I really wanted the Ubuntu phone to succeed. I backed the Indiegogo for their fancy phone, and when that failed I installed Ubuntu OS on a Nexus 5 to play with.

I never activated any phone service on it but I think I would have enjoyed it if I had. It was kind of neat to have a smartphone that didn't hide the fact that it was a computer. Even without plugging it in to a monitor or anything, I was able to play with the Chrome dev tools on the fly and it was pretty fun.

I used to use Ubuntu Touch on my OnePlus One for quite a while and it was very nice! I had to switch away because it didn't really support group MMS (still doesn't from what I can tell), and then later US carriers started requiring VoLTE which it didn't support for a while either. But I still hope to switch back someday, that was the most enjoyable phone I'd used since the N900.

I haven't tried to use it full time for the same reason that I haven't tried to take the plunge to the dumb-phone world.

A part of me occasionally considers buying one of those cheap KaiOS dumbphones, because I figure that the stuff that's missing could be supplemented with my laptop. The problem I'm always afraid of is that I'm afraid that I'll be missing something. I'll start a job and they'll require some specific proprietary 2FA app, for example. Also, I don't think Signal has a dumb phone app yet and that's basically all I use to communicate.

Similarly, I haven't done anything but play with Ubuntu Touch, because I know there would be an app missing that I would need, so I'd be forced to have an Android or iPhone for those edge cases anyway.

Startup costs are enormous to break into the smartphone world now, at least if you want to penetrate into the "non-tech-enthusiast" market.

IMO it comes down to marketing: can't have the kayfabe of selling something that is "not a computer"/"new kind of computer" and have it act like a "computer" too

Marketing played a role yes, but plenty of other phone operating systems failed that had much stronger marketing then Ubuntu ever would have.

Not a phone, but I really liked the Blackberry Playbook's QNX system. It was extremely smooth and easy to use, and was fairly easy to develop for.

They shit the bed by betting on Flash, which was a dying tech, but I was pretty sad when Blackberry just went to "another Android phone", since I thought what they had was pretty neat.

Blackberry definitely had much better marketing than Ubuntu has ever had, and for reasons probably too complicated for me to fully understand, they're sort of a joke now.

I ran Ubuntu Touch briefly on a Nexus 5 as well. IIRC the two issues I had were:

1) calls and MMS not working well.

2) Instead of a normal GNU/Linux OS Ubuntu Touch tried really hard to have an Android style immutable OS. Kind of the worst of both worlds since you have a difficult to work with OS but without the app ecosystem of Android that some people believe makes it worth using.

After that I kind of just gave up on the idea that I owned my phone at all and a few years later I gave up carrying smartphones entirely.

> but the ecosystems are pretty well separated for most

This seems to be a product decision not a limitation. Apple cannot implement this without self-cannabalizing sales of the other.

It sounds like a nice idea in theory but I dont think demand is that high.

Except all those people who don't own a computer and are stuck on old Android phones. Debian can allow lifetime updates btw.

Apple has this out of the box with MacBooks + iPhones. But in their case, it's easy to implement.

You can get like 65% of the way there by just using i3wm with an onscreen keyboard and really big window borders (so you have somewhere to poke to change windows). But you have to contend with the fact that it is a basically fine touch window manager showing you… applications that were designed without touch in mind at all.

You don't have to suffer: Phosh and convergent apps exist.

> A dozen years later, nobody has done that well.

I'm writing this from Librem 5 phone running PureOS based on Debian. No Android dependencies and possibility of full desktop mode. How is this not a success?

You and several thousand other people see it that way, meanwhile the billions of other computer users out here don’t.

Do they have a choice? I mean, an actual one that doesn't take significant technical ability.

Those billions need to accept the facts. They're not getting a Windows phone again, and Apple won't likely cannibalize Mac demand with the iPhone.

If they want to have their cake and eat it too, they can either run Linux in an Android container, or Android in a Linux container.

People don't want to quickly turn their phone into a desktop, they want to quickly turn their phone into a laptop. Apple's laptops are over 75% of its sales. And if you have the screen and keyboard of a laptop, why not put a CPU, battery et al. in there?

I think you are missing the point. People don't want to turn anything into anything. They want a minimal number of devices covering all their needs.

then I guess we'll have to wait for companies like huawei with their mate xt, or more likely given friendshoring, the samsung trifold to give people phones that turn into tablets and have a companion keyboard for them to feel this is enough of a threat to address

Popularity =/= quality.

That being said, obviously the Librem phone is missing functionality.

Most computer users are goo goo ga ga level users. Most people in the US use iMessage, for example.

It's not the best messenger, not even close. It's not the most ubiquitous either - doesn't even break top 5 globally. Doesn't have the most features. Not the fastest. Not the most secure. Not the easiest to use.

They just use it because it's already installed and right there. That is how deep their understanding and comprehension can go.

You picked the exact worst example to illustrate your point. People do not pick their messaging service, it is driven by adoption of everybody else.

That's why you're downvoted.

Right, but as ive stated: thats popularity, not quality. There's a difference. Picking something for popularity is perfectly valid.

For Operating Systems and Ecosystems, Popularity == Quality. The more popular a software is, the more edge cases it encounters and solves issues. More people will write software, plugins, tools for it.

Are you saying that Windows is the most high-quality operating system in the world? This is not even funny. Did you hear about the concepts of walled garden and network effect?

There are a lot of people who are not "goo goo ga ga" level users and really need a device.

I've switched to UMPCs because the constant lies and intentional breakage from smartphone vendors drove me up the wall.

glad to see I'm not the only one, if you don't mind me asking, what OS did you settle on? I'm using mint w/ auto-cpufreq on a donki nanote next right now but it's not perfect.

I use Artix on everything these days. It used to Alpine (and before that I was just making my own distros because I didn't like the decisions eg Debian was making.)

thanks! wasn't even on my radar. mind if I ask the device?

GNOME is still plugging away at this, making sure their entire OS is usable on mobile. Even without a market or audience.

Both impressively and disappointingly, that seems to be literally one guy in Germany.

https://gitlab.gnome.org/verdre/gnome-shell-mobile

relevant https://xkcd.com/2347/

Purism company contributed a lot to mobile Gnome UI.

> Recent iterations of iOS are getting closer to being able to replace a Mac for a class of tablet-owning users who don't need desktop software, but the ecosystems are pretty well separated for most.

Yeah, apple could solve that instantly.

Sadly, they decided to play a cash cow strategy of milking existing users instead for trying to grow market share.

Apple is so frustrating. Amazing hardware but then they complete shit the bead with the software side of things and act actively hostile against their own users.

It's a product decision.

Apple decided when the iPhone first came out to redo the whole UI stack to be touch-first. There are hardenings for security and battery preservation, but that's arguably the biggest difference between iOS and macOS.

Windows and Linux have tried to retrofit their ecosystems to also work on touchscreens. It hasn't gone well. Too many apps assume a mouse-equivalent pointing system.

Adaptive design has fared best on the web, but it's still not settled. See, for instance, the back-and-forth around density defaults for web apps like Gmail. Some people really like their pointer-friendly dense UIs with hover buttons, and that makes them really hard to use on a touchscreen.

Perhaps ironically, Apple is also in the best position to bridge the gap. Since they own the UI stack that renders most apps on Apple devices, they could do something clever like say "a button is 32px tall if there's a trackpad and 56px tall otherwise." Rules like that could produce an app that truly adapts to the user's primary modality.

The other systems have too many UI frameworks for that to work. Apple could only pull it off because ~everyone uses their component set (I think it's called UIKit). They also have a reputation of declaring the future and making the ecosystem catch up (going all the way back to adopting USB exclusively on the first iMac).

There are no new native apps, only Electron or similar.

This is only sort of sarcasm.

> Windows and Linux have tried to retrofit their ecosystems to also work on touchscreens. It hasn't gone well.

Concerning Windows, you're right. Concerning GNU/Linux you aren't: https://news.ycombinator.com/item?id=19328085

That post is 6 years old, and yet Linux still sucks by default on a touchscreen.

I installed NixOS on a handheld nearly 2 years ago and tried a whole bunch of touch-oriented options (including tiny ones like Maui). The closest I got to good enough was running the mobile fork of GNOME, using someone on GitHub's custom flake to pull it in at HEAD. I'm very happy that Valve has released SteamOS for other devices so I can offload that tinkering.

Steam is very usable on a touchscreen, but the KDE desktop mode hasn't impressed me.

You must use the plasma-mobile stuff, not the plasma desktop.

Mobile wasn't packaged for NixOS.

I am curious why/if Valve hasn't used it for SteamOS's desktop mode though.

Package it yourself then!

I don't really know, but I don't think the use outside of steam is an important use case to them, possibly they don't want to make things too easy, or maybe it wasn't packaged for arch either when they created the base image.

I'm writing this from Librem 5 phone running PureOS with Phosh DE. It's convenient and pretty in my opinion. Also it runs desktop Firefox with all plugins you want.

I think the Steam Deck’s touch screen is also one of its worst features.

I will type a message or an email on a phone (Android or iOS) but I loathe even typing my username into the Steam Deck via touch.

I know for a fact that many university students are using iPad OS instead of macOS now, especially for less typing-intensive stuff like chemistry. You might be surprised how much of college runs in the browser now.

Linux barely works on a limited set of hardware that it was designed for. Indeed I don't think it's reasonable to then also make it work on some completely different phone hardware that has a operational life of like 2 years + full of closed source hardware and drivers.

If you want users, the thing has to be usable. For the thing to be usable it needs software with perfect hardware support. Google, Android vendors and Apple (and to some extend System 76) understand this.

For Linux to gain 5% marketshare, I really doubt it "barely works" on a "limited set of hardware that it was designed for". It can run headless on basically anything better than a Pentium, and it mostly just works on average hardware (except fingerprint sensors and Nvidia). I've had no problem with Linux on all my hardware, and I have a feeling the last time you looked at it was 2013.

Besides working with it on servers on regular basis, and having had enough with it on desktops and laptops since 1995, last year I managed to get a NUC, where Ubuntu, Red-Hat and SuSE weren't able to boot from the internal SSD or get along with UEFI, only booting from the external drive worked.

So yeah no problem, and yes I know should have gone to the usual forums asking everyone and their dog if someone before me had ever succeeded installing Linux on this brick.

I have different opinions. Try my NUC. Reinstall Windows 10 on it; you won't get audio support unless you get especial audio drivers from a shady download URL (usually used from pirating services) linked from the OEM vendors. Windows Update won't help. Neither SDI Tool Origin.

GNU/Linux works, HDMI output et all.

Unless I’m mistaken windows 10 is end of life in 3 months so this shouldn’t be surprising ?

Barely no one uses Windows 11.

I think about half of all desktop machines , since MS is pushing upgrades aggressively. Many corporate environments ( have already moved as well.

https://gs.statcounter.com/windows-version-market-share/desk...

Nope. Win 10 is now a free year of support. All the way to Oct 2026. Thanks for pointing out mistakenly your mistaken. And is it surprising than hardware shills are pointing not pointing this out? You want Linux get a Thinkpad, and a mouse, and plan on replugging every time you wake. "It just works." Well guess what?

Microsoft says it ends in October 2025.

What am I missing here ?

https://support.microsoft.com/en-us/windows/windows-10-suppo...

I love how HN nowadays mirrors the Linux forums experience from phpBB and Slashdot days.

Because with NUCs if you dare to gasp reinstall the OEM OS, if the vendor it's gone, with Windows you might get a nice door stopper. You can get audio thru the headphones/speakers output, but that's a shitty experience for your living room.

With Windows 11 it will happen the same.

then don't use it. Linux needs less people like you

And I don't, Android/Linux, WebOS/Linux and VMs are good enough.

If others have fun making it work, good for them, I have better ways to spend the rest of my life on this litte sphere.

As what Linux needs, I thought it needs every little user, to make a difference against Apple and Microsoft, after all we are always told that now is the year everyone is going to flee into Linux.

Agreed - my view is that Linux on the desktop failed and will never succeed and its mostly alright since people have shifted their computing needs from desktops to phones.

So the market for personal computers has grown massively with phones entering it, and Linux has won here.

As to the general state of Linux usability - I’ve been using Linux since 1995, and professionally for more than 20 years - it’s great for professionally administered servers or workstations, less great for regular desktops / laptops where things pretty never work out of the box which is unfortunately not going to bring in extra users …

Linux may have won on phones, but free software did not, which is the thing that actually matters.

That’s a different issue but you have a point esp since the Linux vs windows fight back then was essentially framed as proprietary vs free.

Running as a user where things need to work is not the same as being headless wherrle all you need is CLI access, disk and network ...

My lenovo p14s is a great linux laptop unless you want it to sleep (which it does!) It even wakes up! But 50% of the time the trackpad does not wake up properly ... Making hard to be used as a laptop that I can get things done on

Linux didn’t work reliably on my laptops in the past 10 years. And I mean basic things, like booting up, or showing a desktop without serious glitches. And all the time problems were non deterministic, and printing just generic unhelpful “something is wrong” errors, if there were any. I try it every year, whether the situation is better, and in the past 25 years, the answer is that barely. Yes, you can have terminal with very basic settings almost every time, but if you want anything more, even just like proper resolution, then it’s still a lottery. The interesting thing is that before that I was luckier, I could hack Linux to many things (I would definitely not say “install”).

And of course like in the past decades any time, you can always use Linux in VMs. Very reliably. So I stick to that.

> Linux didn’t work reliably on my laptops in the past 10 years. And I mean basic things, like booting up, or showing a desktop

Which laptops? Do they all have Nvidia graphics? This is really vague. Your comment is not helpful, and it just looks like usual Linux bashing from people who don't know what they're talking about.

FWIW I have Mint on two grandparents laptops and I agree with the parent comment. I gave up doing any updates on them, I just wipe off everything once a year and reinstall from scratch. Even though it's only browser and music player and nobody touches the settings (unless there was a misclick) yet still last time the desktop came to show through the application windows so I had to wipe it again and lo it worked.

1. You didn't name the laptops just like the OP.

2. It seems you're using hardware designed for Windows. It's like trying to install Windows on a Macbook: You're on your own here. Don't blame Microsoft or Apple for your issues.

You know, reading through your replies in this thread is kind of just sad. You selectively pick only tiny parts of comments to reply to, while ignoring the points that are made and the plethora of very, very common issues highlighted.

Sadly, it's also a very expected answer from someone who probably likes Linux a lot. No matter what argument is made, no matter what examples or proof are included, irregardless of the number of such things, it's all just user error in the end.

When the argument is that Linux only _properly_ supports a small subset of devices for _desktop computing,_ and the counter-argument is that it does, yet a person mentions very, very incredibly common issues . . . Is the proper response to say, to effect, you're the one at fault by not using the right hardware? No, no it's not. Given that so, so, _so_ many experience these kinds of basic UX problems means that either Linux doesn't actually properly support a lot of different hardware (despite what many obsessed Linux users screech/preach) or it supports them perfectly fine and the tons upon tons upon tons of people who experience these issues are all the ones at fault and in no way, shape or form does Linux lack from a well known general UX problem.

Even worse is this, honestly, moronic take:

> 2. It seems you're using hardware designed for Windows. It's like trying to install Windows on a Macbook: You're on your own here. Don't blame Microsoft or Apple for your issues.

If I wasn't before convinced that you aren't completely out of touch, I am now. The primary subject of _desktop_ users _are_ Windows users with Windows machines. Goodness gracious, I honestly do not understand this take. Most of Linux's not-always-optimal desktop hardware support comes from a plethora of formerly Windows machines.

It's honestly just so weird to argue that installing Linux on a Windows machine is in any way shape or form as bad as installing it on a Mac.

Even more, Linux-only devices, as in Linux-first devices, are more of a newer thing. There have been initiatives before, that have fallen through or not gained a lot of traction. It's not until relatively recently that companies making Linux-first devices for _desktop_ computing became a real thing. And even then, the Steam Deck blows them out of the water by mostly _removing_ the desktop-part of Linux and replacing it with a homebrewed one. Although you can still use a regular desktop "mode." And yet, even with such companies, Windows machines stand for a massive section of the Linux desktop computing userbase.

Ironically, this is exactly the kind of argument I'd expect from an obsessed and out of touch Linux user. Because it makes no sense, disregards the points other potential users make and doesn't even properly address them.

Thank you for an interesting, thoughtful reply.

> You selectively pick only tiny parts of comments to reply to

Yes, I nitpick on some smaller things when I know the poster is wrong on that matter (and I feel it's sufficiently important).

> while ignoring the points that are made and the plethora of very, very common issues highlighted

It doesn't mean I think there are no issues at all. Often I just don't have anything to add to them. Do I always have to say "it's true" to everything else? I never said "and therefore you're wrong about everything".

> Linux only _properly_ supports a small subset of devices for _desktop computing

This is true. However what you conclude from this may differ. It doesn't mean that Linux is a lost case and everybody should use Windows. It doesn't mean that companies can't switch to Linux. Very often they choose the hardware for the next upgrade and nothing prevents them from choosing a known supported hardware, however rare it could be.

Also, in every Linux thread there is always a large number of people with vague, unhelpful Linux bashing like "I'm trying Linux every other week and there are countless issues every single time". This is not actionable, it doesn't help anybody to choose the hardware. It's just a shallow dismissal, which is against the HN guidelines. I don't understand who upvotes that.

> Is the proper response to say, to effect, you're the one at fault by not using the right hardware?

This is not exactly what I'm saying. I'm saying that you should manage your expectations when using random hardware with a free OS. Not only the Linux developers weren't paid to support every hardware in world -- in many cases, the hardware vendors actively fight against this. Statistically, it must be a really large number of devices reliably working. How else could Linux reach >5% without anything sold in stores?

> either Linux doesn't actually properly support a lot of different hardware

Yes, this is the case.

> Even worse is this, honestly, moronic take

I stand by what I said. You can't demand any OS to work on every piece of hardware in the world; not even mega-corps can achieve that. You have to keep in mind the reality, otherwise this is a Nirvana fallacy. I would understand if someone said "I'm unlucky, Linux doesn't work on my hardware". Instead, they say "Linux barely works on a limited set of hardware that it was designed for" or "Linux on the desktop failed and will never succeed" (actual quotes from this discussion).

> The primary subject of _desktop_ users _are_ Windows users with Windows machines.

This is very true, and it's a sad state of affairs that for Linux to grow it has an impossible task of supporting all those different Windows machines. I recommend every Windows user to try installing Linux. At the same time, I do not guarantee that it will work, but there is a good chance. You will likely loose some features like suspend, but your machine will be faster, more secure, private and supported for the lifetime if you're lucky. Also you should be able to pay somebody to fix your problems, which will fix them for every other user. This is not true for proprietary software btw.

> It's honestly just so weird to argue that installing Linux on a Windows machine is in any way shape or form as bad as installing it on a Mac.

This is not what I said. I said that you're misplacing the responsibility here. The user should accept their responsibility for buying a "Windows-certified" hardware and not blame their problems on volunteers. They should thank the volunteer when it works and understand the reasons when it doesn't (and consider helping to fix the issues, too).

> And even then, the Steam Deck blows them out of the water

What are you talking about? Most users buy it as a separate device to play games, just like a console. It does help Linux to grow, but the comparison to Linux-first general-purpose devices doesn't seem relevant here.

And how it would be helpful for me to specify? Do you think that I just gave up immediately, and didn’t try to solve it for days and weeks? Do you think you can help me more than that without the exact errors which I got? Do you think I care about a solution to laptops which I don’t even have? Do you think any kind of blanket term like “NVIDIA whatever doesn’t work” would be even true?

I specified a few times, and I got zero relevant information. My time is more valuable than reading comments which are like “it worked for me with <something-which-is-not-even-in-my-spec>”.

Works reliably on mine. Windows 11 on the other hand, feels beta.

> I have a feeling the last time you looked at it was 2013.

I literally looked at it last week. Spent multiple days on it. Tried Mint and Zorin (full install, not just live).

This is on a brand new Lenovo p16s Gen 4 with AMD (no nvidia). That laptop didn't even exist before this year.

List of problems I encountered:

-- Multi touch not working (fixed by switching from Mint to Zorin, upgrading mainline kernel in Mint did not help).

-- External monitor not working (had to install display link drivers by going into terminal and running scripts and all of that other classic Linux usability) .

-- Hardware video acceleration not working (scrolling super slow, maps super slow, entire system super slow). Had to install AMD display drivers for that separately. Upgrading mainline kernel worked for Mint, but not for Zorin. Installing AMD drivers in Zorin involved downloading the drivers, !editting an install script that is part of the drivers! and then having an LLM guide me through the rest of the extremely elaborate process of installing the driver.

-- And to top it off, my classic pet peeve: there's no way to configure something as basic as scroll-lines (mouse scrollwheel speed) through a GUI in ANY of the distros. It involves installing imwheel, !writing a script!, setting the script to run on boot and then rebooting (and/or restarting the script).

So no. There's definitely no "it just works". Not even on a laptop that is supposed to have official HW vendor support for Ubuntu.

Also, I only ran it for like a day. I'm sure that I'll run into tons of other issues if I use it a bit longer.

Good for you and lucky you that you got it to work. But for most of us Linux is "nice try, but it's not finished yet" .

> That laptop didn't even exist before this year.

Here's your problem. The hardware wasn't designed to run Linux and you gave Linux no time to fix the related problems. Try older hardware or wait.

It works for me, said every Linux zelot ever. NVidia acceleration is horrid. The only thing that worked was directx on zorin. AMD? Only a few machines, even fewer models. What all the zelots we're running is built in Intel. "It works for me." Followed by bins by the truck full of unsupported hardware.

ThinkPads? Perfect every time? Dell? Have you configured you bios lately? Unsupported AMD? Grab your crayons on a race the old hardware will loose.

Linux has had 10 years to run on 10 year old hardware. Linux is not an OS. It's an attitude, and it's not a reflection on the technology, it's a reflection on the lowest common denominator.

Windows 10? I still have another 14 months left, and mass*.dev to thank.

Hardware that was built to run Linux? Drivers? If you have an AMD card and any interest in Linux, pull the card and prepare to trash the entire system, like the HP sitting on the sidewalk last night. Bye.

> ThinkPads? Perfect every time? Dell?

System76, Purism, Taxedo, Pine64.

Which brings us back to my initial point: it only runs properly without fiddling on a tiny minority of very specific hardware by a tiny handful of vendors (who often ship to only a tiny portion of the globe).

To be clear: I got it to run eventually after hours/days of fiddling.

Do I like it after all this effort?

No.

But sure, I'll wait. I've waited for Linux to become truly user friendy for 3 decades and I really really want to love it, so I'll wait some more. No problem.

Linux runs on everything that upstreamed it's drivers, and then almost everything that didn't. By just reverse engineering and guessing.

And, a lot of this firmware is extremely buggy too. Have you see ACPI tables in laptops? But, they work under Linux. They shouldn't, but they do.

What doesn't work is the intersection of closed-source firmware and extremely eccentric or evil firmware. I think a lot of Android parts manufactures don't want to upstream their stuff because it's extremely bad and probably filled to the brim to vulnerabilities.

But, Intel upstreams everything, and so does AMD - and it's only improved their firmware quality.

Pity that AMD doesn't upstream everything they do with ROCm for everyone.

Linux is used on a far greater variety of hardware than Windows or Mac, from phones to supercomputers, and everything in between. The number of servers, of all architectures, running Linux completely dwarfs either Windows or Mac.

> Linux barely works on a limited set of hardware that it was designed for.

This is demonstrably false, and I don't understand where all these comments from Linux haters came from. My Librem laptops work flawlessly, including suspend.

In the long term markets usually specialize https://ashishb.net/tech/the-android-chrome-merger-saga/

I think the true reason is that onñy a small fraction of people are interested in convergence and most people are fine, or even truly desire, to have different devices and experiences.

Millions (billions?) of people are happy to leave their phone in another room when working on their computer and vice-versa. Sure you could use a do not disturb button but it would be a major PITA to have enough granularity to allow or disallow certain app/services to notify you and you would be certain to forget to activate/deactivate it when you really want.

Billions of users don't have a computer they work on and don't realize that this might be an option.

Imo trying to make a single UX that just changes a little bit to suit different device types is a misguided approach. Using a large screen with a mouse and keyboard is a fundamentally different thing to using a phone with a touch screen.

Using the same hardware for both would be super useful, but the software stack from the desktop environment upwards should be entirely different (yes, including most of the applications!)

There are some fuzzy boundaries - e.g. imo Gnome 3 has proven that a single experience can feel good on both a tablet and a single screen laptop with a good track pad. But I think paradoxically you need to take different approaches on different use modes if you want to provide true unity.

Try the KDE applications designed for this, like kasts, alligator, qrca and see for yourself.

I don't think it can necessarily work for any kind of application, but for some simpler ones I think it's completely fine.

> designed for this

This is kind of my point - I'm not saying you can't have applications that are usable across multiple UX paradigms, and I'm also not saying you can't write a UX library that automatically translates at least simple applications with little manual effort.

I'm just saying this requires active buy in from application developers into the ecosystem - you can't just run everything on all devices and have it magically work (with usability comparable to current state of the art in single device applications).

And to have buy in it needs to exist first :)

Like websites nowadays being usually designed for mobile and desktop devices

kirigami

> I'm just saying this requires active buy in from application developers into the ecosystem

It's true, but on the other hand, it's much easier to adapt your desktop app to a smaller screen than to rewrite it completely for another OS.

> Using the same hardware for both would be super useful, but the software stack from the desktop environment upwards should be entirely different (yes, including most of the applications!)

Phosh (Phone Shell) already exists and works quite well. I'm writing this comment from desktop Firefox running in Phosh on Librem 5 smartphone.

See also: https://videos.puri.sm/pureos/l5-convergence-purism.mp4

Plasma mobile is not flawless but it's very good. I use a tablet daily to do tablet stuff: read news or play sokoban while pooping, watching cartoons…

The people I knew with a Windows phone said it worked awesome for them, the desktop mode.

Samsung DeX is not bad! It is not as good as a real desktop, but it will do in a pinch.

I kind of doubt it did work well for them. Microsoft was having an identity crisis at that time where they didn't yet realize nobody wanted their tablet UI on desktop.

If they allowed straight Win32 ports on ARM it could have been interesting. Or if they did it now, where I think x86 emulation on arm is working well these days.

In fairness, most people I knew with Windows phones worked for MSFT.

But a few (the ones I talked to the most) were just people interested in the Windows ecosystem. One was running Windows Home Server. Those folks all swore that everything was great (except the app ecosystem). I can only go by what I was told.

I tried out that windows phone continuum feature back in the day to try and get actual work done. Here's what I remember of it.

You could connect it wirelessly to your windows laptop and it would take over screen, keyboard and mouse from the laptop. The actual connecting part worked smoothly, but it could only stream one relatively low res screen and even then over a wireless connection it felt sluggish. I couldn't use the laptop while the phone was connected, and this was its biggest handicap because I would have preferred the phone's desktop in a window similar to running a virtualized OS, with easy drag-and-drop and copy-paste.

The experience was that you had the same phone apps with the same feature set as on the phone, but they transformed into desktop layouts. For the first party apps this worked fine, but many of the third party apps didn't work at all or didn't work well, so I ended up largely being stuck with the first party apps, mostly the mobile versions of Outlook and Edge, and the file explorer. At the time these were seriously limited compared to the desktop counterparts. In that version of Edge I couldn't use many of the web apps that I tried. The outlook version was very basic, but I still managed to get some of my email done.

The apps only appeared fullscreen. I get why they did it on the weak phone hardware of the time, but this was very limiting. You could alt-tab however, and there was the windows taskbar. The start menu was the phone's home screen, so you would see the exact same tile layout as you saw on the phone, and clicking a tile would open the app. I really liked this solution because it gave a lot of flexibility while being instantly familiar.

Bottom line it wasn't really suited at being a one stop computing experience, but it was a good way to do the things you would otherwise do on the phone on a larger canvas. It was good for what it was, but it was not in any way a laptop replacement. What really killed it as a useful feature for me was what ultimately killed windows phone: the lack of a decent app catalog. I still think in its time windows phone was the best mobile OS, but the app gap meant that it never stood a chance.

I think it's more than phone hardware is still really locked down/non standardized, so there aren't a lot of viable options to get alternatives on there without also being a hardware designer and having the bank to go with that. Need a big country to force a standardization on the hardware interface/bios and an opening up similar to what we have in pc components and this will get a lot more interesting. Sadly the chance of that happening is near zero and the interests of google and apple lie in more lockdown, not less.

Nokia nailed this 2005 with Maemo actually but didn't have the grit to see it through.

Understandable as the risk of going all in on would have been a hard sell to the board vs a slow cosy death they couldn't see coming.

Grit is one possible interpretation. A bad c-suite culture is another. Same as with Kodak.

What is the point of convergence when you can sell two devices to one customer?

Because I think ultimately it is only good on paper. In practice you want your computer to be able to act on itself as well. So it is hard to get the balance right.

I do think for enterprise and business it should work though. Especially we are all on mobile already and most of the time we need to work on something that is on our phone. I wonder if there could be another blackberry moment on this? Or Something Microsoft Windows could work on, as a way to get back into Mobile. I have long argued they should have spend small R&D on possible Mobile OS just in case opportunities pops up.

Bring back Pocket PC. Tie it up with Azure and CoPilot.