Open Source Smartwatch
pine64.org
pine64.org
Unfortunately, the mediocre hardware used in their devices keeps holding me back.
First came a phone with a 9-year-old A53 Cortex chipset. Then a tablet with the same outdated chipset. Then a laptop that ships with a Rockchip RK3399 with 4GB of RAM, in an era when 4GB barely suffice to run a browser with a few tabs. And now they are releasing a smartwatch with a 64 MHz core and 64 KB of RAM - some specs you would have expected from a very cheap smartwatch released 7-8 years ago.
Why should I use a device whose hardware is almost a decade old at the time of purchase? And how do they even expect anyone to actually use these devices as their primary daily drivers, and not as proofs of concept, nice toys to tinker on every now and then or show off to friends?
I understand that Pine wants to use 100% open hardware, and that comes with heavy compromises, but are these compromises worth it?
If you ask even the geekest user out there whether they prefer a device whose source code is 100% pure and doesn't contain any binary blobs, or a device that doesn't start freezing/lagging after a couple of minutes of normal usage, I bet that anyone would opt for the latter.
If you want to change this, you'd need to either start a semiconductor company, or engineer some sort of coup at one or more of the "most unpleasant to deal with" Silicon vendors (Broadcom, Qualcomm, etc.).
Otherwise, as unpleasant as it can be, you have to make a deal with Broadcom/Qualcomm, or get some other vendors to license their ARM chipset to you.
The latter doesn't seem like the end of the world to me TBH. Of course I'd love an ideal world where open architectures like RISC V have taken over and are actually usable, but until we get there usability remains more important than purity - unless you're RMS of course - even if it involves a small "deal with the devil".
I believe that most of the geeks out there love open-source products not much because they work daily with kernels and drivers (if that's the case then the argument for a pure open hardware without hex bloats makes sense), but because of the flexibility that they offer. Because they can install anything that they want on them and aren't prone to abandonware/planned deprecation like many products powered by Google or Apple. But if you don't give them products that they can use as main daily drivers, you'll rarely get enough momentum outside of the small tinker-and-show-off circles.
My iPhone 6S is more than 5 years old, and is perfectly usable as a daily driver (and my SO's iPhone 6 is too - it's 6 years old, only runs iOS 12, but still gets security updates and is perfectly usable).
It's fine if on the freedom/performance/up-to-dateness tradeoff, you're not pine64's customer.
Many people are.
If you're making a point about 5-year-old hardware being still useful today, I think you need to reference typical 5-year-old hardware. The 6S was weird (albeit in a good way).
My Xiaomi Mi Band probably has similar specs. I bought it last year. You position this as outdated, but for me, the specs mean that I only have to charge it every couple of weeks.
I know this is a bad cliche, but this "decade old" processor is more powerful than the most powerful PDP-11 systems that have ever existed, that were last sold back in 1990s.
That said, I too would love to see higher specced devices from Pine. An update to the Pinebook with a fancy new ARM chip and 16GB ram would be tempting indeed.
It's always hard to stay true to ones beliefs; when it comes with so many shortcomings.
When you have a powerful processor like the Apple Watch does, your power draw is pretty high. This is why the Apple Watch only gets 1-2 days of battery life, while the Pebble has 7 days, and the Apple watch just got a display that is always on, while the Pebble has had a always on display since day 1.
I bought a pinebook pro and i have some SBCs. The PBP is perfect travel laptop with long battery life that can run on my phone charger. Firefox runs buttery smooth on Manjaro i3.
You might be interested in Librem 5 phone: https://puri.sm/products/librem-5.
- This is basically a commercial smartwach (the Colmi P8) that is based on a well documented SOC, the nordic nRF52832
- The open source version comes with an enclosure that is not glued shut, so that it is easy to access testpoints on the pcb.
- Development is done by accessing the nRF51 SWD interface with pogo pins or similar. The programming interface can be a standard SWD interface such as an STLink clone
Ok, sounds easy enough. It should be possible to just buy one of those smartwatches and crack them open? But OTOH the developer version is not much more expensive.
The main question is now: What software to run on it, that goes beyond a few experiments? Is there any kind of open source smartwatch OS that would run on this?
Edit: Corrected typo, of course I meant SWD interface, not SWM. (SWM or rather SWIM is for STM8)
I would advise against the ST-Links though as they expose a more high level protocol and don't always work with non-STM chips. FT232H dongles make for a more versatile programmer.
and the default https://github.com/JF002/Pinetime
My guess would be that Wasp-OS will become the beginner-friendly fully-featured option while Pinetime will be the more battery friendly option.
Could the way that they're sealed become a problem? Description says waterproof glue will be involved in the final version and seem a bit unclear about whether it'll have any mechanical interlock [0]. Sounds like one might have a hard time putting it back together...
[0] https://pine64.com/product/pinetime-dev-kit/?v=0446c16e2e66
https://tinygo.org/microcontrollers/pinetime/
With that, you can write Go code which (when compiled) is then flashed to the watch for running directly on the hw. No need for an OS layer underneath.
Maybe Fitbit can be petitioned to make the Pebble's hardware fully open source too.
I really hope the PineTime will pick up where it left off! But it's only in dev kit stage now.
I was a backer of the original pebble too.. It's a shame things didn't work out. I think they just grew too quickly and got caught up in too much investment money wanting too quick returns.
[1]: https://www.tindie.com/products/ttgo/lilygor-ttgo-t-watch-20... [2]: https://github.com/Xinyuan-LilyGO/TTGO_TWatch_Library
Is the size some standard? I would prefer some other belt but don't know how to search for it?
I'd say that MicroPython is more popular than Lua at this point.
Here are some useful starting points if you want to play with it:
Instructions for setting up a platform that easily allows you to build your own simple apps for it: https://www.instructables.com/Lilygo-T-Watch-2020-Arduino-Fr...
Extended version of the platform with many included apps and features: https://github.com/wfdudley/T-watch-2020
MicroPython port for the watch (have not been able to try it yet and not sure if it is maintained): https://github.com/OPHoperHPO/lilygo-ttgo-twatch-2020-microp...
Also NordicSemi never seemed to be particularly opensource friendly. I'm hoping for a future RISCV edition of this watch with more ram.
240x240 is ~56.25 kB at 8 bpp. the screen is RGB 65K colors, that is, 16 bpp; so the framebuffer size should be twice larger. I assume that the display controller has its own memory. True, you would need another 112,5 kB for double-buffering, but maybe you can do without it.
Oh they could have. The problem is: it's extremely hard to get access to powerful SoCs - the vendors simply won't work with you and most of the documentation is under NDA.
That's a really good point. I remember having the Pebble watch (this was pre-Apple and Android watches) and I think it had 128 KB of RAM total, including OS, background tasks, apps, etc. That was a mass produced and commercialized item though.
Granted that was an e-paper display. (They had color and black and white options)
Recently I got a Mi Band 5 and learned they transfer BLE pairing over platforms - they make devices generate new MAC when unpaired, which is tied to Huami account upon pairing. Gives me different chills than the fact that its sole purpose is to take my health data anyway.
I have a Mi Band 5, and wanted to use GadgetBridge instead of the official Mi Fit or Amazfit/Zepp app.
I was surprised to learn this was not easy to do. I needed to retrieve the Bluetooth MAC from the proprietary app via rooted app, or by logging in via script, which no layman would do.
The watch is now paired with GadgetBridge, but the steps are hard enough to dissuade anyone from trying it.
> I have a Mi Band 5, and wanted to use GadgetBridge instead of the official Mi Fit or Amazfit/Zepp app.
> I was surprised to learn this was not easy to do. I needed to retrieve the Bluetooth MAC from the proprietary app via rooted app, or by logging in via script, which no layman would do.
> The watch is now paired with GadgetBridge, but the steps are hard enough to dissuade anyone from trying it.
There is now also an android app to fetch the key and gps almanac, very simple to do, no root or python required.
Note: I kivy wrapped and apk packaged the above mentioned script.
The Amazfit Bip has an unofficial BipOs, which is a very cleverly created reverse engineered SDK allowing to uploaded custom compiled programs.
This strikes me a really over the top expression of contempt which just isn't necessary or constructive at all. I generally understand Pine64 products to be developed to the point that they are capable of being used by end-users from day one without hacking or fiddling, are basically ready-to-order in the same way as any commercicial product but with a general understanding that they were hacked together, and may require the kind of hacking that only enthusiasts are interested in doing over the medium to long term. To my mind that is enough to elevate it above junk (are there other distributors comparable to Pine64 that I don't know about?), and I feel that it's quite strange that that should even have to be debated at all.
>while genuine Huami goes on to collect data more strategically at a larger scale.
You're saying that as though it's a good thing, which I don't understand.
>Gives me different chills than the fact that its sole purpose is to take my health data anyway.
I also couldn't parse this.
Was this intended to be a reply to my comment?
If someone from there reads this, I'd suggest also updating the product page with a clear statement. Right now it just says "Estimate dispatch in late July, 2020", which makes it look abandoned.
Cheap, sleek, waterproof, gorilla glass always-on display, battery lasts a month, does GPS tracking without smartphone connection and you can sync it via Gadgetbridge to keep all your data private without any account/cloud services.
I have the original Bip and am very satisfied with the hardware (lasts forever, looks pretty good) and somewhat satisfied with the software (good selection of watch faces, it's gadgetbridge compatible, but activity mode is a tad bit too easy to enter and too hard to exit)
I have the Bip S, and was trying to get that for a friend, but as I mentioned... I cannot find them in stock anywhere.
I've been very pleased with it. Great battery life, even with GPS while running. I found it comparable to Garmins in that respect.
The chunky looks could turn people off, but I prefer ugly retro-looking stuff :)
Edit: here is DC rainmaker's review https://www.dcrainmaker.com/2020/06/timex-r300-gps-smartwatc...
Still quite impressive though. It defies belief getting 2 weeks out of a GPS watch charge.
But once you move indoors and lose signal regularly and have a much higher noise to signal ratio, I very much doubt this can still work. I don't know about you, but I don't work in the garden, so most of the day that GNSS unit would still be listening and calculating (as I understand, the code division multiplexing means you kind of brute force the various codes until one produces a signal instead of noise) and might not even get a fix at all (my phone takes a few minutes next to a window when there is concrete on all sides except for that narrow view of the sky, and this 2018 phone is already orders of magnitude better than my previous Galaxy Note 2 was). I'm not saying it's impossible, but I would highly doubt that it consistently gets a fix in 2ms, or even 200ms, in the place where you spend most of your time.
Of course, it's mostly useless indoors anyway, but so you'd have to manually toggle it every time you go in- or outside, or somehow configure a home location where it turns off (would work great during COVID) if it has such a feature, though I wouldn't know how it would figure out when you've left home in that case. (WiFi comes to mind as an obvious solution, but then how do you explain to the user that it won't track your GPS while that WiFi is in range? It just seems like they'd drop the auto-off feature at this point.)
So if you don't need the position in real time for navigation it is quite possible to track location over a week with a tiny battery. This could be time triggered and/or based on MEMs counters (which are also power friendly). Or you could calculate locally if the user explicitly asks.
https://ieeexplore.ieee.org/document/5708033 http://publica.fraunhofer.de/documents/N-464241.html
>The 3-Pack sealed PineTime aimed solely for project deployment purpose only, this is not for end user who is looking for ready to wear Smart Watch.
just like the Pinephone.
It's a little bulky but seems like it could be a fun device to hack around on. I got one recently from Adafruit.
The processor draws about 220uA [1]
so you have about 818 hours (=34 days) of CPU time, but you need to take display current out of that budget too, and an easy source for that probably exists, but I can't find easily.
[1] https://devzone.nordicsemi.com/f/nordic-q-a/7893/nrf52832-cu...
I prefer it that way. would not want to charge yet another device every day or every other day.
The Pebble Time had a Sharp Memory LCD, I'm not sure what the Amazfit Bip uses.
It's baffling that there are so few electronics in general that use E-Paper/translective displays. Does anyone know?
Sure, the image quality isn't super amazing, but it displays color, uses super little energy and still updates quickly. That's pretty much my point.
The HiSense Q5 is a tablet with a transflective display, but for some reason that is black and white.
I've changed the battery once in 2 years (and that was simply preventative). The fact that it just keeps on working for years at a time is a nice contrast to basically every other modern device.
I love mine. I get my notifications, pulse, step, and sleep tracking, in a classy looking package with a grown up battery life.
These might be nitpicks but just because it's open source doesn't excuse it from poor design critique.
Tech wise looks really good for the price.
It's a 3D render. The actual display seems to be more or less centered[0].
> Also the wake up button is on the left instead of the right
One could presumably flip the device and the rotate the screen rendering by 180 degrees. This is how Apple Watch handles right wrist use.
[0]: https://pine64.com/wp-content/uploads/2020/09/PineTive-DevKi...
Good call on the screen rotation, is this something any of the OSes support or needs to be requested to be added in?
When I ordered a PinePhone from Hong Kong, it was just too much of a hassle for my liking. Waiting for an indefinite amount of time until it gets shipped and then deal with customs.
I wonder how many customers there are per country. Would it make sense to have local distributors around the world?
Or maybe at least one per continent?
https://www.pine64.org/2020/12/02/pine-store-community-prici...
These new stores will offer a higher standard of customer service, but the products will also be priced higher. The original community Pine Store will remain available in its current form.
I personally had no issues ordering stuff from Pine64, albeit it did take long.
I got my Pinebook from Hong Kong and was stunned at how quick it came. I was expecting a month or two, and if memory serves it was closer to a just a couple of weeks.
I wonder why phones are more of a hassle to order from Hong Kong than laptops?
But doesn't quite scratch the itch that I've got for a higher level open source wearable or tablet (think IP67 Raspberry Pi tablet or watch) The key is that like this watch, the hardware is done and "consumer" ready.
A Pi tablet would be a very fun toy. I’d be surprised if there wasn’t something available for that now but may be bulky.
A slightly alarming statement on the page:
"Small numbers (1-3) of stuck or dead pixels are a characteristic of LCD screens. These are normal and should not be considered a defect."
I've bought dozens of screens in my life (phones tablets monitors TV's) and have never had a stuck pixel. Is it actually normal?
I don't know where pine falls, I would guess they buy straight from the factory without extensive QC and might seldomly have stuck pixels.
I guess if I were out hiking it could be pretty nice to see where I've hiked. But that seems like such an edge case for 97% of people.
You could also use it as a bike speedometer or a running pace monitor.
(Though, the battery may be too small for all these use cases.)
I ended up buying a Fossil Hybrid HR instead. Its a decent watch, just a tad expensive.
https://www.tindie.com/products/ttgo/t-wristband-diy-program...
Hats off to the Pine folks for having a smartwatch, but for God's sake if they are going to develop their own devices from the ground up please include a temp sensor at the least.
Used it once to pay for dinner and jogging a few times and found it to be not useful as I keep going back to my smartphone which I am not keen to carry while exercising.
I'm hoping that Smartglasses won't be a flop but can't help but feel we are no in a plateau of really incremental innovations moving the screen from desktop to our pockets now to our wrists, our glasses and perhaps if Elon gets his way our brains in the future.
Still would want one tho as it appears it is very light unlike many major brand ones that seem so bulky and large on my tiny wrists
I've also seen a DMV worker use the Apple Watch to distract himself with information every spare second he had in his interaction with me.
I see SMS notifications and can even give a monosyllable response while the phone is stowed.
That's a close to the Full Dick Tracy as I get.
All the time. I use it to read texts (a lot of time they don't require a reply, they're just informational). Set timers. Check the weather before I head out. Track my workouts. Listen to music. Log meals. And I don't carry my phone with me when I'm working out, so sometimes I take calls on it.
Lots of Apple Watches and a decent number of Garmin watches. Don't see a lot of anything else. I suspect the fact most of them are bulky and have shitty CPU options is a big part of it. Garmin users are willing to accept the bulk for the functionality. Apple Watches aren't that bulky.
I use mine for workouts, taking calls so I don't have to pull out my phone, sending text & receiving text messages, and controlling music. For my light mobile phone use, it's nearly suitable as a replacement for my phone.
Some basic things you wouldn't expect, like crypto for HTTPS, take a LOT of work on a tiny wrist-mounted CPU.
I wish smartwatch templates adopt more multiple programmable buttons. Nice to have a functional media remote during winter with gloves on. Pepple got so much right.
Sadly moved to an Apple Watch, since the Pebble ecosystem started to die.
I'm really just looking for something that can track my health data + sleep and has an api to get the data in a raw form for further analysis.
But probably even if it had it it wouldn't work with Google/Apple Pay, for me payments using my watch would be the best application of "smart" together with notifications.
Defy not with 4MB of storage.
On the other hand good luck with getting such a device accepted by anyone in the financial industry. (On the other hand you probably could build something that proxies to your real physical card)
Cryptocurrencies are better in that regard but few businesses accept cryptocurrencies, GNU Taler would also be better but I think that requires banks to be on board with it which they're also unlikely to do.
What's flashed into the device is the whole system image, and it's up to you (that's the whole point!) to decide what this image does.
Anything sample code they provide is to be seen as a sample only.
Some might even say that it fails right at the start at being a smart watch since it doesn't have an always-on display.
There's still a long way to go before it's viable for the average tech-geek to go for a fully open-source setup (phone, watch, tablet, etc), so if your question is if it's good now, then no
If you're asking if this will turn into something good over time, then yes... A million times yes
The open-source space will most likely always come with its own set of problems and I don't think we'll ever get an environment that has the same level of seamlessness as android or iOS.
I'm hoping the next 2-3 years will get us to a point where open source can get a user 80% of the way out-of-the-box, at which point I'm definitely ready to ditch android and move everything to an environment, that's built to respect my privacy
What does this even mean. Already fragmented apps before it even starts, what else to expect.
Yes, open source leads to experimentation and individualization which leads to fragmentation. It's a trade-off, but and there are already plenty of other smart watches going for a less fragmented but less open ecosystem
I didn't mention anything about walled gardens.
It doesn't work by installing "apps", but rather, the whole system image is built from _your_ code. Which makes sense on a device that has 64KB RAM.
Any code samples they provide are just that: Samples.
If what you're looking for is a general public oriented device, this is not it. There's plenty in the market to pick from.