Search results for:

What happens if you connect with the IDE and paste this code in:

GB({"t":"notify","id":1575479849,"src":"Hangouts","title":"A Name","body":"message contents"})

With 'Notifications Fullscreen' I get a fullscreen pop up when I type that in the IDE.
Tried to g.dump() it but that would not work when the pop up is displayed.

However when I disconnect from the IDE and connect through Gadgetbridge - call my phone , nothing happens, when the phone stops ringing nothing happens. Same test works fine if I connect GadgetBridge to my Bangle 1. I get a pop up for a missed call. No notification that my phone is ringing (thats a different issue).

When connected to the B2 I can see that Gadgetbridge is getting battery level notifications from the Bangle 2 as the battery icon says 61% which is what my Bangle 2 is displaying.

I get this with 'Notifications (default)'

>
>GB({"t":"notify","id":1575479849,"src":"Hangouts","title":"A Name","body":"message contents"})
Uncaught Error: setLCDMode is unsupported on this device
 at line 1 col 378
...;Bangle.setLCDMode("direct");let x=0,y=320-size,h=size,b=y+h-1,...
                                 ^
in function "show" called from line 1 col 173
...e("notify").show(currentNot);if(!(require('Storage').readJSON...
                               ^
in function "GB" called from line 1 col 94
...,"body":"message contents"})
                              ^

You said that B2 only works with fullscreen notifications right now - so thats a known issue.

Any tests I can initiate from the phone other than phoning it ?

So I'm thinking of adding a health event once per minute

Sounds a good idea.

Bangle.setSteps/etc

Not so keen on that. It means an App could reset the step count for the day.

We've had a longstanding issue that the step counter is 'steps since Bangle.js started' and
apps like pedometers have to do their own work to calculate steps per day.

Could the firmware not manage the reset at midnight as per current locale setting ?

Apps would be much simpler if we could just do Bangle.getSteps() rather than have to manage a listenner.

3 hrs isn't enough to recreate the problem.
The problem occurs somewhere between 3 to 4 hours.
3 hours works fine, 4 hours does not.

I have never been able to record a log of 4hrs (144K seconds) or over.
I have recreated this 5 times now. I have tried starting with no log at all and also overwriting an existing log on start up.

I use a single press to stop the logging. Is it possible that I could be getting a bouce and restarting the logging ? BUT I would see a new screen starting up saying RECORDING.

I will try a long press to stop the log after 4hrs and report back.

Updated to latest code but I'm still not getting any notifications on a Bangle 2.
Are there any firmware dependancies ? I'm on 2.10.51.

Bangle 2.

Bangle 1 would run out of storage well before 4 hours.

On a B2 I cant get notifications to work at all.
On a B1 Full Screen notifcations work BUT the default notfications ends up with a corrupt part of the screen.

Looks like both versions are broken.

When I make a test call - I only get a missed call notification - was expecting I would get a buzz while the phone was ringing.

Are you using Bangle 1 or Bangle 2 ?

I have not had a lot of success with Gadgetbridge / notifications so far.
Part of this is confusion over what and how it is meant to work.

I have just looked at the code and "Notifications (default)" installs a module called notify that is mean't to pop up a notifcation at the top of the screen. "Notifications Fullscreen" replaces the notify module with a set of library functions that use the full screen instead. So it sounds like there is a fault with "Notifications (default)". This would fit with my observations so far as I have not been able to get notifications to work on Bangle 2.

Looking at the code I can see that 7 days ago changes were done to both versions of the notify module to make it work with Bangle 2. The changes are to adjust colours and screensizes so that they adjust properly for Bangle 1 and Bangle 2. EG the code now uses g.getWidth() to get the screen width instead of hard coding it as 240 for Bangle 1 etc.

Not having any success getting notifications through to the bangle 2.
Can see battery level status of the bangle on Gadgetbridge on the phone.
The Widget Settings says I am connected.
Can use the find phone capability and make the phone ring.
So seem to be connected but when I call my mobile I am not getting any notifications.
Have checked phone / App permissions for Gadgetbridge and access to phone / SMS is enabled.

The Anton and Wave Clock seem to be using easch others icons.

Just downloaded widpedom 0.19 and last full screen notifications.
Still getting a problem with the widget size when the gadgetbridge icon displays.
Have yet to test notifications.

Do I need gadgetbridge installed or does full screen notifications suffice. Its not very clear if Full Screen Notifications is all you need.

Shame about the housework one though.

Here's the baseline results through the test harness.
we can see 102% (+2%) accuracy on the 10K step log.
But 277.97 % (+177% over) on the housework log.

X_STEPS = 6, RAW_THRESHOLD = 14
File, Expected, Simulated, Diff, %, (Orignal)
HughB-walk-6605.csv, 6605, 6397, -208, 96.85 %, (3223)
HughB-walk-2350.csv, 2350, 2243, -107, 95.45 %, (1042)
HughB-walk-a3070-b3046.csv, 3070, 3013, -57, 98.14 %, (1909)
HughB-walk-a10021-b10248.csv, 10021, 10253, 232, 102.32 %, (12222)
HughB-drive-36min-0.csv, 0, 160, 160, 0.00 %, (1199)
HughB-drive-29min-0.csv, 0, 192, 192, 0.00 %, (1153)
HughB-drive-a3-b136.csv, 3, 124, 121, 4133.33 %, (535)
HughB-work-66.csv, 66, 97, 31, 146.97 %, (980)
HughB-work-66.csv, 66, 97, 31, 146.97 %, (980)
HughB-mixed-390.csv, 390, 541, 151, 138.72 %, (1871)
HughB-general-a260-b573.csv, 260, 578, 318, 222.31 %, (3854)
HughB-housework-a958-b2658.csv, 958, 2663, 1705, 277.97 %, (5762)
TOTAL DIFFERENCE 4013

Now it is possible to improve the housework log to +32% over by increasing the thresholds.
BUT this is at huge cost to the overall accuracy when actually walking.
I would not be happy with only 85% accuracy over a 10,000 step walk. Needs to be 98-102%.

X_STEPS = 8, RAW_THRESHOLD = 17
File, Expected, Simulated, Diff, %, (Orignal)
HughB-walk-6605.csv, 6605, 6012, -593, 91.02 %, (3223)
HughB-walk-2350.csv, 2350, 2198, -152, 93.53 %, (1042)
HughB-walk-a3070-b3046.csv, 3070, 2891, -179, 94.17 %, (1909)
HughB-walk-a10021-b10248.csv, 10021, 8542, -1479, 85.24 %, (12222)
HughB-drive-36min-0.csv, 0, 13, 13, 0.00 %, (1199)
HughB-drive-29min-0.csv, 0, 46, 46, 0.00 %, (1153)
HughB-drive-a3-b136.csv, 3, 52, 49, 1733.33 %, (535)
HughB-work-66.csv, 66, 56, -10, 84.85 %, (980)
HughB-work-66.csv, 66, 56, -10, 84.85 %, (980)
HughB-mixed-390.csv, 390, 356, -34, 91.28 %, (1871)
HughB-general-a260-b573.csv, 260, 226, -34, 86.92 %, (3854)
HughB-housework-a958-b2658.csv, 958, 1266, 308, 132.15 %, (5762)
TOTAL DIFFERENCE 3671

So I think without a radical rethink this is the limit of what can be acheived with this algorithm.

But on the screen itself it showed that it had recorded lots of samples?

Yes - 144K+ seconds worth of samples.

If you toggle it on and off when you're connected to the IDE, do you see any errors?

None

And does require("Storage").getFree() show that there's a decent amount of free space?

>require("Storage").getFree();
=3775872

I am deleting old logs as I go along so that I never risk running out of space.
I will need to run for 4 hrs again to check the value of getFree() at the time I want to stop the recorder.

Can you try a test and see if you can get a 4 hour log ?

Here's the baseline set of test results for the step counter in the current firmware (2.10.27 and beyond).

Ran again for 150K seconds (4 hrs plus).
Before I stopped the logger I connected through the IDE to see if there were any errors.
There was nothing displayed when I selected the STOP button.
When I went to view the log it said there were 2 records - when downloaded it looked as follows.
It looks like the first record only was recorded.

Hi @Gordon - run into a problem with accellog between 3-4 hours.

I have just been away on a 4 day walking and aimed to get some 10,000+ steps logs.
I managed to get one log of 11450 seconds (3hrs 10 mins).
But on two other days when I stopped the logger after more than 3.5hrs and viewed its size it came out as 2 records only.
I then repeated a test of just running the logger up to 4 hours and can recreate the issue.
The Accelog app will keep incrementing the sample and second count but it appears that after a while
it reaches some limit that causes a problem. Not sure if this is an issue with the storage on a B2 or with accellog itself.

Agree, ease of use, accessibility, consistency, then clarity and elegance.

Does this make sense to you too?

Not for me.
a) BTN3 is a reset and only largely used as many Apps dont provide a proper way to return to the launcher. Thats my interpretation.

b) a long press also means longer navigation to get out of the App. Instead of selecting Back using BTN1 all the way out of a deep sub menu and finally exit you would now have to go BT1, BT1, BT1, long press BT3, which is just as awkward as having to navigate to the last back menu item on the current settings app.

Agree, think this comes down to a readable built in font of the right size.

6x8 font scaled 1x2, but I'm not entirely sure it's an improvement.

There is a small improvement.
The existing font height is 15.
The 6x8:1x2 is 16, it looks heavier and hence easier to read, it results in 7 lines rather than 8.

Here are the 6x8:1x2 (top) and the existing (bottom) screen dumps.

Conclusion - some marginal improvement by switching to 6x8:1x2 - at low cost effort to implement.

6x8 font scaled 1x2, but I'm not entirely sure it's an improvement.

6x8 :1x2 results in a font size of 16 high - so its still around the same size as the existing font.
I think we need a font height of 20 to be readable (for me anyway).

I tried setFont('6x8:1x3') - the results looks ok in the screenshot not so readable on the wrist.

@MaBe - I think what you mean is the spacing (black) between the line area (yellow) as in the screenshot below which is from an Apple Watch (BTW I don't have one, just grabbed a screenshot off a YouTube video). The advantage of what you suggest is that the text might still fit on the line. But if you are struggling to read the text in the first place there is no change to the font.

Personally if the selected option line is highlighted using contrasting colours I'd rather have a larger font than a dead region between lines.

The interesting thing about the Apple screenshot is that they have gone for 4 lines in the scrolling window for lists. The Apple watch settings works the same way, when you select the line you jump to its options list; though some have selector slider switches for on/off etc in the same line.

'g.dump()' in the IDE and it'll give you a nice screenshot

Gave it a try, really good. Thank you for reminding me of this. I remember the announcement on a different thread but never checked it out. I remember you said it was a bit slow, but it did it in less than 1 second. Thats going to save a lot of time for me, when I do a README file.

I think Vector font, size 20 was significantly more readable.
Could it be made to work with options 2 or 3 above ?

Are we thinking we change it on the first menu in settings too?

In my opinion Yes - not having it there means extra navigation to exit settings.

Sounds brilliant.

the difference between "make connectable" and the activation in ON of the bluetooth?

I'm not sure. I never use it.

fixing the "back" selection, whether it is always the first or the last.)

So the proposal is for always First. This is because pressing back from a submenu - selects the first entry in the parent menu on return. In this way you can move back up through the sub-menu's and exit just by pressing button 1, instead of sometimes having to navigate to the bottom option.