-
Hi - there are no immediate plans for a Bangle.js 3, in fact I'm due a delivery of another large batch of Bangle.js 2 in a few weeks so there definitely won't be anything in 2024.
While I'm interested in getting a wishlist for a Bangle.js 3 to see what people like and don't like, and doing one at some point, I want to release some other Espruino-based devices before I even consider a v3 Bangle.
Right now Bangle.js 2 does a lot of things right, and while it can doubtless be improved the cost of developing a brand new device is huge, so I can't afford to just keep pushing new ones out every 2 years unless it's going to be a vast improvement :)
-
-
-
I think I was wrong to write this off - while the
Uncaught undefined
itself is ok, I just flashed a version onto a device and did a Factory Reset, and I found it happened when loading every new app. Was this what you were seeing?I believe the issue has been that the pretokeniser was pretokenising empty strings when it shouldn't have, and this caused an error - which wasn't reported properly because it was inside
bootupdate.js
. Once you updatebootupdate.js
manually it's fixed.It seems the alarm app may have been broken by default anyway. I've just fixed the ide and the app loader, as well as the factory update image in cutting edge builds. Luckily I'm not sure how but it seems that 2v21 (including the factory image) escaped ok
-
-
Ok, I think that's fixed now - realistically the 6x15 font isn't that great anyway. The plan eventually is to try and replace them with the fonts from https://github.com/pebble-dev/renaissance which at least all share the same style
-
-
Sorry to hear that - it's probably better to respond on the post linked above that you just posted on though: https://forum.espruino.com/conversations/394518/#17309516
-
it's 5 reports over the course of about 1.5 years and some thousands of bangles
Yes, it's pretty much 10,000 Bangle.js 2 sold now, and 5 with this problem (and I think one or two more via email). For something that spends every day on your wrist being knocked around it's not that high a failure rate - but as you say it's worth looking at strengthening for a v3.
It's worth noting that most people report their issues publicly on this forum (often in multiple posts), and I make a point of not deleting posts because I think it's important to be honest with everyone.
Unfortunately that means that people come here and see posts on a broken bit of watch body, or bluetooth, or the button, and get the idea that they're unreliable, despite there being a pretty low failure rate. I honestly believe these watches are no less reliable than the competition though, it's just that we don't hide any of the complaints.
-
That's odd - while we have different fonts for different sizes generally the characters should all be the ISO8859 codepage so should display the same, but I just tested with:
g.clear(1).getFonts().forEach((f,i) => { g.setFont(f).drawString(f+": àéèêùç ",0,i*20); });
and I see the attached, so it looks like the 6x15 font has an issue with
à
- but is there other corruption you see?If all software is up to date, you can install this font: https://banglejs.com/apps/?id=fontext
and messages will then use only that font which should render anything. You'll need to update your Bangle.js build to a cutting edge release to allow it to render the font in different sizes though
-
-
-
It should be live on the Play Store testing track now anyway, which you can sign up to with https://play.google.com/apps/testing/com.espruino.gadgetbridge.banglejs
I've tried to submit it to everyone but it's having to go for review again
-
Do you have the bit that came off? Because it's ABS plastic it should stick back on with acetone pretty easily.
One of those '3D pens' might work as well to melt plastic back on.
It's tricky - in Pinetime I think they have more space to mount the strap inboard because they don't have a GPS aerial to worry about. If we extended the Bangle.js 2 case downwards I bet most people would complain it was too big.
It does seem you've been unlucky though - there aren't very many of these watches it's happened to given how many we've shipped out.
-
-
-
Yes, that sounds like a good idea. It's something that would need to be done inside Gadgetbridge though.
I'll take a look and see how easy it is to get rid of stuff like that - however if Google change the format of invites then we'll probably have to change the regex again to try and strip them out.
-
-
-
-
Hi - that's great - thanks for posting up!
As you say 6uA is not such a big deal, but if you did want to reduce it further, as far as I can tell the ADC pin on the Espruino Pico is 5v tolerant (even if it won't measure above 3.3v), so you could attach the GND part of the potential divider to a GPIO, and then let that GPIO float normally and pull it to GND when measuring.
Also putting a small capacitor between A5 and GND can help with ADC measurement accuracy if you need it (they get less accurate when they're measuring a high resistance).
-
-
-
If you're seeing
>Puck.IR(mode)
presumably it's because you were trying to output the IR data?Is the issue that you are trying to record IR data at the same time as sending it, so you're basically recording what you're sending?
I wouldn't worry too much about
FIFO_FULL
though - it's letting you know that some data it was receiving got lost because too much came in too fast. Realistically that's probably not a problem though - worst case the signal doesn't work and you have to re-record it.On new firmwares you could also try JIT which should improve capture speed:
setWatch(function(e) { "jit" d.push(1000*(e.time-e.lastTime)); }, D1, {edge:"both",repeat:true});
I already sent him one - but I imagine there might be someone out there who wants one!