Looks like Google is coming out with Wear OS 3 soon, so I don't want to invest any real money in a Watch until then
51 USD will get you a sealed smart watch and a dev kit :)
Once it's shipped, it's 60$ or so venturing close to what a WearOS watch cost ( I'm seeing older ones at 80$ or so)
I also want to find a good smartwatch, tired of waiting for wear3.0, but an always-on display is a must have for me.
The dev kit seems a bit too low level for me
https://github.com/daniel-thompson/wasp-os
i have personally made a metronome app for it
https://gist.github.com/mishushakov/7af0c459178f152a27005d31...
AFAICT, it's fully open source software (except, presumably, the nordic hal stuff?), with a JS SDK and a hosted IDE. But, I haven't used it yet because it requires WebBluetooth, and I refuse to use chrome, and I haven't figured out if the rasppi hosted IDE is still supported or not, it seems that the `nw` package doesn't support arm (anymore?).
But this was the best option I found when I was looking awhile ago.
The other things I considered were the Pinetime (software is still a bit too early from what I've seen) and "Paul's Open-Smartwatch" https://open-smartwatch.github.io/, but I really wanted GPS and that version's design isn't finished yet.
These are all problems fixable with software updates, and it's very early stages of the project. The app store grew from 80 to 130 apps in the last 3 weeks.
The hardware on the other hand is great. It doesn't feel cheap, but it's really lightweight, which I enjoy. The e-ink display is crisp and clear, easily readable in direct sunlight, and it has 8 colors, which is a cool feature for an e-ink display.
Didn't get my hands dirty with the SDK yet, but will hopefully find some time during the holidays. It looks well thought out though, and the web IDE and emulator seem useful.
Really looking forward to dig into it.
Can you modify the OS itself.
I want to program a watch that doesn't actually tell time.
The code was on a Raspberry PI in which a 3G/4G dongle was plugged, and it connected to the watch, uploaded data, was robust do disconnection events either internet or BLE (used Thespian and the Actor model to create and kill actors). The devices were geographically distributed on different timezones (a fucking nightmare) and basically needed to work for users with zero technical skills. It needed to be plug and play and it just works. Even configuring the pairing between a Raspberry PI and the watch needed to be low touch (I used the analogy of a dog and a hand: you put your hand out for the dog to smell and recognize you, you put the watch nearby the Raspberry PI and it computed the RSSI and duration to detect the "intent" so it didn't pair with "drive-by" devices). Kind of like NFC, but for BLE. The data was sent to a backend, then displayed on a mobile application (the prototype had a Grafana dashboard that showed your activity in near-real time).
Talk with them. They'll probably send you the PDF spec and you can do whatever you want.
That said, given how cheap these watches are I'm open to trying anything. Were you able to write your own custom app too, or just consume existing data ?
As mentioned in my original comment with additional: I wrote an application that ran on Raspberry PI devices in different continents, countries, and time zones that connected to trackers worn by people with zero technical skills and worked nonetheless surviving disconnection events and automatically upgrading the software, communicating with the trackers and sending serialized data (protobufs) to a server application in Scala my colleague wrote, that then made analysis and aggregations and sent result to yet another mobile application (Android) another colleague wrote to display graphs and other data. The application I wrote worked with several models of this company's products, and at some point, we thought of making it more generic and use it for video and sound feeds for another project ("sound event detection", let's say).
Yes, I was able to build a custom app and it worked not only on my data, but on others' data as well.
The funniest conversations we had were about time and time zones. At some point, we started using timestamp timestamp and timestamp not timestamp because working on data from multiple time zones makes you question what on earth is time, and what actually is now ("Now for whom?" "Now where?").
I had literally nightmares about time and time zones. No, I am not kidding.
Check out the watchX: https://youtu.be/AKSIGRdL-ts You can code it with Scratch or Arduino. There are a lot of examples available: https://community.watchx.io/t/watchx-watch-face-collection/6...
There's also Bangle.js to keep an eye on.
Highly recommend the anodized aluminum case. The plastic one is sturdy enough, but with the thickness of the material, makes the watch feel bulky. Also, the aluminum one has such a good fit (they are precision CNCed) that it can handle water splashes with no problem, and a few gaskets can get it to IP64 or 65
IMO, it still is one of the best smartwatches.
[1]: https://www.ebay.com/sch/i.html?_from=R40&_trksid=p2380057.m...
On December 7th, 2016, Pebble announced that they were ceasing operations. Two days later, on December 9th, 2016, the Rebble web site went live, announcing our intention to "maintain and advance Pebble functionality, in the absence of Pebble Technology Corp." We have fulfilled much of this goal via Rebble Web Services, which restores most of the functionality that the Pebble community knows and loves.
Wear OS 3 is already out.
I’m in the Apple ecosystem and very happy with the watch, even though the newer versions are surely nicer.
