This is nice and all, but what I am waiting for is the phone to power my desktop setup. I walk up, drop the phone into a dock, and my three 27" monitors all wake up with the desktop exactly as I left it.
To get there I think the GPU will have to be embedded in the monitors, and some interesting operating system hoops will have to be jumped through. Do cords have enough bandwidth now to support this? I honestly don't know the answer to this question. I feel like the answer is yes.
The other idea is just to mount the phone to a head mount display and go VR or AR. But I feel like that is going to require way more advancement in resolution on the phone screens to get right.
We can get there. We are super close. The iPad with an M1 chip probably already has the power to do it. Maybe one or two more years and this is not only feasible but purchasable. I'm a buyer.
Scan your retina and have the screens broadcast your zero latency always on desktop running in the cloud that seamlessly updates, never has to stop or reboot and backs up every single interaction you have done in the past decade.
The screens turn off and disconnect as you walk away and reconnect to your phone (now merely a pocket-screen) with all applications adapting to whatever size of the screen.
Oh and of course your earbuds have been connected the whole time broadcasting any audio that might be playing in the desktop without interruption.
Microsoft appear to be moving this way as well with the recent Windows 365 announcement [1]. I tried self-hosting my own cloud desktop a few years ago on a raspi, and was blown away by how much work I could get done remoting in from my phone or old laptop, not to mention the seamless transition when switching clients.
> This is nice and all, but what I am waiting for is the phone to power my desktop setup. I walk up, drop the phone into a dock, and my three 27" monitors all wake up with the desktop exactly as I left it.
This has been attempted multiple times. (Motorola's go at it and Windows Phone are two recent attempts that come to mind) Each time it has failed.
Reasons:
1. The upgrade cycle on phones is much shorter than on laptops/desktops.
2. Phone manufacturers want you to use their own proprietary docks, which means investing in their ecosystem, which is a non-starter for most people, especially since history shows such products will be discontinued after one, maybe two, generations. The investment just isn't worth it. (chicken and egg problem here).
3. Corporations already have fine grained infrastructure in place to manage Mac/PCs, and employees are pretty used to it. Of course your work laptop has restrictions on what you can do, and corporate VPN software, etc etc. People generally expect their phones to belong to them, they tolerate PIN policies and remote wiping, but that is about it. Newer android versions do support a dual work/personal mode, but that still isn't on the level of what exists in the Mac/PC management space.
4. Thermal limits are a thing.
5. Related to #4, pushing a phone at its limits all day long will shorten its usable lifespan. LiON batteries don't enjoy living next a hot CPU that's running at full load for hours on end.
As for cords, thunderbolt will take care of your bandwidth needs. Even USB 3.2 (I think that's the right version?) can handle dual screens, USB 4 can certainly handle it.
Is it doable with todays tech? Of course. Is it even more doable with some sort of cloud compute dumb terminal setup? Naturally, could have done that a decade+ ago.
I use DeX quite a bit - it's super handy for viewing photos/videos captured on the device (Galaxy Note 9) on a big screen with desktop like controls.
Also I have the Wacom One LCD drawing tablet that I often connect to my Galaxy Tab S6 and draw on in in DeX mode with ClipStudio. It effectively turns this setup to a nice portable drawing studio, with the added flexibility of being able to work on your projects also on the go with just the Tab S6.
Funnily enough, you can kinda already do this with a USB-C dock. I have one for my MacBook Pro, and when I plugged my phone in it launched a desktop interface that ran on my phone. It could only use one screen, though.
It's my hope the future will be more decentralized than that. Perhaps a smartphone is a portal to a dashboard coordinating a fleet of machines -- some in the cloud, some in the community, and some under your personal control in your property.
A high definition display (perhaps a projector or OLED screen) embeds a streaming device of which there are already plenty on the market. A computer with integrated game controller and advanced graphics sits charging on a dock, ready to pair and cast it's intensively, locally rendered graphics to a screen of the user's choosing. A room off to the side contains 5 or more smartscreens, ready to accept display from a personalized repository of cloud containers.
A touch-tablet hangs from a swivel arm under the cabinets above the kitchen stove, a swipe away from picking up a webpage from a smartphone or another on-network device. A UV light is embedded under the cabinet to help sanitize the tablet after cooking.
I think the intelligence gets decentralized behind APIs -- some hosted locally, some hosted through the internet. The user interfaces become more focused on context. You have various levels and kinds of input and edge intelligence based on the specific goal, all backed up by some handful of cloud applications. But a fleet like this becomes too hard to manage for an individual. So the specific deployments, allocations and budgeting is managed by an AI system that you set "moods" for through a smartphone app with some -- but limited -- advanced, specific settings available through a hamburger menu.
I just wish they’d make a monitor that doesn’t cost $5000. It’s basically impossible to find a 27”+ monitor that works optimally with a Mac because they’re all 4K, and the way macOS handles scaling means you have to scale by a non-integer factor.
Just give me the 5K iMac panel in a basic shell, that’s all I need.
If we can send 4k over a cable it also has the bandwidth to handle 3 screens at HD.
I am not sure that there is a phone that can do that though. I can plug my Android into my USB-C dock and it can display a copy of the screen on the screen, but that is about it. Doesn't even seem to work with the USB keyboard I have plugged in.
Your phone does not support USB 3.0 then (common with e.g. Xiaomi - these ports are USB-C but USB 2.0). The USB-C on my Samsung can do everything the USB-C on my PC does, including a 4K screen in dual display mode.
I would be a buyer too. I suspect the only reason it hasn't happened is that it would mean less revenue... why would I buy a laptop/desktop AND a phone then?
This rumour about Apple sticking one of their SoCs in a monitor [0] popped up the other day. Relevant here.
And their current Lightning->HDMI adapter uses an embedded chip that decodes a h264 stream of the "screen", potentially the exact same stream that would be sent via AirPlay to an AppleTV. CarPlay does something similar.
I was hoping this would be the general direction of Windows Phone. I want my big monitors at home and at the office (split time 3/5 and 2/5) and just want to carry my phone...without having to login to Windows 365...or a VM in the cloud...to do small to medium .NET development. Looks like Apple/VSC may get me there quicker.
I don’t know why you’d want that. The place where I have my big monitor is where I also have the space for a computer that has enough room to be always fast and quiet for my tasks that I can leave running things, while my phone is something that I need for communication which can help with information/scheduling and that I would like to have with me at all times. Having all my work on my phone would make me leave that phone at home and get an extra phone, I just wouldn’t trust myself with all that even if I have backups.
It’s a 400W+ machine blazing under my desk, with relatively low performance per watt and per unit of cost and high depreciation per hour of usage. With 6 ms latency to my major IX and the stability and backup professionalism cloud providers can offer I’d be all in for online desktop. Perhaps I’d invest in a bit more beefy Synology as an internet-down local machine, but my phone or iPad could offer that functionality as well. Going from full machines to terminal-screens would also save my household of 5 many kgs of electronics. That times a few hundred million households sure adds up.
I got a GPD Micro PC a couple months ago and I'm in love.
It's a magical size - exactly small enough to slip into a back pocket like a chunky phone, and exactly large enough that it's actually practical to use desktop software on it. It's ergonomically first-rate - I've done CAD work on it. And it's durable too - I've dropped it several times already and it's barely even scratched.
It's also unlocked a class of activity I didn't know was possible - long-form writing while out walking. I can get 50wpm with great tactile feedback on the tiny keyboard. It's all in on thumb typing, and having struggled with a Cosmo Communicator for a year I can absolutely say that that is the correct decision for a mobile device. The vast majority of the time there's no handy surface nearby.
Now that I essentially have a laptop on me at all times, phones hold no appeal for me anymore. None. This is the class of device I've always wanted. All it really needs in order to replace a phone completely is a cell modem which can wake it from suspend, and ideally the standard "mobile" sensors such as GPS, cameras, and gyros. And waterproofing, I guess.
How's the battery life? Can you replace the battery if/when it gets old?
That looks like something I could buy. I've really wanted a tiny console PC with RS232 and ethernet for a long time now. I thought about building one into a custom keyboard using something like a raspberry pi but I haven't had the time to follow through. This thing looks like it could make my diy project unnecessary.
I also haven't found phones particularly appealing.. mine gets 5-6 days on a charge because it's mostly just an alarm clock. Any time I try to do anything with it, I get frustrated by how clumsy it is compared to an actual computer. I do like being able to look at maps or make a call in an emergency, and it's handy having a camera always in the pocket, but that's pretty much where the appeal ends.
Battery life is a solid 7 to 8 hours with real-world, screen-on usage, provided you're not compiling software or pushing the thing really hard. I've turned the TDP down on mine because I almost always prefer it to be cool rather than fast. You could probably drain the battery in a 2 or 3 hours if you really went for it.
Battery replacement is a matter of unscrewing the case. It's glued to the interior, but I've seen teardown videos of people just peeling it off so it's not a big deal. In lieu of hot-swapping batteries, I sometimes carry a USB-C PD power bank - although the Micro PC is low power enough that it will actually charge off a regular phone charger.
There were lots and lots of problems with the Cosmo as a phone, build quality first and foremost, but I'll focus on the screen/keyboard combo since that's the point of comparison with the GPD.
First, the keyboard: The Cosmo keyboard is not designed for thumb typing. You can do it, but it's a bit awkward. The phone is not comfortable to hold in this configuration - it's thin, wide, top heavy, and the hinge has sharp edges. The generous key travel also works against you when thumb typing. It seems that you are basically always meant to put the phone on a flat surface and touch-type - but it's not very good at that either. It has no feet on the (slightly curved!) bottom surface, and while the rear hinge is meant to act as a kind of kickstand, the design means that force of typing transfers to the screen, causing it to wobble with every keystroke. The keys themselves are as nice as is possible in that size, but touch-typing at a table while on the go is a very narrow use case. The device feels delicate and fragile. Also - while this isn't a completely fair comparison as I've only had the GPD for two months - after a year the keyboard began missing keystrokes (and this was a replacement keyboard, since the one it came with did that from day 1).
Next the screen - a super-high aspect ratio smartphone screen turned landscape is actually really annoying. You have almost no vertical real estate, whereas most phone apps assume that the vertical axis is where all the real estate is and consequently put menu bars at the top and the bottom, eroding it further. There's very little gap between the touchscreen and the keyboard, so accidental touches are common. Switching between touchscreen and keyboard interaction is even more annoying that switching between mouse and keyboard on a desktop, either because you have to transfer the weight of the device to one hand if you're holding it, or if it's on a table because you must gently prod at the screen without tipping the thing over, or brace the screen from behind. Also the screen only opens to one very wide angle (cheerfully described by Planet as "the optimum angle"). There's volume buttons, but they're awkwardly on the back of the screen, which would make sense if they worked when the device is closed, which they DON'T.
Now, the GPD Micro PC: no touchscreen, which is actually great because all interactions that change the state of the device require definite action and make a physical "click" - you don't have to worry about accidentally brushing it the wrong way. This actually takes a lot of cognitive load off. It's comfortable to hold in two hands for thumb typing (but also has rubber feet so you can set it down). The touchpad is accessible with a fractional movement of the thumb, so switching time is minimized. There's dedicated and easily accessible volume buttons. The screen has a comfortable aspect ratio and can be set to any angle, where it remains stiffly. And most importantly, it runs desktop software. With these controls, it's more ergonomic to use desktop software in miniature than it is to poke about at a touchscreen.
Thanks indeed for the review! I came across these devices once, and I always wondered if it was more gadget-like, or really useful. Have you tried running Ubuntu MATE on it?
I was never brave enough to tinker with GNU/Linux on the Cosmo while it was my daily driver. I'm pretty sure that using the device that way comes with some pretty sever compromises.
On the Micro PC, I run Void Linux with KDE Plasma. I didn't do anything special, everything's mainline, so I'm sure at a technical level Ubuntu MATE would work perfectly well. The only interface tweaks I've made are to set the task manager to "icons only" to save space and added thermal/RAM/CPU monitors in the taskbar.
As for "toy gadget" vs "useful tool", I'd say that if you're the kind of person who likes to take your laptop everywhere, it's definitely for you.
Thanks. I meant on the Micro PC. I guess you already mentioned KDE Plasma somewhere. I definitely don't take my laptop everywhere, but I'm contemplating. Such a Micro PC looks very promising, exactly for the reasons you stated!
Quad core and 8gb of RAM is underpowered? It's not the best gaming machine to be sure, and I have to be a little more careful with my browser tabs than on my 64GB workstation... but there's not really anything glaring it outright can't do.
Hard to answer what I do on it, since the answer is "everything". I run Plasma 5 on mine. Right now I have a lot of firefox windows open, Quassel (IRC), Qalculate, some file managers, some terminals, and some text editors. For heavy duty I sometimes remote into my workstation (graphically) which is fast enough on a LAN to use CAD and all the rest of it. I've even done ANSYS simulations this way.
It was around 2000, after buying a Cassiopeia E-100 that I realized a keyboard makes any computer massively more usable. With that I bought a Psion 5MX, used it great success, and never considered a device without a physical keyboard again. Switched off to Nokia Communicators, then several Android Sliders. All worked great. I was always the guy in the conference room who could login to servers and look-up info, despite everyone else having phones, too.
A (pocket-sized) bluetooth keyboard sounds fine in theory, but it's a hassle to keep it charged and to carry around a bulky extra device you'll only use occasionally. So it'll get left at home, not there when you need it, and nearly worthless.
It's only now, with every US carrier switching to VoLTE-only that I'll soon have to retire the old Android sliders and find a new option. Everything seems an unfortunate compromise. The keyboard on the Blackberry Keyone/Key2 isn't great, missing critically important keys like arrows and OK, and without being rootable, you can't remap them to make it usable. The form factor is... awkward as well.
The Gemini PDA is likely to be my next stop. Unfortunately the form factor is kind of anti-phone, making it a poor choice to quickly look at your screen to check some info, or make phone calls, and only good when you really need to do a lot of typing.
It's a shame and a contradiction how the smartphone market has ballooned for the past few years yet choice is smaller and smaller.
How can it be that there are 100s of smartphone models out every year yet every single one has the same form factor? You cannot even get a phone smaller than the phablets of 6 or 7 years ago, let alone a phone with physical keys...
The technology is so complicated that costs in design/patents/royalties/manufacturing make it unfeasible to produce anything without economies of scale. Coupled with insufficient number of buyers willing to pay a lot, the potential profit margins are so low it simply is not worth the risk reward ratio.
Before buying anything from Planet, see how well they have done at maintaining older devices. Latest release for Gemini PDA is Android 8.1, three years ago.
I'm sure all things are possible. I just felt like it was indicitive of a feeling like they abandoned the device soon after launch to work on the next shiny thing.
Yeah true, I had that with the gdp pocket which I used a lot. I can replace and update the software (Linux in my case), but hardware support was dropped very fast. Since then I am a bit more careful. I like real open projects like the Pandora or the Pyra as hardware support is there for so long (more than a decade for the Pandora) that it keeps going with small replacements. However, those devices are just not as high spec or practical. Open source hardware and software but with great specs and formfactor are far away it seems.
Ah my previous planet device was stolen a few weeks after I received it. I hardly had a chance of using it.
The Slide looks more like a usable phone indeed. I never tried Linux on the previous one (which was my plan); with Linux I guess the 5g doesn't work? Anyone knows if wifi works with Linux on the Gemini or what works and what doesn't? For me to have a usable device, it would need to run Linux for somme things; I'm fine with Android for all phone user, but Linux with Wifi would be needed to do anything useful besides that.
If there would are (closed source) drivers for the 5g chip for Linux then this would be the 3rd new Linux phone on the market?
I have an alternate solution to propose to you that I used for a couple of years quite happily:
- Smart Watch for Voice/Text
- small Tablet with keyboard cover and LTE in your pocket for everything else
These days I have a different situation where I prefer not to have anything if I’m away from computer and I just use the Watch. I have the Apple Watch / AirPods which works quite well, but I know not everybody wants to commit to the Apple ecosystem. When I was doing the tablet-in-pocket thing, I was a bit more cross platform and it still worked great with Google Voice as the central hub for everything.
My experience with some early Android sliders (HTC Dream and HTC Desire Z) is that the keyboards don't have all the keys I need. So in the end, typing the usual array of Unix and C graphics characters was a bit of a hassle.
For years now, I have used the Android "Hacker's Keyboard" on lots of ssh sessions.
Though I guess I would take a slider if it had a full (5 row) set of keys.
> My experience with some early Android sliders (HTC Dream and HTC Desire Z) is that the keyboards don't have all the keys I need.
That's what the Sym button is for. Pop-up a menu on screen showing other, less-frequently used keys. Several apps included similar workarounds.
The LG F3Q did a good job with their keyboard, only omitting tab and pipe, but it was possible to remap keys to get those.
But then again, I'm from the dumb terminal days... I'm used to systems which are missing common keys, and using combos (e.g. Ctrl+i for tab) instead. Doing that on a physical keyboard was infinitely more productive than an on-screen keyboard, even Hacker Keyboard.
For a really long time, I never left town without my laptop, even if it was a pleasure trip. I have lots of stuff there I might want on a personal level, plus for the last 14 years I've been in a broad role in a small software company that kind of requires it. (I mean, it doesn't, and I'm not griping, but it's easier for everyone else if I limit the amount of time I'm truly unreachable, so I pick and choose.)
The incredible growth in on-device capabilities from my iPad + a keyboard went a long way towards making that unnecessary; adding a remote desktop client (for access to the back-end Windows server things I might want) closes the loop.
Unless I'm dead sure I'll NEED a full computer on a trip, I don't take one now. The default has become "nope." Even -- especially! -- as a photo management platform, because my iPad has ample storage and I'm a Creative Cloud user. I used to have to carry the laptop to dump and process photos during the trip (to avoid dealing with 2,000 all at once when we got home). Now I can go day by day, sync to Adobe, cull and do some processing, and just take the final pass once I'm home.
That's great.
(Ironically, this same era has seen me become increasingly deep into Orgmode, which is annoying, because as great as it is on a desktop there's still no truly great way to work with it from mobile short of setting up sync to a Linux box and using mosh/ssh.)
I’m curious why you prefer the iPad and keyboard over a laptop since adding the keyboard typically makes the iPad both bigger and more unwieldy than a laptop.
I ask because I also like the vague idea of not taking a laptop, but I’m not going to do any serious typing on a screen keyboard and adding an external keyboard makes the iPad less portable than a laptop (plus you probably need a stand and mouse as well to reach the physical usability of a laptop).
I'm the same way. My iPad is smaller, lighter, has cellular so I can get in generally everywhere. The battery (before my M1 Macbook Pro) lasted much longer, and didn't put out heat. It also had my movies for the plane. So it was killing two birds with one stone.
I dunno what laptop you use, but my iPad Pro + keyboard is materially smaller and lighter than my full-sized 15" laptop.
"plus you probably need a stand and mouse as well to reach the physical usability of a laptop"
Apple was first, but there are now SEVERAL keyboard cases that function more or less like Apple's fancy & expensive "Magic Keyboard." Like Apple's, they usually include a touchpad.
My iPad Pro in my Magic Keyboard folio is not much bigger than the iPad itself, but opens up into a very very usable and stable configuration. The hinge holds the iPad above the keyboard, so it's at a more comfortable heigh. I can easily use it on my lap, for example, which is typically a use case that tablet-keyboard combinations fail at.
(The only downside of this style of keyboard case is that you can't fold it all the way around; to work with the iPad in portrait mode, you take it off the keyboard -- which is also simple, since it's just magnets.)
IOW, the "physical usability of the laptop" threshold is absolutely reachable with a single accessory.
Also, even though I have a fancy iPad and fancy keyboard case, my iPad solution is materially less expensive than my laptop, so taking it places represents less of a risk. And obviously it almost never needs to be plugged in. (This gap is narrowed for M1 hardware, but I'm still on Intel.)
Similar to what I’ve been doing lately. Bought an M1 iPad (different story…) and have been using MOSH and Blink Shell to access my computer at home running emacs. Has been a blast doing journaling, logging my time in org-mode, writing notes in org-roam for uni subjects. Has been a game-changer for me. I know it’s just a ‘dumb terminal’ but the experience is so seamless it’s incredible. MOSH makes sure the connection doesn’t drop out when I lock the device and come back half a day later. I can be on the train home and fire up that buffer I had going this morning writing a paper in LaTeX, with proper syntax highlighting, completion, snippets… incredible.
This is exactly my setup, and I love it. Seriously, gentle readers: if you haven't checked out MOSH yet, do so. It's game changing:
* It doesn't care if my IP changes. For example, I can start a session at home connected to Wi-Fi, switch to tethering to my phone as I take a bus, switch to office Wi-Fi when I get to work, and the connection never drops.
* It stays connected forever. Switch to a Blink tab with a connection I haven't used for a week, and I'm instantly looking at a live shell prompt.
It's magical. I use MOSH everywhere I possibly can now.
I also use Blink with Mosh, but what’s your physical setup? You mentioned using it on the train, so are you using an attachable keyboard like the Magic Keyboard? At that point, isn’t the iPad just bigger and heavier than a laptop?
I don’t ask to refute your experience; I’m honestly curious since I like the idea of just using the iPad, but the ergonomics just seem inherently inferior regardless of software concerns.
It’s heavier but not bigger. It’s not ideal but I like that it is more versatile (for me). I can do some coding/org stuff, then detach from the magic kb and start doing practice problems from a textbook (engineering drawings etc). Laptop will always be better if I know I won’t want to do any pencil stuff. Otherwise, iPad is compact and modular, and I feel more like bringing it with me places. I like to leave my laptop ‘docked’ and use it as a desktop often.
I sometimes sit my HHKB on top of the magic keyboard, but not often. If I’m going out I just use the magic kb and trackpad.
Not the person you were asking, but I have a Smart Keyboard Folio which is super thin and light, and that folds back and out of the way so I can use my iPad like a normal, bare iPad when I want to.
What is the UX for switching between apps or moving the screen like?
Can you do this with keyboard shortcuts or do you find yourself switching between keyboard and touch screen for some interactions?
I find this idea of using a keyboard super appealing but don’t like the idea of switching between keyboard and screen all the time. At least a virtual keyboard is on the screen!
I often use an external keyboard with my iPad (but not really on the go). Both multi-tasking and keyboard shortcut support have gotten a lot better recently (with an especially big push in iPadOS 15), but I find the biggest blocker that iPadOS still hasn’t been decided with a keyboard in mind.
For example, tons of apps—including Apple apps—don’t highlight text fields by default as they would on a computer, so you have to tap (or click) the field before you can type. You can very rarely tab to switch focus between different buttons and menus. You can’t type to jump to an item in a list, etc.
MacOS has so many little productivity enhancers, especially in regards to keyboard shortcuts and keyboard control, and the people designing iOS/iPadOS have just never made it a priority so even with an external keyboard, I find myself having to use touch or the mouse so much more often.
You can CMD-TAB to other apps as on macOS. There are many of the same keyboard shortcuts translated to iPadOS, even some emacs movement keys in all text boxes!
There isn’t enough room to have more than one thing open usually, so lots of CMD-TAB action.
Normally I’m doing one or the other - typing or annotating/drawing with the pencil. If I need touch-interaction when typing I can use the trackpad.
I can’t say it’s a perfect experience, or even a practical one, but I still enjoy it. Maybe it’s just the novelty of the setup. It weighs more (with the keyboard-case) than my MacBook Air, but it does have a nicer screen and it is waterproof. It lends itself more to portability for sure. I tend to just chuck it in my bag more often than I did with my laptop.
I'm doing something similar: PC at home, Chromebook with Crostini, running ZeroTier and nss-mdns with mosh on top or ssh for X11 forwarding (I use git gui a lot). Cheap fast and can work on battery forever.
I've been saying this for ages — the future setup is going to be phones as a singular device. When you get home, you'll dock it to be able to use your peripherals or extra monitors. Phones will become even more powerful, but even then anything heavy can be done in the cloud or at the network edge (with minimal latency thanks to 5g, see cloud gaming). OSs will become more convergent (Windows tried it, Huawei's doing it, PureOS if you're into Linux) but that won't even matter as the browser will be the de facto OS if it isn't already. Responsive web apps (through PWAs) will get native-level capabilities with the browser as a compatibility layer, and traditional OSs will simply be browser-running machines.
To be able to dock it you need a screen and keyboard (or a laptop-style "shell"). That means you need a device that takes just as much space as an independent device but can't be used unless you plug your phone to it. The only benefit is a slight lower cost - but then you need to make sure your phone is powerful enough so a cheap Android is out of the picture.
You've been saying this for ages, and manufacturers have been trying it for ages (Motorola Atrix, Samsung Dex...) but it doesn't catch on. So maybe you should reconsider your prediction!
> you need a device that takes just as much space as an independent device but can't be used unless you plug your phone to it
Flip side: it’s a device that just works, doesn’t need to be kept patched and updated and where syncing is never an issue. It’s an external monitor for your phone plus a keyboard.
The prediction has fallen on its face due to the mobile component not being central enough. If half my life is on my laptop and half on my phone, a desktop makes sense. (Though I go for external monitors to my laptop.) But if all my stuff is on my phone, losing the syncing tax becomes more meaningful.
Now that I'm using a Surface Book 2 as my primary computer, with a Surface Dock set up on my desk at home with screen/keyboard/speakers/etc., I'm more confident than ever that this future is coming eventually - but also confident that it will take a long time. Having a single device makes a surprisingly big difference from having two separate devices (home PC and travel laptop), but is also not worth sacrificing power for. The technical users who will become early adopters want a first-class home PC, so just as laptops only started to actually displace desktops once they had enough RAM etc. to match them, it'll be the same for phones, and they'll probably need to be x86-based phones for the sake of being able to run old programs.
Apple already made it possible on macOS to run x86_64 software on their ARM computers, with Rosetta 2.
Any company with enough resources could do the same if they felt so inclined.
And if Apple wanted to, Apple could make Rosetta 2 run on iOS devices. For now they want to keep them mostly separate, and are more about running iPad apps on macOS than the other way around. But there is nothing stopping Apple from running desktop software, whether compiled for x86_64 or for ARM, on iOS devices if they wanted to. And if convergence turns out to be the future then I think we may see them do that eventually. But I am not convinced that it will.
> they'll probably need to be x86-based phones for the sake of being able to run old programs.
Maybe its the other way around?
Maybe we need to wait until the old way of doing things is so forgotten we can accept the phone-way as a valid way. As long has people have their x86 applications, they'll want to use them. Once people stop caring - well then there is an app for that.
Years ago I had a similar thought to parent, only without the dock. I figured phones would have micro projectors to throw a display on any convenient wall and a laser keyboard[0] to throw onto a table.
These days I no longer think that'll happen largely because the kinds of things people want a desktop computer for are really much better served by an actual computer, while mobile users have adapted to not even wanting a desk.
We are already there with laptops and laptop docks. It could certainly be a cellphone dock in the future but the phones just aren't powerful enough now. Maybe an iPhone with an M2 or M3?
1) This would be tolerable with some PinePhone successor, but proprietary solutions like Samsung DEX are conducive to lock-in, more difficult for developers to target, and tend to go unsupported after a few years.
2) Phones have to become more durable. I expect my primary computing device to last half a decade. My phone's USB-C port is flaky after just over 2 years of charging it and carrying it around. How is it going to last if I'm plugging and unplugging it to/from some desktop rig?
3) They have to become more repairable too. "If it's broken, throw it away and buy new" just isn't tenable for a primary PC.
4) Getting everybody on board with "live in the pod, eat the bugs, use the cloud-connected thin client" is not going to happen.
Concerning 2) magnetic usb-c connectors are nice and wireless is an option, e.g. see the Lenovo wigig dock for an implementation.
Concerning 3) I think things are moving in the other direction. With immutable base OS and data seamlessy backed up (encrypted, of course) switching to a new device should be fairly frustration free.
For 4) I think we will see more and more of a mixture. For instance better integration of cloud storage, things like game streaming, more video streaming etc. . So not thin client, but not fully featured offline either.
Hell no. Real techie software is basically a third-class citizen on smartphones at this point, with the promise of being locked out even more. The cloud is not a solution to everyone's problems, thanks to what it can or cannot do (technical problems), and thanks to the fact that some just don't trust it or don't want a recurring subscription (socioeconomic factors). If you ever want to upgrade the hardware, you'll also need somewhere to install it, and dongles won't suffice.
Some people may be happy with this future, much like people who create and process lightweight office documents on iPads. But there will always be demand for singular PCs or laptops.
There's also still a market for horse drawn carriages. I don't meant that flippantly, but that the advancements in technology make things like the Novena laptop a possibility. It sounds like something like the GPD Micro PC is much more your speed.
> anything heavy can be done in the cloud or at the network edge (with minimal latency thanks to 5g, see cloud gaming)
It's interesting that people often handwave latency like this as if it was a solved problem. As instructed, I have seen cloud gaming: it's an unmitigated disaster, held back by mere trifles such the limits of information travel speed imposed by our physical universe.
Even if it did, that's only talking about latency pushing the image down. I wouldn't expect input latency to ever improve to the point that most gamers play exclusively via cloud.
I'm not convinced. Not because it's technically unviable, but because I know that when companies like Google are involved it'd be far too restrictive for my taste. And as soon as you also want some local computing power (even if just for the situations it's offline) it's more expensive than hardware in a larger format like laptops - even more so when you need an upgrade, because phones can't be upgraded (or even repaired in some cases).
> the future setup is going to be phones as a singular device
I had a four-day weekend where the only device I had was my Android phone. I needed to work on a release of our software, so I had to fix a few bugs, write a press release, get content ready for the website. On the way out, I grabbed the laptop bag, but forgot the laptop. On some of the higher-end android devices, if you plug them into a monitor you get desktop mode. So down to the hotel "business center" I went with my LG Wing and plugged in a monitor, keyboard and mouse using the usb-c dongle I usually use as a hub for my laptop. A lot of the apps played nice, so email (Fairemail is a fantastic piece of software), word processing (Google Docs App), editing screenshots (PicSayPro) and SSH(Connectbot) were all fantastic. Could Android desktop mode use a little polish? Sure. For any serious development, I'd prefer to edit and test on localhost... but honestly, the experience wasn't bad, and from time to time I'll go phone-only to the office if I know I'm not going to be coding all day.
I don't think it's going down that path. We're seeing people use phones more and more, and a lot of people don't seem to need a computer for most tasks. Why buy a dock plus keyboard plus mouse plus monitor when you can just use your phone?
People that do need a computer-like device probably won't have enough with a phone, and will need something more powerful than browser-based applications. Not to mention that a lot applications would be far more expensive if moved to the cloud (imagine a video editor where you not only pay for the software but also for the storage + computing power to the provider).
Samsung Dex is pretty usable IMO. Unfortunately phone hardware seems to be just a bit lacking when you try to do desktop-like things. My galaxy S10 can keep up for a short time, but having lots of browser tabs open, a vs code instance, whatsapp open in the background can degrade performance after a few minutes when the phone starts getting hot.
It seems to me the CPU/GPU can burst pretty well (maybe approaching ultrabook-style laptops) but sustained performance is just not there yet.
I am using old Android Xperia 10 with Termux, Apple Bluetooth keyboard and Chromcast screen to 27 inch TV. With Termux I will ssh into my server (using mosh).
Pretty decent and stable setup… And free as I have just gathered some old hardware pieces…
Have you tired the Logitech K480? It met all my needs for pairing to my phone and an Android tablet with a nifty rotary switch for changing inputs easily. They keycap shape was lackluster but overall it was a novel solution.
The biggest problem with using your phone as a computer is not the screen size, or app availability or anything else you might commonly think of when considering arguing about this subject.
In my opinion, today it is one and only one thing - the battery. I'm presently typing this comment on a desktop, because my phone is on the charger.
That's not consistent between phones. I have a low end Redmi 9T with a 6AH battery. It lasts 2 days of relatively heavy phone usage. I'm guessing it would last a whole day in a Termux shell.
Edit: if anything the real issue is thermals. I also have a high end phone with a Snapdragon 888. This thing has a faster CPU & GPU than my Intel desktop. However, it's only fast for seconds, until it gets thermally throttled.
I don't see why this is an issue. If you connect by usb-c that also charges. If you want it wirelessly you can still lie it down on an inductive charging pad, right? Or would that interfere with the connection?
I've been looking around at old sharp zaurus organizers and psion mx5 devices for the use case of offline portable word processor that can save text files to eventually go to PC via CF / SD card. for this years nanowrimo.
Most psions are 300 bucks, i can't justify that though i love the idea. And the 15 buck zaurus organizers either only have a 30 character 'memo' function or are so old they only run DOS and have to sync via serial cable.
Anyone have any suggestions on a palmtopish machine with qwerty keyboard that can write txt or rtf files to an sd card and can be had for 100 or less on ebay?
I'm currently resisting trying to hack a Smart Response XE and instead trying to put a pi zero w and a 1.3" inch screen and battery into the front margin of a Unicomp Model M and just viming my way to victory whilst syncthing makes the file magically appear on my pc whenever the device is on home wifi, but an integrated device would be more pocketable. the model m word processor is more for taking over long writing sessions where i currently lug an actual wheelwriter around. this small device would be even more portable, sitting at the beach, in the car etc.
I think a lot about PinePhone, PinePhone Keyboard, terminal and emacs org-mode. As a journal, to-do list, notes etc. I'm very tempted to buy them, when the keyboard is released.
But I have a Gemini PDA which is basically the same thing but more integrated and I never use it.
I did write a blog post or two on it on the train using termux and nano. That was cool. And the occasional SSH session.
But as a phone running Android it's not quite there. Features like the LED controller app are half-baked vanity apps, and I'd rather they'd spent the time getting the lower level stuff right. Low quality phone call audio, random glitches, and software updates were mostly forgotten about. If I did want it as my primary phone I would also want a camera. The only camera was front-facing and low quality, which ruled it out. So it became a novel supplementary device rather than a daily driver.
To whit, the hinge mechanism isn't well designed. In theory the spine rotates back to form a stand, but it's incredibly wobbly. So it's not much use for typing on a table. And it can't be held in the hand, as there's no stiffness.
Why do I think PinePhone will be any different? The prototypes I've seen look like the hinge is stiffer. I'm not a kernel-compiling Linux enthusiast, so I don't expect to use a PinePhone as a real phone. But I do hope an SSH terminal to be simple enough to just work. So that puts it into the realm of mostly offline with sync. And org-mode is both simple and exciting, in a weird kind of a way.
And that appeals to nostaliga for Psions, Palm PDAs and Newtons.
Do you know how viable (if even possible) it is to run Linux on it? There are some notes for Kali[0] and Debian[[1]. I've come across comments that it's far from ideal due to Halium but I'm yet to understand what that means in practice.
I do this too! I have a Logitech MX Keys, and while trying to separate work and personal devices I put my phone on my laptop stand and with 1 button I can switch the MX Keys to the iPhone and Signal/Telegram/Whatsapp/Email/Termius away.
It works for more than I would have thought but I still really miss things like alt-tab, or in general the ability to navigate the UI with a keyboard. I mean why would the win/super key not open the multitasking overview? Currently it does nothing... Also, in Termius backspace works, but not delete, and page-up/down produces a ~. Little annoyances like that are everywhere.
I also bought an MX Anywhere 3 to go with it, but it didn't switch device with the MX Master switches, a missed opportunity imho. Moreover, at the time I was still on Android and for example in MineCraft the mouse pointer would still behave as a finger that drags stuff. (I eventually returned it because I also found it too flat for a mouse.)
So there is still much be made better but the basis is there. I guess the demand is still very low. But I bet it is also a chicken/egg story waiting to be resolved. I mean, iMovie on the iPhone is super powerful but on this screen and with touch... ugh.
Now just get me a monitor that also switches with my MX Master Keys switch keys to a larger view of my Phone and we've got ourselves a (partial because I love Linux, and yeah I was so looking forward to that Ubuntu Phone Convergence thing) laptop killer.
Man, a personal hosted VPN for all my junk is my dream. Can't wait to get something selfhosted, tail-scail-like setup. Something i can get into from anywhere and work from anything. I'm tempted to order a bluetooth keyboard like this post just to have it in the car and be able to work from my iPad/iPhone in a pinch (assuming internet).
For someone wanting an easy, at home VPN for setups just like in the blog - is anything on the market competitive with Tailscale's UX? Ideally there would be no proxy, no VPN hosted on DigitalOcean that i fear being a weak point.. Instead, i'd like:
1. A redirect hosted on DigitalOcean, acting as a self hosted DynIP. No security issue here i'd think?
2. After hitting my real IP, connect to a VPN on a predetermined port.
3. Get access to bells and whistles now on the priveledged network. Bonus points if i could assign DNS entries like `ssh me@workmachine.fake` or `http(s?)://videostreaming.fake`
Tailscale makes me a bit nervous, as cool as they are i'd prefer entirely self hosted. Though i may give them a try just to experience this UX.
For the last decade or so I have been running cheap Asus routers on my home network. They support dyndns and OpenVPN out of the box. I assume that other manufacturers offer the same thing.
I haven't poked around in the guts of their vpn setup. But it's generally password protected and I would assume in the clear. So I replace accounts frequently and only leave it running if I am out of town and have some project I may need access to.
Basically you vpn to yourname.dyndnsprovider.com and you are done.
The rest is DNS Setup which.is sort of up to you but you can easily configure to something like machinename.home.mydomain.net
In 2013 Mark O’Connor wrote a series of articles about his experiments with nomadic computing using ipads, Linodes, and bluetooth keyboards. He was kind of a pioneer in this area:
Logitech has a few models with a builtin phone stand, too, so you can prop up the phone or tablet as you type. They all have number pads, though, instead of the compact layout.
Update: If you click to the "Tablet Keyboards" section instead of the regular "Keyboards" section, there's the K480 which is a compact model with phone stand: https://www.logitech.com/en-us/mobile/keyboards Why that's not listed in the regular keyboard section remains a mystery...
Do mouse cursors work on iPhones yet or just iPad OS? Because there's other models that have a trackpad on the side, which could be neat. Ah, there's an accessibility setting to enable it, but it's not the same as the iPad OS mouse support.
This is neat, but both Android and iPhone are really poor choices if you're looking to travel with a computer. Particularly on the iPhone side of things, you're basically screwed if you lose your network connection, since you aren't going to be doing many interesting things with the local shell. Android is a little better, but you'll still end up SSHing into another machine to get any valuable work done.
Good, cheap x86 laptops are like $80 right now. That's about how much you'd pay for one of those fancy Logitech keyboards, but it comes with an entire computer, too.
Agree that the freewrite looks a little massive compared to the OP's travel-oriented use-case here, and that getting a bluetooth keyboard paired to something is a great use of available items.
There is alse a Logitech keyboard (Multidevice K380, under $50 apparently) with a built-in notch in which to place a phone or tablet.[1]
How feasible is to split/use compute power of multiple devices?
I would love to travel with a 15" ultraportble which is good for mail/browsing but will crawls while trying to play Flight Simulator. When I reach home, it seamlessly uses power of 599$ XBox compute centre to run it on 60FPS. When I reach office it compiles millions line of code using power of a 500 core office server.
Central server can be shared across devices so all devices of all members of family run at same speed/capability.
You can buy an e-GPU, if your device has thunderbolt and isn't an M1. Compatibility isn't great and it is obviously more expensive, but it is more or less what you suggest, with the limit that you can only use it with one computer at a time.
I’m a big believer in this idea; so much so that I have a Bluetooth keyboard for my Boox Note 2 e-ink tablet. Very writing, reading books and note taking via the stylus focused, though e-ink refresh even on a device like this might be too slow for some (depending on what refresh setting you choose), it’s entirely replaced the need for paper in my life — and for travelling at least around Queensland it’s been brilliant.
For me, a folding bluetooth keyboard was what made the difference. It really tilts my default time-to-fill behavior from consuming towards producing. I'd also recommend having a ring on the back of your phone, so you can prop it at a good viewing angle.
I've been writing with obsidian, coding via ssh. Jupyter notebooks and other browser-based environments are also good, though much smoother with a tablet-sized screen.
Uber had trouble with their app being hard to manage deployments for large for sanctioned downloads over cellular. I suggested they opt for a thin client approach. It can't work they said because it requires the internet to work. I'm unemployable in tech because of good reasons.
The Remarkable 2 runs Android and has a USB-C port so it might support a wired keyboard.
Although I’m not sure if the screen refresh and cpu are fast enough for comfortable typing. I already get frustrated using the onscreen keyboard for typing a 10 character notebook name.
You can SSH into it in order to add software to enable an external keyboard. But that's not the point of the device, which is that it's a digital pad of paper, and all the freedom from trying to draw perfect drawings in Powerpoint/Visio/Whimsical.com using a mouse.
The problem with the device is that, entirely unsuprisingly, it turns out I can type way faster than I can write with a stylus after years of typing. But I really like my Remarkable2, I just haven't figured out how to fit it into my life - I've spent so long wanting a device like that, that I've got all these other things in place that I don't seem to need it. I journal using google docs, and the Remarkable2 doesn't have any integration there. I use it for quick notes, drawing diagrams and also learning to draw still life.
I'm very happy with my purchase, but it is very much a niche product. I didn't buy one to get an eink terminal, so I haven't tried it out as such, and can't report on how well it does as one, but I don't think it would do well at scrolling pages like you would a terminal. (But I'm a fan of Alacritty on OS X, so my demands on a terminal are also niche.)
It doesn't run Android. It runs Linux. There is already similar devices with front lit eInk screens and support keyboards out of the box too with Bluetooth or USB C. The Boox Note Air is an example. It's considerably more powerful than the Remarkable 2 and in the A2 refresh mode typing with a keyboard is not too bad.
I tasted this with the Display Dock for Lumia phone. My initial setup was USB/Wifi KB + miracast + android phone. And it worked just as well. Overseas phones seem to have more capabilities than what we get stateside.
Unfortunately Logitech k480, as good a concept it might be, has a very mediocre button feel and also feels very bulky . That was probably on purpose, so that it can support the weight of a tablet. On the bright side it can be found for pretty cheap now.
There must be generation coming which has never used physical keyboard and do not ever want to. Even now, when my home automation needs some adjustment, it easier just to SSH to Emacs than to wake up, put some clothes on and go sitting by the screen. Especially with (global-set-key (kbd "å") 'hippie-expand) .
Wait until they pass 35 and find that getting up and moving about is not just an annoyance but vital to keeping your body away from a state of constant little nagging pains ;)
I can type with my thumbs but it's far slower than using my left thumb and right index finger. I am almost 50.
However, what does that have to do with 2 handed use? Thumbs typing is also 2 hands. Or you mean typing with 1 thumb? I see people, especially Chinese, writing on their phones with 1 thumb and holding it with the same hand (so 1 hand use); you have that too with swipe keyboards though for western languages. This us definitely not faster (or, for me, nicer) than just typing with 1 thumb and 1 finger.
Curious what other people do... I am on a tetras on a busy street now and wasting what (very) young people do: I see them all typing with 2 hands, two thumbs.
Not sure how that works with a lot of noise and people (subway, busy streets, airplanes etc); it seems to type out complete crap random people around me are shouting. So in reality touchtyping is faster.
To get there I think the GPU will have to be embedded in the monitors, and some interesting operating system hoops will have to be jumped through. Do cords have enough bandwidth now to support this? I honestly don't know the answer to this question. I feel like the answer is yes.
The other idea is just to mount the phone to a head mount display and go VR or AR. But I feel like that is going to require way more advancement in resolution on the phone screens to get right.
We can get there. We are super close. The iPad with an M1 chip probably already has the power to do it. Maybe one or two more years and this is not only feasible but purchasable. I'm a buyer.