Search results for:

Its fixed and works fine. Well done ;-)
Connection issue did happen one time only today.
Usual problem, if you want to show strange behaviour, it never happens :-(
Poor signal should not be the problem, distance between board and laptop is less than 20cm

Hmm, some more (may be helpful) info
One more problem I have around is disconnecting of UART on MDBT42Q.
Did not encounter this with Puck (yet?)
From time to time it happens, that connection is kind of frozen. There is no echo in terminal window anymore. This happens at crlf or in the middle of a command.
Or, another problem, sending even some lines of code, less than 3kb, from source window to Espruino stops somewhere in the middle.
And one more, sometimes after reset, startinfo comes in very slow only, char by char.

Just seen new version(V2.06.81) in binaries.
I did some extended testing with other boards

board V2.04 V2.06
========= ===== =====
Espruino ok ok
ESP32 ok ok
MDBT42Q ok bad
Puck ok bad

Based on that, my first guess would be somewhere around implementation of UART for BLE

Tested with Android and nrf
Got same problem, so its not a windows problem.
With Android its even worth, first time it takes >30 secs

Hmm,
tested with 2V06,78 from http://www.espruino.com/binaries/travis/master/
Still same problem.
Latest entry is from 2020-07-30
Looks like there is no update since end of July

During testing a rotary phone, I ran into problems recognizing pulses.

After some testing brought it down to following samples.
Duration is always much more than expected (by me).

Interesting is last example where only first timeout is wrong.

Testing is done with WebIde. First examples use terminal window for input.
Last testing is uploaded from editor window and called from terminal.
Board is MDBT42Q with 2v06, downloaded from binary

for(var i = 0; i < 10; i++) {var dt = new Date();setTimeout(function(){console.log(new Date() - dt);},1000);}
=undefined
2159.7900390625
2161.0107421875
2162.90283203125
2164.21508789062
2165.67993164062
.....

var dt = new Date();setTimeout(function(){console.log(new Date() - dt);},100);
=1
958.46557617187
>var dt = new Date();setTimeout(function(){console.log(new Date() - dt);},100);
=1
1476.01318359375
>var dt = new Date();setTimeout(function(){console.log(new Date() - dt);},100);
=1
1280.94482421875
>var dt = new Date();setTimeout(function(){console.log(new Date() - dt);},1000);
=1
1865.69213867187
>var dt = new Date();setTimeout(function(){console.log(new Date() - dt);},1000);
=1
2315.673828125

function tst(duration){
  var pnt = 0;
  function tstTimeout(){
    var dt = new Date();
    setTimeout(function(){ 
      console.log(new Date() - dt);
      if(pnt++ < 4) tstTimeout();
    },duration);
  }
  clearTimeout();
  tstTimeout();
}
tst(1000);
=undefined
1828.857421875
1001.43432617187
1001.46484375
1001.43432617187
1001.46484375

Hmm, I will check. Thanks for this hint

Sounds like an OGP (old guys problem), that I would like to get rid off.
Problem: If I have to synchronize serveral places, sooner or later this will fail.
Anyway, for now, there is no nice solution. I will follow the easy way, as you recommended.

Hmm, good answer ;-)
I'm working to get efeu running again.
For efeu, a lot of stuff, like compilers and precompiled libs, are moved to a special folder.
ESP8266/32 use exported pointer to this folder.
This makes it easier to support several Espruino versions (edge, release, develop).
Do you have a nice way to move targetlibs to my special folder and have an exported pointer to that ?
Or would this be too much effort, having in mind, efeu would be the only tool to benefit from that?

targetlibs holds a lot of nrf relevant files.
patches to these files (based on $PROVISION_NRF_SDK15) are installed in provision.sh
Is there a special reason why these changes are not directly in targetlibs ?
Is there a reason, not to install on default ?

@parasquid, which version of ESP-IDF did you use ?

Sorry, misunderstanding.
'y' means it should skip the error/reset in case of missing PSRAM

The only way, I know is to change sdkconfig

CONFIG_SPIRAM_IGNORE_NOTFOUND=y should manage the problem

May be pico board is not supported well from esp-idf ?

Sorry, did not take care on EFEU for some time, since feedback was close to zero.
Will take a closer look soon, and do an update.

There is a nice test of bangle https://www.golem.de/news/bangle-js-im-test-die-smartwatch-fuer-bastler-2006-148951.html
It's a german page, it starts with:
The Bangle.js by Espruino inventor Gordon Williams is charmingly different from conventional Smartwatches - and invites you to tinker and program.
BTW, I've something in mind about replacing watchstrap (?)

How could I reset bangle to clock mode after reset ?
Steps I did:

• connected with web ide, bangle shows clock
  
• send reset() from terminal window, bangle shows logo
  
• disconnected web ide, still shows logo
  
  

Only (?) click btn3 on clock brings me back to clockmode.
Is there any other way ?

@paulc,
you brought me to the right train ;-)
There is an option for tensorflow_converter, --quantization_bytes=1 (or 2)
May be, this will help to reduce the size. At least it should reduce size of bp file in H5

Switched scripts to :

https://cdn.jsdelivr.net/npm/@tensorflow/tfjs@2.0.0/dist/tf.min.js
https://cdn.jsdelivr.net/npm/@tensorflow/tfjs-vis@1.1.0/dist/tfjs-vis.umd.min.js

Training runs fine with updatetd scripts
Next converted output from JSON-Format to Keras H5 format.

tensorflowjs_converter --input_format=tfjs_layers_model --output_format=keras ./model.json ./model.H5

Then converted Keras to tflite file

tflite_convert --output_file c:\temp\bangletf\Model.H5\lite.tflite --keras_model_file c:\temp\bangletf\Model.H5

Now I've a tflite file, but cannot test. I'm not at home, but thats where my bangle is :-(

@PaulC
Actual converter does not accept model file, created by your app, any more.
Digging deeper, saved model has an array, where converter expects an object.
Going back to your app, you use tfjs@1.0.0, actual version is @2.0.0
Did you check/try/have_experience/whatever with actual version ?

To give an update about my next steps:

• recording and save samples locally works fine
  
• train samples and store resulting model locally works fine. Two files are stored, first is a JSON file with a lot of description, second is a binary with weights.
  
  

Ok, whats now. Tool from @PaulC takes the model, reads gesture data from bangle and recognizes gesture in browser. What we would like to do is to have both steps, read gesture data and recognize gesture on bangle. Therefore we need a TFLIE model.

As @paulc mentioned there is no way to convert TFJS model to TFLITE.
First I tried to find something like that in TFJS, no success.

Next installed tflite_converter for python.
Failed on my UBUNTU server, but got it running on windows.
But again, no success, tflite_converter expects model in pbtxt or pb file.

Next found another converter which converts tfjx model to Kera model, which then could be converted tflite (?).
Startet tensorflowjs_wizard, and ....
Unfortunately, this converter runs into an error
AttributeError: 'list' object has no attribute 'items'
No idea what that is, but now its weekend :-)

@ConorONeill,
thanks for your hint. I see the problem to understand what a gesture is.
@brainfart-be added a chart in his tool, which could be a first step to visialize a gesture.

I could like to have something like a 3d drawer to show gesture. But did'nt find anything yet.

I'm working on my own version based on @PaulC and @brainfart-be, mainly to get a first understanding. Recording part and save in local files (not in local storage) is running in a first version.

Wow, GREAT!!!
Just tested drawing circles above my head and in front of my stomach.
Works fine on first test.
So, at least one member of the community is interested in that ;-)
What I would like to have is

• an open database which collects samples for gestures also from other users
  
• collect my own set of gestures for my app and train based on database
  
• generate a TFLite model
  
  

For the beginning, testing etc. we could use my shared server.
Feel free to contact me directly, if you are interested.

From time to time I spent an hour or so for speech recognition.
There is an example https://github.com/tensorflow/tensorflow/tree/master/tensorflow/lite/micro/examples/micro_speech for tensor flow lite.
One of my open questions is microphone.

• simple microphone would expect fast A/D conversion.
  
• I2S microphone (like INMP441) needs an I2S interface.
  At the end, both solutions need lot of memory, which will not be available on all boards.
  Last not least, see problems with adding tensorflow on MDBT42Q board in other chat.
  Any nice idea, how it could go on ?
  

Same from my side, I do not want a refund.
Its not your fault, and last not least, in times like this we have serious problems.
Nobody should waste time for peanuts like a refund.

@Jonaswinz
Are you working with something like this https://www.aliexpress.com/item/32653944927.html ?
AFAIK, you would first need a driver for this board before using something like SmartMatrix
For ESP32 there was a solution from ESP_SPRITE, see https://esp32.com/viewtopic.php?f=17&t=3188
I've got it working for Espruino some time ago, see https://github.com/jumjum123/EspruinoExtensions/tree/master/Family/ESP32/LED32x64
Did not follow , but have in mind to need a ESP32 board with PSRAM