Pinetab – 10.1″ Linux Tablet with Detached Backlit Keyboard
store.pine64.org
store.pine64.org
One typically works.
List: https://en.wikipedia.org/wiki/Archive.today
Cloudflare is in a grudge match with archive.today:
https://community.cloudflare.com/t/archive-is-error-1001/182...
As a portable SSH machine, a kid's first computer, or a nice portable machine with a screen for hobby projects, I think this will be fantastic.
This saddens me...
I wish there was a lite version of the Web with enforced resource constraints
There are lots of reasonable use cases for JavaScript. Think of something purely event driven like sorting a table by clicking on columns or an input field with autocomplete where you see a preview of results as you type.
8GB RAM, 128GB MMC, USB-C for up to 300$ and it will be the perfect Microsoft Surface Go's contender I'm waiting for!
Why MMC when there is NVMe?
I just wish it was easier to try out git kernels (ones with the prototype patches for video decode) on ARM devices. Having to deal with hacking up device tree files is a bit daunting. It's a good learning experience nonetheless.
It's nice to be able to collect random patches for things and apply them to your kernel and pretty much being good to go on x86. With ARM you have to ensure you instantiated the driver in the device tree file which gets stuck in the boot partition along with the kernel .
Meanwhile with linux your memory needs are generally modest.
Until I had to work with Angular apps. Then I suddenly needed twice the cores, twice the RAM. I was honestly surprised.
Looking at the resulting code, I also have no trouble believing this cost is also brought on to users of Angular apps.
For a framework which (IMO) brings so little, the overhead seems surprisingly large.
It works reasonably well. I have touch screen driver support (even managed to get it working with Wayland, once, though that was too slow for the Atom processor), the keyboard is good, and the speed is o.k. The major downside is that the touchpad on the keyboard is terrible. They also went with a custom connector, not the Surface Go one, so you're stuck with a subpar touchpad. Since Linux is not always very touch friendly, it is a bit tricky to use (when no external mouse is available).
Edit: It also has one USB-C and two USB-A and a SD Slot, which is quite handy for a device of such a size
What I'd like to know is if it will work with the battery compartment empty on AC power only.
I'd like to have "smart" displays wall mounted around the house with dashboards on them. I've read about so many instances where people used wall mounted iPads and Android tablets where the batteries would swell after a period of use.
The Pinebook laptops can work without a battery while connected to AC, but not out of the box. A simple hardware mod is necessary to enable this, that's documented by Pine.
I wouldn't doubt it's possible to draw more power than the adapter can give you if you're pushing the MBP full-bore, but I'm pretty sure that under normal power loads, this doesn't happen.
My 16” shipped with a 96 watt charger, and I’ve not seen a laptop with a higher powered one. I’d have to contend with that assertion.
<100W adapters were mostly for potato laptops, until USB-C came into picture which supported up to 100W and now manfucaturers adapted (and probably wait for 200W PD).
I must say that I love that I just plug out the laptop power adapter and plug it into my phone (or mouse) and it charges. Even when it comes with lower performance of the laptop.
I've also had it happen a few times when running unit tests for a different project in Docker, but that's likely just because Docker for Mac makes Electron look resource efficient.
Also one of their older W models that will only run on battery if a standard 90W power supply is connected and will only charge the battery if the device is off. Running on external power is only supported with the 135W or (stock) 170W PSU.
You have 60W chargers usually for systems with max power draw of the CPU+GPU that can reach around 200W peak and in the mid to high 1XXW sustained if you have a DTR/high-performance “mobile” CPU + dGPU.
Even if your laptop comes with a 90W charger it’s still a problem because it’s not enough to power a 65W series H CPU and a high end GPU that draws another 80-120W.
Most laptops will always use the battery directly and trickle charge it due to the power draw.
Power(hungry)ful modern gaming/DTR style laptops can use both concurrently, heck older gaming laptops used to come with multiple power supplies that were needed simply because even the battery wasn’t enough at the time to serve as a buffer, so you had to plug in 2 power adapters to power the laptop fully.
Even without this constraint having a single power path is cheaper and simpler to design so most manufacturers just only draw from the battery and have a separate charging circuit that charges it without having to add power balancing and switching circuitry.
IBM thinkpads used to advertise that you can switch the battery while the laptop was plugged into the charger most laptops of that period didn’t support this either.
If you go even further it was even more of a headache because older laptops used LiPo batteries which were even more finicky.
Some curious digging found two examples (with pictures):
https://www.gamersnexus.net/news-pc/2612-why-2-power-adapter...
https://hothardware.com/reviews/dell-alienware-area-51m-revi...
I was rudely surprised by my first Core 2-era machine that wouldn't boot without a battery installed, because prior to that, from Pentium-based Toughbooks back to the 486-based Noteflex back to various 8088 and even pre-x86 machines (PC8201A, club100 represent!), they were all absolutely fine running from the power inlet alone with no battery.
The Zenith MiniSport was the sole exception because the power jack was actually physically on the battery, but you could still gut an old dead battery and make a dummy, and it would run from the brick just fine. (This is documented in the very first issue of Brian Mork's excellent Minisport Laptop Hacker zine, available wherever fine textfiles are sold.)
I think going with an android tablet is fine and headache free. I just don't like aesthetics of that solution.
The problem with many tablets is that leaving them plugged can cause batteries to swell, which can end up being a huge headache.
I've heard stories of offices that used iPads as wall mounted displays having this problem, so I don't think any particular make or model is immune to this issue.
I considered Pi+Monitor, but for me, something like the Pinetab is a cleaner, simpler solution. And at $99 without the keyboard, that's a pretty decent price.
Of course you should go for what is easier for you, and the pinetab does sound easier, however I'd argue that it is not cleaner.
Separating the screen and main unit will allow for a more serviceable solution, changing either piece if it breaks, or upgrading the main unit if you need more power/different connectivity.
Made that mistake myself. I now have a wall mounted iPad with the charger plugged into a timer powerpoint that only switches on for 1hr a day. Been running like that for 5+ years now without issue. (I'd planned for that to just be a proof of concept, and to use a wifi powerpoint that'd turn on when the battery dropped to 20% and back off where it hit 80%. The easy solution has turned out to be "good enough" and has remained in production, like that cronjob perl script from 2004...)
The timer switch I grabbed to do my proof of concept only has 15min resolution (it's a weird old mechanical one where you move plastic pins in the stop/start time rings). I recall setting it to 45mins because the docs says the iPad gets 80% charge in 1 hour, and it'd occasionally run down and switch off, so I bumped it up to 1hr and haven't touched it since.
I just checked now, it charged for an hour overnight finishing ~10.5 hours ago, and is currently saying 59% charge. I honestly have no idea how low or high the charge level gets, just that I haven't needed to wake it back up from a shutdown in over 4 years...
I've come to doubt the details there are important. Don't leave it on charge 24x7, that'll kill it fairly quickly (~9 months in one data point of mine). Charge it long enough and regularly enough to keep it running, but at the low end of that spectrum. Keep in mind my goal here is less "extend the life of the battery", and more " prevent the battery from swelling up and damaging the device/catching fire", but I suspect there's a fair bit of crossover between those goals, since I've not needed to increase the charge time in response to any battery capacity decrease over 5 years (which in retrospect surprises me somewhat).
If I was aiming to maintain the best possible range in an electric vehicle, I'd probably pay much more attention to max charge and min discharge levels, but "the simplest possible thing" has been working well enough for me to just leave it alone.
If it dies on me tomorrow, I'll be happy enough with the use I got out of it and the tradeoff of time I spent futzing with it.
(Note: I'm also quite familiar with abusing LiPo batteries way past safe limits, I raced FPV quadcopters 5-6 years back, and in my worst excesses, I'd be aiming to completely drain a flight pack in ~2 minutes. They'd end up too hot to touch, and start to decline to the point of not being race competitive within 20 or 25 cycles. They'd also catch into unextinguishable fire in big crashes and occasionally while charging. I have a healthy fear/respect of high capacity and high current LiPo batteries...)
One of my use cases is to monitor my surveillance cameras, which drains my batteries quickly, but I have unused wemo switches that I could put on a schedule to just cycle power more frequently for intermittent charging.
I was thinking I'd get cheap Android tablets and root them, but I'd much rather support something where it's built for user control from the beginning.
Yeah missed the Kanban part
Had seen a simple conf room notification system on those lines (which just mentioned who has blocked the room, when is the next blocked meeting by whom, calendar etc)
2GB of RAM is pretty limiting but 64GB of storage is quite respectable at this price point. The 6AH battery pack is quite beefy too.
no or low amounts of QA significantly reduce cost
https://support.lenovo.com/ca/en/solutions/ht035306
https://appleinsider.com/articles/10/11/05/leaked_apple_dead...
The webpage for the product is https://www.pine64.org/pinetab/
I would rather use i3/lighter DE + touch screen support for some GUI apps?
...well that was... interesting...
Now, this is what I need Pine64 - 'A small pocketable, user repairable linux laptop with GSM module' which can replace smartphone. Current options are GPD[1], One-netbook[2] both of which are expensive, not-user repairable, built for windows.
Then there's planet computer's devices[3], but they seem to be android first although they offer linux as multi-boot option.
Why pocketable linux laptop when there's PinePhone, Librem5? Well, the main bottleneck in the linux app ecosystem is adapting it to the touch screen of the smartphone. I think a pocketable linux laptop could serve well in the interim, may be the form factor will prove better for productivity and will stay as a separate profitable segment.
I would love to enable a few kids and engage with them so I might at least vicariously get some of the fun of tinkering.
Samsung Series 7 Slate XE700T1A
If this functionality was demoed by Apple on a new $999 "MacPad Pro" there would be arguments over if it was the end of the PC and the beginning of the next generation of computing, a leap like the smartphone revolution.
Edits: price, s/this/this functionality/
I took this out of context (sorry), it's a very pre-production device that has received a lot of work since then and will receive much more. https://www.pine64.org/2020/05/15/may-update-pinetab-pre-ord...
(Oh, I guess you were talking about the functionality, not the latency. Ah well, I’ll leave this comment up anyway.)
Apple would be KILLED if they released anything with that much touch lag.
I really want something that I hold on my lap for some emails.
Most models have that back supporting leg which is for table use only.
The biggest difference is that the Magic Keyboard never fails. Not like it works 95% of the time or even 99% of the time. 100% of the time it types when you expect it to, connects when you want it to, disconnects easily, doesn’t double inputs or skip inputs. There’s nothing I would change about it.
To the parent- the surface works fine on a lap, provided you use a case with a solid kickstand (the stand on the actual tablet wants to fold in far too easily). For reference, the reason I abandoned the thinkpad is that I am managing a warehouse and spend ~50% of my day using the surface either standing, on my lap, or haphazardly resting on a pallet of product. I wanted something light and portable, I haven't regretted the tablet decision yet. I seriously considered iPad, but unfortunately it is essentially a phone OS. Win10 isn't the greatest, but it is a full-fledged OS for real work. My only gripe is that it isn't 100% touch optimized, so I can barely use the surface if I don't have the keyboard attached. Oh, and I do dearly miss the trackpoint. However, the keyboard has been great, and with WSL I don't really have any issues with getting the development side of my job done either. I have had a /. account long enough to be considered slightly 'rabid', but as much as it pains me to say Microsoft have made an excellent product.
These OSes are touch-optimized and recommended:
UBPorts (community-driven Ubuntu Touch)
postmarketOS
Arch Linux ARM
I know that desktop Ubuntu has screen reader (Orca) installed by default. No sure about Ubuntu Touch. My Google-fu shows very outdated answers.
That gets me wondering... what would be the ideal here? I don't have much experience with blind-oriented usage but I get the impression blind users only keep screens around for when software/the OS falls apart and orientational cues completely break down... but perhaps sighted assistance is needed less often than I presume, and blind users simply use phones and laptops with screens because that's what dominates the market.
Maybe you could viably get away with having a quick-access mechanism that lets someone use their own device's display whenever assistance is required. I can totally see a device that serves VNC over HTML5 to any web browser via a built-in hotspot, that would be pretty universal.
Which leads me to the question of how useful touchscreen gestures are - since that's all touchscreens are useful for in a blind context - and if there isn't a better input method than that. I get the impression keyboards are pretty much "it". It would be pretty cool if your entire computer was a keyboard with a 3.5mm headphone jack :)
The Pinetab is basically a highish-end ARM system with its own bits of novel hardware added on, so the base question becomes "how usable is the mainline AArch64 Linux experience for blind users?". I suspect it's somewhere between "poor" and "fair"; and bespoke hardware in this device may add a small amount of extra strain on top of that.
Now I'm puzzling over which mechanical keyboards are the thinnest, and if you could cram a power bank and ARM development board into the footprint space underneath.
These things have a very narrow audience so they're one digit more expensive than their regular counterparts.
My use case is basic browsing, taking notes in meetings, reviewing papers (using something like xournal), SSH'ing into remote machines, a small amount of dev'ing (compiling on remote) and perhaps presentations (has HD output). I travel alot and having a machine that doesn't break my back would be cool. If the keyboard is half-way decent I may even use it for writing too.
The reason for this device over other existing solutions are many: low price, built for Linux, lightweight, low-power (good battery), form factor (where did you go netbooks :( ) and a great community!
- random colored text
- line drawing fun
- circles
- z buffers
- convex hulls
- ray tracing spheres
(Checkerboards are too noisy though, and no Newell teapots. Alas, if only my browser supported PBM, the graphics format of winners.)
So yeah, bring it on! Doing more with less feels like a journey worth traveling. I want one!
I am interested to learn low level stuff like CPU assembly, operating system and VGA, VBE, linear frame buffering graphics. I want to see how far I can go with low end x86, arm, misp, and fpga devices
The PineTab – $99.99 The magnetic backlit keyboard – $19.99
That's interesting. I love my iPad mini 5, but at this price I may have to pick one of these up.
I use a Surface Book. I’ve been interested to discover that I would quite often like to have the keyboard continue to work after detaching the base, because when drawing on the device a keyboard is still useful for switching between tools, activating things, typing in precise values when necessary, &c. To be sure, a large part of this is just that I’m using apps that haven’t been optimised for keyboardless usage (I’m talking things like Krita, Inkscape, OpenToonz), but even if they were, a keyboard would still be useful. As it stands, I’d need to buy a second keyboard, one that worked by Bluetooth, to achieve this goal.
(In the particular case of the Surface Book, detaching the base isn’t quite as useful as you might imagine, because ¾ of the battery life is in it, so you’ll get ~3h at most, <2h if using it much.)
source: https://www.pine64.org/pinetab/ ("Learn More About It!" link at the bottom goes to forum post from 2019)
Do you know if these binaries are required to run the Pinetab?
I know USB-C has a bad rep but it is honestly just amazing having one charger that does everything. I don't want to regress from that
If there's anyone from pine watching: Please, add a normal credit card option. Paypal is a horrid company. I'm prepared to pay directly with a credit card via them, but I _will not_ make an account with them just to buy your stuff. Please let me give you money without making a legal agreement with a third party :(
Pine has (still) not enabled it:
https://i.imgur.com/Y1OhCkX.png
In case anyone from pine is watching, here's the doc on how to do it:
https://developer.paypal.com/docs/integration/direct/payment...
It's like....5 clicks.
EDIT: it is indeed gated by some vague heuristics[1].
>Buyers don't always have the option to complete their purchases without using or creating a PayPal account. This option is presented based on several risk factors, including but not limited to the buyer's PayPal purchase history, PayPal cookies stored on the buyer's computer, the buyer's location, or a credit assessment.
[1] https://www.paypal.com/uk/smarthelp/article/how-do-i-accept-...
That's probably why. I use addons that autodelete cookies as soon as I leave webpages because no one has the luxury of pretending they don't know how badly cookies are abused at this point.
I guess the inevitable next step was for companies that abuse cookies to start punishing people for refusing to take their shit.
I reiterate: paypal is a shit company. Pine, please provide an alternative.
That would make this tablet the ultimate Pocket CHIP[1].
[1] https://en.wikipedia.org/wiki/CHIP_(computer)#Pocket_CHIP_an...
Its gathering dust because its so underpowered, could not find a use for it as a GUI user device.
I think some kind of kiosk / iot hub with display application could work, if you put in the effort.
Won't a generous allocation of swap compensate?
You'll also rapidly exhaust the SD card's flash lifetime, and all the spare capacity... which will make the SD return read errors, nicely corrupting the contents of RAM.
You can verify the slowdown/lag/pain factor easily enough: boot Linux on any PC with `mem=2G`, attach an SD card you don't mind nuking :), and go nuts.
Is this more of a hacker device (IR blaster, odd assortment of ports, full OS, etc.) or consumer? It should ideally pass the wife test...
[1] https://store.pine64.org/?product=sopine-a64
[2] https://wiki.pine64.org/index.php?title=PINE_A64-LTS/SOPine
Looks like Resterampe/left-over sale to me.
Come on, 2GB? In 2020?
If I'm expected to go root out some forum thread to monitor progress on things I purchase from the pine64 store, that's worth noting here to inform everyone's expectations.
They could post a shipping update on the wall of their local laundromat for all the good that did me when I'm not monitoring their message forums.
If not for this thread, I'd still be operating in the dark having basically forgotten I even bought the thing. I appreciate your comment pointing me at the forum, pine64 should have emailed that link to everyone who ordered one in the affected time period.
I buy mine because I'll jump through any hoop to not carry around a fortune 500 in my pocket
For the consumer market -- runnability of android? hardware quality? low price? Some compatibility factor that only an open company can hit?
It'll come, eventually, I hope.
I'm tired of not being able to uninstall Dropbox because Samsung doesn't want me to.
I actually really like this system and wish that I could trade-off further down the spectrum on quality. After all, to get a zero-dead-pixel device you just have to check in the second-hand market like people do for high-binned CPUs. The volume of true demand for those CPUs is high enough that there are retailers like Silicon Lottery that explicitly look out for that.
It turns out that most consumers don't really want a zero-dead-pixel guarantee though.
Alternatively they just buy their displays from brick and mortar companies that allow you to return for any reason at all and just return for replacement any device that has any number of dead pixels.
I've never bought a display that had any number of defective pixels in almost 3 decades of using computers I'm betting lots of people are in the same boat. Logically the price is built into whatever such places are selling them for based on some percentage of users returning anything with any number of dead pixels.
> As of 6/7/20, the top 20% of tested 10700K fell within our 5.1GHz bin.
Interesting!
Just for the sake of argument lets suppose it had a 1% chance of failing between 30 days and 1 year even though I would hope it is far less when it would normally have been covered. Assume you remedy the issue by replacing it out of pocket.
Expected cost is 0.99 * 100 + 0.01 * 200 = $101
Also keep in mind that brands pay for warranty services out of profits earned from higher margin sectors and volume neither of which is applicable here.
I didn't down vote either your prior post or this one. I value durable things. My 8ish year old desktop just died. Both my phone and my laptop are 5 years old. On the other hand I see the value in cheap things that will be affordable to all made by small distributors that value our freedom to use our devices. Everything in life is some degree of trade off and I see room for this one.