FOSDEM 2024: PineTime Talk | Hacker News Reader

FOSDEM 2024: PineTime Talk

blog.mlich.cz

96 pointslupyuen32 comments

Bought a PineTime for $27 plus shipping. Wear it while running to measure my BPM. I send the BPM to the Gadgetbridge Android app via Bluetooth. Upgraded Infinitime via Bluetooth. Have not messed around with the OS code yet.

Have been happy so far.

I love my little PineTime. It's a great little watch. Took a little bit to work through to start up, but after that it's pretty golden. Simple and has the basics. I don't want a watch that can do everything, that's what my phone is for.

Does it do sleep tracking? How long does the battery last? I still daily drive a Pebble 2HR with over a week of battery life but am looking for a replacement compatible with Gadgetbridge, and there really aren't any that can track sleep in a way Gadgetbridge can understand...

If I do cardio exercise for an hour and am measuring my BPM the whole time, a little less than 20% of the battery is drained. They changed the BPM algorithm in the latest Infinitime update, not sure what the battery hit was in the old one.

I turned off Bluetooth and put the watch on the shelf five days ago, the battery drained maybe 14% over that time. I mainly use the watch to measure my BPM, but it can do other things (measure steps walked, get notifications from an Android phone, stopwatch, tell time of course, a somewhat weak flashlight app, it even has games like 2048 and Pong).

On Android there is a Sleep as Android app, which I have never used, which says it integrates with PineTime. Reading what people on the net say, some have been happy with it, some have encountered problems. I have never tried to get sleep readings on Gadgetbridge and don't see any talk about Pinetime and Gadgetbridge for sleep on the net. I know someone who has a Fitbit where the sleep measurement from watch to Fitbit app is very extensive, and even tries to estimate when someone is in REM sleep and such - Pinetime doesn't have anything like this yet.

To me it's a $27 (plus shipping) hackable watch that does the main thing I want it to, measure BPM. As people code for it, it will probably get more capabilities - someone already improved how it measures BPM in the latest version of Infinitime by doing some fast fourier transforms and the like. Some of the more expensive watches seem to have more capabilities, but I like it for my purposes.

From looking at the slides it looks like a really impressive project.

But I can't help but feel like very low cost full-blown Linux devices are right around the corner. Ex: https://www.cnx-software.com/2024/02/08/licheerv-nano-low-co...

If you can run full blown Debian on your wrist soon, why bother with this severely restricted device that are a pain in the ass to develop for?

But I can't help but feel like very low cost full-blown Linux devices are right around the corner.

why is that a but?

i don't think this is a case of, if we had waited longer we could have made them cheaper, but i rather believe that pine64, fairphone, and other projects going all the way back to openmoko were laying the groundwork that make the next generation of cheaper and better devices even possible.

so yes, very low cost full-blown Linux devices may very well be around the corner, but that's no reason to dismiss the pine64 products, on the contrary, they should be lauded for being pioneers

I'm not any kind of authority on this, but for the most part the hacker/OSS community is taking cheap chips and repurposing the hardware for other stuff. So progress in the space happens on its own. For instance the chip in the Lychee RV Nano is made for IP cameras. I guess my point is that once you can run Debian on your wrist (which will happen for external reasons), all these complex setups with customs RTOS stacks that people are working on may become kind of irrelevant. It's always a shame to see people's hard work become obsolete

You've made a very cool suspension for your carriage, but Ford is about to mass produce the automobile

i get your point, but i actually think that RTOS solutions will remain interesting as i believe they should be able to run on less power thus enabling uses that devices which need more power can't. (like a wrist watch that doesn't need to be recharged as often)

Yeah, totally could be :)

Total layman guess, but I think with an e-ink display you could get away with the chip sleeping most of the time

Battery life would be a major reason.

Battery size and battery time are the primary constraints for a wrist watch. The power efficiency of a SoC that runs full blown Debian, as well as Debian itself is much worse than the NRF52 used in PineTime. Without doing calculations, I would estimate 10-100x. With corresponding increase in battery size or decrease in battery time. That said, I believe there are smart watches that run Android on Linux, and they do manage 1 day of use.

A better development experience would be to use MicroPython. That will require a microcontroller with more RAM and FLASH space, but could equally power efficient. When using the LVGL UI library, and decent drivers for the peripherals on the device, and a live she'll connection, it is a very productive and pleasant developer experience.

The TTGO Watch S3 is an example of such hardware based on ESP32-S3 chip. Hardware is nice and cheap, but they have a hodgepodge custom SDK with forked MicroPython :/

I agree with sibling comments that battery life would be the real issue here, but I was astonished to see the $9 price tag on this. Incredible.

As someone who’s worked on both embedded and Linux-based smartwatch firmware, yeah it’s mostly battery life. The nice part about developing for a lower power microcontroller is that you can basically aim to squeeze as much performance out of the chip as possible, and the power draw will still end up pretty good. On Linux devices, you need to an architecture early on for how you’re not going to blow your power budget immediately. Even stuff like entering/exiting suspend can be problematic in how many Joules it takes.

it took me two weeks to get the sealed pinetime working back when it launched. for some reason it would get stuck on the initial boot screen, and it would be in a state to where I couldn't update the firmware. eventually in a mindless "flap/spin it in a circle with my hand while I was thinking about work" I felt it vibrate and to my surprise, it booted for the first time ever.

It turned out, even with a firmware update months later applied to it, that whenever it gets rebooted, I have to do the above movement until it finally decides to boot, and it's random. Sometimes it takes a few spins/flicks, sometimes I'm doing it for minutes on end.

I confirmed this isn't a red herring by leaving it in its frozen state and motionless, and it would never recover, even 24hr later.

This was the second Pine purchase of mine, the first being the PinePhone Pro, which A) I couldn't get activated on AT&T (not Pine's fault with the 3g phase out and whitelist approach, but I digress), and when I used TMobile, I couldn't make a call with it without having audio problems or etc.

Look, I wanted to support them, and the idea of less expensive open hardware/software, even if it meant rough edges, but completely non functional for even the most basic usages... I might as well just build my own and understand all of the reasons why something doesn't work instead of spending days or months getting it to do what it said it would do on the tin out of the box.

That pretty much sums up their early adopter experience.

The PineTime ecosystem is pretty neat nowadays, just try out InfiniTime: https://github.com/InfiniTimeOrg/InfiniTime

Reply on news.ycombinator.com