I too wonder why more people with design skills aren't attracted to working on FOSS stuff.
[+] [-] _0w8t|4 years ago|reply
I can understand that on 4” inch phones there were no space for extra buttons, but with 6” screens this became ridiculous. So much for “good” design.
[+] [-] Valmar|4 years ago|reply
Being able to design something friendly to the average user, yet powerful enough to be friendly to power users, is a very difficult task to accomplish.
Developers understand the latter quite well, I think, as they themselves are power user-types. The former... well, that requires having dedicated designers who know how the average user thinks, as well as perhaps guinea pigs in the form of the average user. And that might require a lot more funding, perhaps.
[+] [-] robolange|4 years ago|reply
* The leadership was terrible. They had no clue what it took to make a mass market product. They just assumed that they would sell hardware and a community would provide a working operating system and apps magically.
* The hardware was buggy. There was one issue that if you let the battery drain fully, you could not get the phone to recharge it and had to use an external charger. Another issue was that the GPS receiver was accidentally surrounded by metal, so barely functioned. There were all kinds of problems with the radios in the early days. Oh, and that touchscreen -- I guess it was typical of pre-capacative touchscreens, but it was hard to use without a stylus and impossible to hit widgets near the edge of the screen.
* The hardware was massively underpowered (compared to competitors) by the time the Freerunner actually shipped. Weak CPU, little RAM, 2G cellular radio in an era when 3G had become standard, so like 5kbps max data transfer.
*Because of the failure of Openmoko leadership, the community fragmented a hundred ways. This meant that there were a dozen or more "distributions" of an OS for the phone, and none could do more than one or two of the things a typical user wanted in a phone at that time. Then there was finally a big bloated distribution (SHR if memory serves) that packed in enough libraries to make a more or less "usable" device, but doing so maxed out the phone's meager storage and RAM, making it nearly impossible to do anything "smart" with it.
* Once again, in absence of strong UX leadership, the community resorted to dumping X11 apps without modification on the tiny screen. Think impossible to read fonts and dialog boxes that ran off the screen with no scroll capability. The vast majority of devs seemed to only use it by hooking it up to a computer via USB networking and SSHing into it.
As a technical user, I could live with this. Kinda. Sorta. Using it was an exercise in masochism. I was embarassed ... no ... humiliated when a nontechnical person compared their iPhone with the OpenMoko that I had talked up so much (before receiving it).
I had planned to destroy the phone in some fantastic fashion (e.g., melting it with a laser) as soon as I got a real phone. But by the time I could afford an Android, I was so done with it that I just dropped it in the trash (after wiping it, of course).
[+] [-] heavyset_go|4 years ago|reply
I also had the Palm Pre and Pre 2, both of which were based on Linux (and apps were made with HTML, CSS, and JS in 2009!), and those were also great.
[+] [-] seba_dos1|4 years ago|reply
No, not really. They had an exact opposite problem - they made several iterations of the default operating system, starting almost from scratch at each iteration, which burned quite a lot of energy and willpower of the community, which in turn focused their efforts on alternative distros like SHR or QtMoko.
> had to use an external charger.
Fortunately, you could use a standard Nokia BL-5C battery with the Freerunner. You didn't even need to have an external charger, just a charged spare battery would suffice. Also, IIRC this was an issue only with the first batch (so a small minority of produced phones).
> Another issue was that the GPS receiver was accidentally surrounded by metal, so barely functioned
Not really. The GPS problem was because of microSD clock interference. You could solder a resistor on microSD slot pins or use a software workaround that clocked the reader down enough to not interfere.
> it was hard to use without a stylus
I have programmed quite a lot on that touchscreen with OSK without using a stylus. It worked fine, but yeah, it would be much better if the screen weren't recessed (N900 did that well, that touchscreen was excellent).
> so like 5kbps max data transfer.
More like 100kbps.
> Then there was finally a big bloated distribution (SHR if memory serves) that packed in enough libraries to make a more or less "usable" device, but doing so maxed out the phone's meager storage and RAM, making it nearly impossible to do anything "smart" with it.
Uhm, no? SHR was a light OpenEmbedded-based distro with E17-based window manager. It was one of the snappiest and most reliable distros for that device, I used it for a few years as a power user and was very happy with it. And maxing storage wouldn't be an issue anyway since you could boot from an SD card.
Eventually I've switched to a N900 because of the Freerunner's slowness. If Glamo wasn't so slow I guess I would use it for a few years more before switching. I still have it and it still works, although I don't really use it anymore.
[+] [-] serf|4 years ago|reply
next time you have an urge to throw a battery-equipped device into the trash(!?), consider donation.
you may not like the device but someone out there may be thrilled at the chance to de-solder some useful components.
[+] [-] fsflover|4 years ago|reply
This is exactly how it works with Pinephone now, and it works really well.
[+] [-] moepstar|4 years ago|reply
I've also had a Freerunner, in theory the ideal pocket computer with Linux on it.
Practically, you could barely have a phone call with all the echo going on and navigation-wise it took _ages_ to find its GPS fix (no AGPS iirc) - the shielded antenna, mentioned down-thread, probably didn't help.
And yeah, whatever you had on screen was probably way too small to be read or interacted with :(
I think i got rid of it on eBay after a few weeks...
[+] [-] ericlewis|4 years ago|reply
[+] [-] jchw|4 years ago|reply
On the other hand, there is one confounding software issue that I think makes it the hardest as a full replacement for a modern smartphone: push notifications. An equivalent for APNS or FCM that allows for relatively low power consumption, low latency push notifications would need to be devised, financed and actually adopted. It’s nice to imagine a world where apps on phones are like how apps on desktop are and just keep, for example, active WebSocket connections at all times, but this doesn’t appear to be something practical with current technology.
Maybe federated, E2EE push notifications could come into existence? A sort of ActivityPub of pushes... It’s a pipe dream, but one can hope.
[+] [-] cookiengineer|4 years ago|reply
I don't think that people will use the phone solely because it's Linux. They'll use it if it can solve their tasks at hand.
Personally I think that Linux will never take off if we won't create financial incentives to make and deploy apps, additionally to fix the mess that's user rights/permission management in GTK and QT.
Currently there are no working native alternatives on Linux, libhandy is still a joke for simple tasks like a fading sidebar (or even swipe gestures!) and QT can't be used for anything serious due to their license.
My hope for servo, FirefoxOS and WebOS was that there will be some day an alternative to Electron that's focussed on permissions and sandboxing, and that allows to be externally configured and is more modular on the environment (VM) level. Basically like a settings app on Android that can toggle GPS, toggle Networking access, toggle Camera access etc.
Layouting-wise CSS has won the masses. Everyone that tries to reinvent it for their own opinionated views has failed and given up (including me who spent over 6 years developing an isomorphic App Engine full-time). It's time to let go. The web has won and there's no point in denying it.
React has won most of the developer crowd because of React Native and convenience of staying in the same language and more importantly, being able to reuse the same code architecture.
I think in order for Linux to succeed there has to be a compile pipeline (and runtime?) that allows to deploy those apps easily. If that's not possible (due to whatever reasons) the platform will never really take off.
Privacy and Security and Openness is just no argument for the average user that doesn't care and just wants to get their tasks done.
[+] [-] ldiracdelta|4 years ago|reply
[+] [-] mch82|4 years ago|reply
I get the memory/performance issues with Electron. Those can be solved. I’m able to open large XML files in VSCode that simply crush other apps like Notepad and XMLSpy.
[+] [-] lytefm|4 years ago|reply
I agree. An important difference to the situation a few years ago is the availability of actually viable cross-platform development frameworks like React Native, Flutter, Capacitor,...
If all I need to do in order to support a new niche OS is configuring one additional build target, I might just go for it. Otherwise, certainly not.
[+] [-] hutzlibu|4 years ago|reply
Of course they will. It is just that hardcore linux people are not enough, to bring enough money in.
" Privacy and Security and Openness is just no argument for the average"
And it actually is a argument for average users, it is just that their priorities to get their tasks done are higher, which is rational.
And since their tasks usually involve whatsapp and co. they won't be able to use a Linux phone.
[+] [-] mpol|4 years ago|reply
But if a smaller company like Jolla can do it, it can be sustainable. If there are users and enough money coming in, all it needs to do is exist. And maybe some day there will be a company like Nokia putting its weight behind it and it can gain marketshare in big numbers. But even without it, it is a viable platform, just with some drawbacks.
So yes, for the near future I keep on using my Sony Xperia with Sailfish OS. I also keep an eye on the Pinephone and Librem 5, but today they are not ready to be a daily driver for me.
[+] [-] tedchs|4 years ago|reply
The majority of phones sold are "Linux phones" -- it's the kernel for Android!
The top selling laptops for years provide "Linux on the desktop" -- Chromebooks run ChromeOS, which is Linux!
Nearly all host servers, at nearly every cloud provider, run Linux. The vast majority of cloud VMs run Linux.
The vast majority of set-top boxes, IP cameras, IoT controllers, and other embedded devices are Linux based.
Linux is even powering an autonomous helicopter on MARS, that has been wildly successful in its mission.
The narrative of "poor underrated Linux gets no love" was plausible until maybe around the mid-2000s. But it's long past time to drop this idea as a community.
Now, I think it IS fair to say that several forms of userspace ecosystems, based on the Linux kernel, have failed to get traction over the years. Maybe we can say "Ubuntu Touch" failed to get adoption, but that has little to do with some characteristic of Linux itself.
[+] [-] teddyh|4 years ago|reply
[+] [-] david-cako|4 years ago|reply
I'm interested to see a stronger focus on privacy and liberty in consumer Linux products like smart devices and IoT. Somehow iOS/MacOS, a markedly closed system, is the only one that really seems to be going in that direction right now.
[+] [-] pjmlp|4 years ago|reply
Any application that tries to use private APIs might be killed by Android Sandbox infrastructure.
As such Google can replace it by something else, e.g. Fuchsia, and app developers won't even notice that the kernel has changed (ART is being ported to Fuchsia).
Then on IoT space, the competition from BSD licensed POSIX clones is heating up, most OEMs don't want any GPL tainted OS on their devices, just in case.
As for the desktop, I already gave up.
[+] [-] Siira|4 years ago|reply
[+] [-] marcodiego|4 years ago|reply
[+] [-] squarefoot|4 years ago|reply
> ...
> Nokia (Maemo/Meego): Change in corporate strategy (new CEO),
> ...
Nokia Linux based phones didn't fail simply because of a new CEO, but because of a partnership with Microsoft that brought a CEO and a deal with Windows Mobile on Nokia phones in exchange for phasing out of Linux and Symbian devices. If it wasn't for Microsoft, Linux on Nokia phones would have succeeded.
Remember this every time you work with WSL thinking that Microsoft loves Linux and the Open Source model; they don't. They simply aim at controlling it, this time also from the inside (see Linux Foundation Platinum membership).
[+] [-] rbanffy|4 years ago|reply
[+] [-] megous|4 years ago|reply
If you have Linux desktop, having a GNU/Linux phone is nice too. Not having to search for apps for every stupid little thing on some cesspool of an appstore would be great too, if you can just write a little script or whatever to scratch your itch, and be able to trust it wholly.
[+] [-] f6v|4 years ago|reply
[+] [-] apatheticonion|4 years ago|reply
Though I love Linux, it's not because it's Linux - I just want to live the one device life. If IPhones could run MacOS on external monitors, I'd buy an iPhone.
Alternatively, a web-based OS (with desktop linux on a hub) would be awesome if all my apps were available as web apps - but they are not so I'd still need a mobile OS and a desktop OS for the usbc hub life.
[+] [-] izacus|4 years ago|reply
However it seemed to not been interesting to the users because noone really used it and they cancelled the project.
Their phones (the Galaxy series) still do support projecting a desktop mode when you plug the phone into a USB-C monitor. They now run "just" Android apps though.
[+] [-] ajot|4 years ago|reply
https://maruos.com/
https://github.com/maruos/maruos
[+] [-] noisy_boy|4 years ago|reply
[+] [-] perardi|4 years ago|reply
I have no idea how it solves any job to be done for almost anyone. Does it have the apps I want, including for my bank? Does it have a great camera? Does it have fantastic battery life? Right now, the answer to all of that is definitely not, and it becomes a chicken and an egg problem, where you don’t have the investment to build all that.
And don’t give me the obvious and trite answer of “privacy”. Because we have been bombarded with ”privacy” for years, and yet Google and Facebook keep raking in the cash and the users. I see no evidence the vast majority of people care at all. (And even then: prove to a user, or me for that matter, that a cobbled together Linux stack is in fact more secure than iOS. Or, explain how Linux magically protects me from all the ad trackers on the web.)
[+] [-] 29athrowaway|4 years ago|reply
Is there an app store + readily available sandboxing solution that allows people to install some random guy's game on their phone without getting their credit card stolen?
[+] [-] dannyw|4 years ago|reply
[+] [-] officeplant|4 years ago|reply
Hopefully we can get some solid modern hardware support in a linux handheld eventually. It would be nice if Qualcomm could open up a few things and make it easier for open source phone hobbyists to get things going.
The Pine64 community has been steadily growing and at least gives me confidence in what they can do with the older hardware the pinephone is working with.
[+] [-] dehrmann|4 years ago|reply
[+] [-] hpoe|4 years ago|reply
I've looked a Librem but $1500 is kind of hard to justify when I have all my other expenses that come with a young family.
Anyone have any suggestions?