Acceptable Clones? Perhaps passed through Gordon?

nm. @fanoush, good job popping the enthusiasm. Super cool. Ideas are bad.

@user156811 did some work to facilitate this kind of thing on https://forum.espruino.com/conversations/393880/ . I think he's working on something a bit different now though.

nm

I'm mainly meaning there are a lot of clones out there

Can you show one such example?

nm

nm

Those are not clones. Those are completely different watches. You said "I'm definitely not meaning that we should port it over to something else"

But anyway, yes we have Espruino working on couple of nrf52840/32 watches, maybe after some more googling you'll find some of them, then you may possibly also search this forum and you may find more. Then you may check again that topic that was already advised. And then you may explain again how " that's a little different to what I'm thinking.".

I just love it when fanoush gets serious :)

it is not as easy as you think @andiohn, only the watch named bangle.2 in this forum has the swd exposed, from the rest only the ones from dafit are hackable due to the efforts of fanoush and atc1441, and then again, they are not very similar to the bangle.

Jeffmer has done some work on porting some banglejs only functions to generic watches, but what you hope for is not easy at all.

https://github.com/jeffmer/WatchApps

Maybe to add to that: even of you get bangle apps to work on another watch, you'll find out that many apps have been written for a specific watch (bangle 1 or 2 respectively) and that most apps just won't work nicely with another setup (e.g. rounded screens, more or less buttons, different screen resolution etc).

So getting the runtime to work is already a lot of work, but then getting the User-Level software to work is another herkulean effort.

But believe me, I also would love to have some more choice on the hardware side. Since you look for ruggedized, maybe my ruggedized cover[0] is something you'd like to have manufactured for you.

[0] https://forum.espruino.com/comments/17490753/

So getting the runtime to work is already a lot of work, but then getting the User-Level software to work is another herkulean effort.

And sadly it is moving target. Both the apps and the firmware itself (including the bangle UI specific stuff) is constantly improved, so one needs to really like that other watch to keep maintaining it. Personally I think watch without sunlight readable always on screen is not that great. The other watches are nice for hacking and reusing for something but I didn't find other watch that I'd like to wear on my wrist daily. And I think the pure javascript way on top of generic Espruino that is used for those watches that jeffmer has there is easier and more fun to do than maintaining forked version of stuff in https://github.com/espruino/Espruino/tree/master/libs/banglejs

Yes, @jeffmer has done a bunch of work on different watches that might be worth taking a look at.

But as mentioned they are usually a bit tricky to flash firmware onto, and then the screen/buttons/etc are different enough that many existing apps may not work quite right.

Right now I get a decent amount of people trying to write Espruino onto non-official devices and then demanding support if it doesn't work or they can't return them to the way they were before. On Bangle.js 1 we updated firmware wirelessly and that failed around 5% of the time and bricked the device, requiring us to open them up (and that was once we'd really dialled everything in). Now imagine what'd happen if even for 5% of the devices that were bought they got bricked during the firmware update - it'd be a nightmare. I think I'd lose more money than I made!

You did go a bit nuts posting on 7 posts in the 24 hours since you had the button issue - but I think your 'stuck button' is most likely just a failed firmware update that can be easily resolved. Let's discuss that in https://forum.espruino.com/conversations/374874/?offset=50#comment17606350 though.

nm

I'm not looking to port it to a different chip

why not, it is a nice idea, I think you should go for it.

I didn't expect people to disregard the idea that close neighbor watches could have a plugin to get closer compatibility and not have to maintain a fork.

Looks like you give your own meaning to the term 'plugin' just like with 'clone'. So far it is just a lot of hand waving.

Perhaps it could start with jeffmer watches and see what it would take to just have an 'enablement' plugin instead of a fork?

Good idea, and at some point you may also check that topic suggested in post #2, it may be related.

Nm

that project to support the P8

the p8 is long gone, out of sales. The same goes for most of the nrf hackable watches, very few exist nowdays, you are 3-4 years too late.

you are also missing an important point. Espruino is funded by hardware sales. It makes no sense trying to bypass the means of funding. At the end, if one succeeds in porting to a cheap alternative, he will only harm the project. I think this is mostly the reason that very few people will be willing to offer what you ask for. If you want to put some effort, espruino is ready for some 4-5 watches, if not, the bangle.js2 is fine as is.

Gordon, the bangle.js2 software could literally power 10s or 100s of watches. I'd would have loved the project to have that type of attention.

I'd have liked that too, and Espruino was originally created as a way to bring together a bunch of third party hardware that was otherwise hard to program, but I had to swap to selling hardware because staggeringly few people were willing to pay anything for it (we're talking me making $30 in an entire year)

I'm sure there are one or two people like you that might support it, but the reality is that the majority of people wouldn't - and as I mentioned in my post above, the process of flashing software to these watches isn't always entirely reliable, and I don't want to be held responsible if/when it doesn't work and the watch is bricked.

Alternative NRF52840/32 Watches compiled list for easy discovery

This is a pretty good start:

https://jeffmer.github.io/WatchApps/
https://github.com/jeffmer/WatchApps

staggeringly few people were willing to pay anything for it

This is so true, I too feel guilty for this. I learned so much using your software, there were over one thousand installations of the eucwatch when I stopped the project, but you got nothing out of it, nor did I, buy maybe a few banglejs sales were lost in the way.

I made a paypal donation a few weeks back, I used my UI for the magic3 for the cat toilet and felt that this time I should return something, I hope it reached you.

I'm interested in the "Alternative NRF52840/32"'s that are available.

There are only a few.
52832:
P22 - similar to P8 and Pinetime, can still be found on aliexpress, hopefully still nrf52832
S28 - relatively new find by @yngv126399 https://www.aliexpress.com/item/1005006404242146.html
both are DaFit so no need to take apart, S28 even has screws so easy to open
52840:
C17/Magic3 https://www.aliexpress.com/item/1005005373233740.html or https://www.aliexpress.com/item/1005005935723945.html

And that's about it as far as I know. I also have few spare unused watches still in a box like QY03/P16, DK08, SN80, GW32 or fitness trackers like B5 that I could sell, if you want to experience the older stuff and possibly continue. But as those are out stock there will be no new users except possibly you.

52840:

this one too (Z004), I haven't tested it, but it looks like a rock/magic3 board?

https://www.aliexpress.com/item/1005007497635449.html

I've kept my work on other watches relatively quiet, as I didn't want to steer anyone away from supporting Gordon monetarily. I think everyone should buy some kind of Espruino hardware to keep the proverbial lights on.
I feel fortunate to have learned enough from folks here HOW to build Espruino for various watches (all nRF52), and for me, the fun is in the discovery of the hardware internals. Like @fanoush I don't wear any of these hacks.

But since much of the cat has left the bag....

I'm working with two others right now:
P4: https://www.aliexpress.com/item/1005007150275911.html - 52832, but DaFit so hackable without opening. It seems to be having power issues (likely needs some pin pulled up/down - not sure). Classic DaFit: shared SPI for LCD and Flash, for which I use @jeffmer 's build to manage. This is something neither Bangle had to contend with, so it was the efforts of folks here who made that happen.
TK78G: https://www.aliexpress.com/item/1005008054139256.html (and above from @enaon )
52840 with tons of goodies like GPS, temperature, HRS, accel, two buttons and touchscreen. I've got the touchscreen working 100%, accelerometer usable (don't know what it is, so not sure what options I can set like changing sensitivity/power settings). I have been trying to figure out the GPS, as I know what pins are being used, but I can seem to turn it on. But if you want to explore, and learn, this is a good one to work with!
And as mentioned above: Pinetime (https://pine64.com/product/pinetime-smartwatch-sealed/) very similar to P8 and very well documented. New ones arrived sealed, which makes things a bit more tricky.. it takes a bit of effort to put the correct bootloader on this for updating to Espruino.

But the only way you'll get a full support watch is Bangle2. Everything else is "roll up your sleeves", and quite a bit of swearing.

. I've got the touchscreen working 100%

Nice! Is it the magic board, or a new one? Is the touch the auto sleeping type? Is the screen brightness more magic3 like or rock like? :)

nm

Not Magic/Rock board, it's totally different, different pins all around. (https://github.com/yngv27/EWatches/tree/main/TK87G).
Screen is ST7789 240x280, very nice, I'd say it was Rock like. Don't know if touch is auto-sleep, I just put it to sleep on screen sleep (I adopted @jeffmer 's wOS object). It's light weight, like the (ironically named) Rock, much lighter than the Magic3.

Not Magic/Rock board, it's totally different

I think you also mentioned before that it is not DaFit despite the listing so needs to be taken apart? @enaon just FYI