#circuitpython-dev

1 messages · Page 68 of 1

onyx hinge
#

as for the pauses trying to download an mp3 from http .. I don't know if my local network is particularly bad .. I stream audio and video fine all the time on wifi, but that's on the 5GHz and not the 2.4GHz signal.

slender iron
#

I think it's good to test with that

#

hiccups are fine IMO

#

its crashes and stops that would be good to fix

manic glacierBOT
urban knoll
#

Is there a built-in routine for setting up a debug uart? Not seeing one in the repo

slender iron
#

looks it up

#

CIRCUITPY_CONSOLE_UART_RX and _TX

#

console_uart_printf() will print to the uart only

#

mp_printf(&mp_plat_print, <normal printf>) will go out all serial

urban knoll
manic glacierBOT
manic glacierBOT
#
Timeline8
[adafruit/circuitpython] New comment on issue #9291: ESP32-S3 problem with adafruit_thermistor.mpy library

Yes, I still have the WiFi settings auto connecting like that. We did talk about disabling that over at the forums but I don't think I tried it. So that is now on my to-do list for tonight: Run both the Waveshare S3 and Feather TFT S3 without the .toml file.

If there is a conflict between the two (ADC and WiFi), that will be disappointing for me since I need both for my application, so I won't be able to simply not use WiFi. But the problem can't be fixed later if we don't narrow it down ...

#
dhalbert
[adafruit/circuitpython] New comment on issue #9291: ESP32-S3 problem with adafruit_thermistor.mpy library

I was just testing on a QT Py ESP32-S3, with the very simple test program below. With CIRCUITPY_WIFI_SSID and CIRCUITPY_WIFI_PASSWORD commented out in settings.toml, it runs indefinitely, with millions of conversions. With them not commented out (so web set up in advance) , it crashes hard almost immediately. I am using board.A2, which uses ADC1 on the QT Py S3.

So now I know how to reproduce this. It is strange because ADC2 is supposed to be shared with WiFi, not ADC1. But mayb...

#
bill88t
[adafruit/circuitpython] Pull request opened: #9309 Update readme to show the support status of the new esps

H2 and C2 are alpha.
C2 outright "halts and catches fire" in some of my tests.
Memory/instruction errors, heap errors, all sorts of goodies.
I will be filing issues once I hopefully make repro code.

C3 and C6 are still fine being beta.
There is a lot of weirdness that isn't easily reproducible and can't be filed atm.
Also thread/zigbee is not implemented.

S3 is just as stable as S2 and ble is somewhat implemented now right?

OG ESP32 still not stable due to serial.

manic glacierBOT
manic glacierBOT
#
Timeline8
[adafruit/circuitpython] New comment on issue #9291: ESP32-S3 ADC use causes crashes when WiFi in use

Hi Dan, Thank you very much for confirming this. Since I first reported this on the forums and after days of replies there and here, since no one seemed to be interesting in actually running the code to see if it could be reproduced, I was starting to wonder if it was me and everyone was just being polite by not telling me I'm the idiot. ;)

You and I actually had a conversation on the forums on the Wifi vs ADC about two months ago because I had read about the possible interference bet...

manic glacierBOT
humble shadow
#

@slender iron hey, I came across an old form post of yours from 2016, I think I’m having the same issue. I have a fork of CircuitPython in my GitHub. I want to add my board to Micropython so I cloned it, made my changes, then tried letting GitHub desktop create a fork in my repository so I can make a pull request; however, GitHub hits me with a “nah man, this is actually circuitpython” and I’m like, “No, GitHub, you’re drunk, this is most definitely Micropython. Listen to me, I want you to make a fork of Micropython.” but git’s all like, “Sorry man, couldn’t hear you over the 7000 conflicts I’m trying to resolve.”

Anyway, if you remember how to fix that, I would greatly appreciate it lol.

manic glacierBOT
#
dhalbert
[adafruit/circuitpython] New comment on issue #9291: ESP32-S3 ADC use causes crashes when WiFi in use

@Timeline8 Rest assured we were interested, but if we can delegate some testing to eliminate possibilities, then we try that. (We have all too many bugs to look at :slightly_smiling_face:). For instance, was displayio involved or not? And I was hoping it was really fixed in 9.1.0-beta.3, since we'd upgraded the underlying Espressif software (ESP-IDF) in that release. I also spent time looking for similar reports in the ESP-IDF repo, but could not find any.

I hope we can figure this out s...

#
dhalbert
[adafruit/circuitpython] New comment on pull request #9311: Web workflow: use absolute paths to avoid paths being adjusted by current directory

@tannewt One confusion I had when tracking this down was what MP_VFS_ROOT meant, as returned by mp_vfs_lookup_path() in extmod/vfs.c. It seems to mean only /, and not the whole default filesystem. I'm thinking that even CIRCUITPY is a VFS that is mounted on MP_VFS_ROOT. So it seems MP_VFS_ROOT is only the / mount point, and not the filesystem itself. Do you remember the details here?

orchid basinBOT
wraith crow
# humble shadow <@252717193496756235> hey, I came across an old form post of yours from 2016, I ...

I am not a GitHub expert and don't use GitHub desktop but in order to create pull requests for both Micropython and Circuitpython from my repo, I've had to fork Micropython and then use a branch of that fork for Circuitpython PRs using the following:

git remote add circuitpython https://github.com/adafruit/circuitpython.git
git fetch circuitpython
git checkout remotes/circuitpython/main
git switch -c projectName #Creates new branch 'projectName' in Micropython fork
git push --set-upstream origin projectName #I guess this one actually creates the branch 😁

GitHub
GitHub - adafruit/circuitpython: CircuitPython - a Python implement...

CircuitPython - a Python implementation for teaching coding with microcontrollers - adafruit/circuitpython

#

You have to fumble a bit with the branch comparison screen when submitting a CircuitPython PR using the web github interface because it defaults to trying to submit the PR to Micropython.

manic glacierBOT
#
juansulca
[adafruit/circuitpython] Issue opened: #9312 QT PY RP2040 Pins A0-A3 not capable of enabling internal pull up/down resistors

CircuitPython version

Adafruit CircuitPython 9.0.5 on 2024-05-22; Adafruit QT Py RP2040 with rp2040

Code/REPL

import board
from digitalio import DigitalInOut, Direction, Pull

btn_0 = DigitalInOut(board.D0)
btn_0.direction = Direction.INPUT
btn_0.pull = Pull.UP

btn_1 = DigitalInOut(board.D4)
btn_1.direction = Direction.INPUT
btn_1.pull = Pull.UP

Behavior

When I measure the voltage between GND and pin D4, it is set to +3.3v as expected b...

#
RetiredWizard
[adafruit/circuitpython] New comment on pull request #9311: Web workflow: use absolute paths to avoid paths being adjusted by current directory

I only looked at the changes so I don't have the context for what the variables are, but the comments are talking about mount points and the issue seems to be more general as simple sub-directories exhibited the issue.

I did test the artifact and it resolved the problem demonstrated by the test code.py in the original issue, however now I don't think a file can be deleted using web workflow if the folder has been changed from the root folder using os.chdir.

#
dtorres-sf
[adafruit/circuitpython] New comment on issue #9302: EINPROGRESS when trying to connect to wifi the first time each day

Last night I did not leave the USB connected to the docking station and this morning I did a full power cycle (disconnecting the battery) and started back up. It appeared to still hit the issue. I cannot be 100% certain since I wasn't connected to USB/serial to see the stack trace, but it hit the assert at the exact same time it usually does (and the app running has been unchanged for months and is pretty stable other than this issue).

orchid basinBOT
manic glacierBOT
#
bill88t
[adafruit/circuitpython] New comment on issue #9291: ESP32-S3 ADC use causes crashes when WiFi in use

I just ran into this with the cardputer.
Spamming the adc with wifi connected produces instant resets and hangups.

I just finished the battery driver, which uses IO10 of the ESP32-S3.
When connected to a network, after about 30 seconds it resets (NO SAFEMODE, RESET) or hangs up (neopixel went white, not a color I use, and stayed there, unresponsive).

The polling rate was 30 samples/s.

The code accounts for None reads.

manic glacierBOT
#
bill88t
[adafruit/circuitpython] New comment on issue #9291: ESP32-S3 ADC use causes crashes when WiFi in use

Attempted this patch, which reduces the points of failure:

--- a/ports/espressif/common-hal/analogio/AnalogIn.c
+++ b/ports/espressif/common-hal/analogio/AnalogIn.c
@@ -115,10 +115,10 @@ uint16_t common_hal_analogio_analogin_get_value(analogio_analogin_obj_t *self) {
     #endif
 
     uint32_t adc_reading = 0;
-    size_t sample_count = 0;
+    int sample_count = 0;
     // Multisampling
     esp_err_t ret = ESP_OK;
-    for (int i = 0; i < NO_OF_SAMPLES; i++) {
+    whi...
manic glacierBOT
#
bill88t
[adafruit/circuitpython] New comment on issue #9291: ESP32-S3 ADC use causes crashes when WiFi in use

Yep, fun stuff:

I (21616) EXAMPLE: ADC1 Channel[3] Raw Data: 562
I (21616) EXAMPLE: ADC1 Channel[3] Cali Voltage: 481 mV
I (21626) EXAMPLE: ADC2 Channel[0] Raw Data: 537
E (21626) task_wdt: Task watchdog got triggered. The following tasks/users did not reset the watchdog in time:
E (21626) task_wdt:  - IDLE0 (CPU 0)
E (21626) task_wdt: Tasks currently running:
E (21626) task_wdt: CPU 0: main
E (21626) task_wdt: CPU 1: IDLE1
E (21626) task_wdt: Print CPU 0 (current core) backtrac...
slender iron
humble shadow
#

Thanks @wraith crow & @slender iron !

manic glacierBOT
#
bill88t
[adafruit/circuitpython] New comment on issue #9291: ESP32-S3 ADC use causes crashes when WiFi in use

I got my YD-ESP32-S3 which is a dual usb-C board, and is excellent for debugging and built some debug builds.

import wifi, board, analogio;wifi.radio.connect("SSID", "PASSWD");a=analogio.AnalogIn(board.GPIO10)
while True:
    a.value

This reliably crashes it. Maximum 15s.

Decoded backtrace of debug build (clean tree, current master):

0x4037cc7a: ram_chip_i2c_readReg at ??:?
0x40378aa0: regi2c_ctrl_write_reg_mask at /home/bill88t/git/circuitpython/ports/espressif/...
manic glacierBOT
#
Timeline8
[adafruit/circuitpython] New comment on issue #9291: ESP32-S3 ADC use causes crashes when WiFi in use

@dhalbert, no problem on me doing some of the upfront testing. I know from my own experiences learning CircuitPython and also reading other people pleas for help, 99% of the time it is user error. So no harm on pushing back a bit on the user to kind of "prove it".

As for my project, it is for aquarium monitoring and eventually some controls later. Therefore I am sensing water temperature. Thermistors are well suited for this. Simple to use, and easy to waterproof if needed. I would love t...

manic glacierBOT
manic glacierBOT
#
tannewt
[adafruit/circuitpython] New comment on pull request #9311: Web workflow: use absolute paths to avoid paths being adjusted by current directory

@tannewt One confusion I had when tracking this down was what MP_VFS_ROOT meant, as returned by mp_vfs_lookup_path() in extmod/vfs.c. It seems to mean only /, and not the whole default filesystem. I'm thinking that even CIRCUITPY is a VFS that is mounted on MP_VFS_ROOT. So it seems MP_VFS_ROOT is only the / mount point, and not the filesystem itself. Do you remember the details here?

I don't remember. Sorry!

#
dhalbert
[adafruit/circuitpython] New comment on pull request #9303: picow: turn off CIRCUITPY_USB_HOST and CIRCUITPY_PICODVI

@tannewt

Board PICODVI (off by default) USB_HOST (on by default) notes
cytron_edu_pico_w + + lots of frozen modules, no display
pajenicko_picopad + - gaming board with display
pimoroni_badger2040w - - e-ink badge
pimoroni_inky_frame_5_7 - - e-ink picture frame
pimoroni_inky_frame_7_3 - - e-ink picture frame
pimoroni_pico_dv_base_w + + point of board is DVI , non pico-W build availa...
manic glacierBOT
manic glacierBOT
orchid basinBOT
manic glacierBOT
manic glacierBOT
#
bill88t
[adafruit/circuitpython] New comment on issue #9291: ESP32-S3 ADC use causes crashes when WiFi in use

Doing the adc sampling into the core also doesn't fix this.
There is nothing that can be done from the CircuitPython side to workaround this.

This needs to be fixed from ESP-IDF.

Here is my experimental patch with an updated ADC api that internalizes the sampling.
(So that python doesn't loop over adc reads)
<details>
<summary>Patch</summary>

diff --git a/locale/circuitpython.pot b/locale/circuitpython.pot
index 0e3878bc3d..407d3713a9 100644
--- a/locale/circuitpyt...
manic glacierBOT
tulip sleet
#

@brazen hatch I am baffled as to why your latest PR says "First-time contributor". Did you change something about your account?

brazen hatch
tulip sleet
#

did you switch to "Pro" recently?

brazen hatch
#

nope, I literally have touched / changed nothing on my acc

manic glacierBOT
brazen hatch
#

i've always been "pro"

tulip sleet
#

that is a weird glitch

brazen hatch
manic glacierBOT
manic glacierBOT
manic glacierBOT
orchid basinBOT
manic glacierBOT
manic glacierBOT
manic glacierBOT
manic glacierBOT
manic glacierBOT
manic glacierBOT
stuck elbow
#

what is the status of the paralleldisplaybus on the esp32-s3 boards? should it be working?

tulip sleet
stuck elbow
manic glacierBOT
#
Timeline8
[adafruit/circuitpython] New comment on issue #9291: ESP32-S3 ADC use causes crashes when WiFi in use

@bablokb , Yes I am aware of the DS18B20 but I just happen to have a LOT of thermistors on hand that are waterproof, so that is one reason. No matter how cheap I can find DS18B20s, no one is going to beat FREE. 😉 While the DS18B20 are fairly simple to use once you get them setup, even they are no match for the dead-simple lowly thermistor. Plus I have seen enough reports of less than reputable supplies (counterfeits?) of DS18B20 being on the market that gives me pause.

But fundamentally, t...

slender iron
stuck elbow
#

I'm getting espidf.IDFError: Requested resource not found -- are there any restrictions on which pins can be used?

#

or do I need to configure memory in settings.toml?

slender iron
#

no memory config needed

stuck elbow
#

weird, after a reset I stopped getting that error

stuck elbow
#

oh, I see there it t-display s3

slender iron
#

ya, the s3 version

#

could have broken since I did it though

#

did you release displays?

#

maybe that's their version of in-use

stuck elbow
# 
displayio.release_displays()
bus = paralleldisplaybus.ParallelBus(
    chip_select=board.IO43, # CS
    command=board.IO44, # RS
    write=board.IO36, # WR
    reset=board.IO18, # REST
    data0 = board.IO6,
)
display = busdisplay.BusDisplay(bus, _INIT_SEQUENCE, width=320, height=240)
#

it's an ili9341 that I had working with rp2040 before

manic glacierBOT
#
justmobilize
[adafruit/circuitpython] Issue opened: #9317 Getting a `Hard fault` when using SSL with a WIZnet5k

CircuitPython version

Adafruit CircuitPython 9.1.0-beta.3 on 2024-05-22; FeatherS3 with ESP32S3

Code/REPL

import alarm
import board
import busio
import digitalio
import time

import adafruit_connection_manager
import adafruit_requests
from adafruit_wiznet5k import adafruit_wiznet5k

# delay so you can connect
print("starting:", end="")
for i in range(10):
    print(".", end="")
    time.sleep(1)
print()

print("getting radio")
chip_select ...
manic glacierBOT
orchid basinBOT
#
makermelissa
[adafruit/circuitpython-org] New comment on pull request #1420: Add ThingPulse Pendrive S3

Thank you so much! I was hoping that this pull request would also allow people to flash the board directly from https://circuitpython.org/board/ThingPulse_PendriveS3/ How can I get that section to show up?

You need to have the board family and bootloader id, which it looks like you have. You'll need to add your board to https://github.com/adafruit/tinyuf2, which I don't see in the latest release, but maybe you added it already. And it looks like you'll need to wait for another release of...

manic glacierBOT
lone sandalBOT
orchid basinBOT
manic glacierBOT
#
deshipu
[adafruit/circuitpython] New comment on issue #8942: espcamera ESPIDF Generic Failure and Memory Error on ESP32S2

Have you tried this with espcamera and CP9? I'm just curious if you will have the same issues with memory.

Sorry for the delay, I didn't see your comment. I just tried with 9.1.0-beta on a Lolin S2 Mini, with this code:

import busio
import espcamera
import board

i2c = busio.I2C(scl=board.IO35, sda=board.IO33)
data_pins = (
    board.IO21, board.IO17, board.IO16, board.IO18,
    board.IO37, board.IO34, board.IO36, board.IO39,
)
c = espcamera.Camera(
    data_pins=data...
#
justmobilize
[adafruit/circuitpython] New comment on issue #9317: Getting a `Hard fault` when using SSL with a WIZnet5k

and verified. This doesn't crash:

import alarm
import board
import busio
import digitalio
import time

import adafruit_connection_manager
import adafruit_requests
from adafruit_wiznet5k import adafruit_wiznet5k

# delay so you can connect
print("starting:", end="")
for i in range(10):
    print(".", end="")
    time.sleep(1)
print()

print("getting radio")
chip_select = digitalio.DigitalInOut(board.D10)
spi = busio.SPI(board.SCK, board.MOSI, board.MISO)
radio = a...
orchid basinBOT
manic glacierBOT
manic glacierBOT
#
dhalbert
[adafruit/circuitpython] New comment on issue #9291: ESP32-S3 ADC use causes crashes when WiFi in use

I think it starts executing arbitrary code.

That's not what I'm seeing (using a Metro ESP32-S3)

import analogio
import busio
import board
import digitalio

led = digitalio.DigitalInOut(board.LED)
led.switch_to_output()

analog = analogio.AnalogIn(board.A5)

uart = busio.UART(tx=board.TX, baudrate=115200)

count = 0
while True:
    count += 1
    if count % 100000 == 0:
        print("count", count)
    if count % 100 == 0:
        led.value = not led.value
    ...
#
bill88t
[adafruit/circuitpython] New comment on issue #9291: ESP32-S3 ADC use causes crashes when WiFi in use

With my example:

while True:
    adc.value

It either halts or safemodes.
I should note, I only tested from REPL.

If proc'ed by my os and the "halt" scenario occurs, the neopixel, instead of heartbeating in 2 different blues, it becomes 255-value white (not the REPL white). USB in this scenario is disconnected.

Also I have already determined the watchdog timeout is cause of the calibration init functions.
Under the adc pr, it doesn't happen, since those only run once dur...

brazen hatch
#

@tulip sleet would it be useful to video the different crash scenarios?

Also sidenote, the cardputer has the screen. If it crashed to REPL, the displayio root would swap back to CP terminal.

tulip sleet
#

it's not crashing to the REPL. It's continuing to run. It just loses USB connectivity. I see that and I also see the Internal Watchdog safe mode, both when using an ADC1 pin.

#

I don't think we need a video. I am reproducing it. I am going to look at the changes in the underlying code from when it used to work.

brazen hatch
tulip sleet
#

yes, I have seen that too

brazen hatch
#

this is what I refer to as the "halt" scenario, that the pr doesn't fix

tulip sleet
#

the UART going to garbage is really strange. I saleae'd it to make sure but I I can't tell the new baud rate or if it is wandering

#

something changed in ESP-IDF. I'm surprised no one else outside of us has seen this.

manic glacierBOT
#
Timeline8
[adafruit/circuitpython] New comment on issue #9291: ESP32-S3 ADC use causes crashes when WiFi in use

If proc'ed by my os and the "halt" scenario occurs, the neopixel, instead of heartbeating in 2 different blues, it becomes 255-value white (not the REPL white). USB in this scenario is disconnected.

I would see this mode when the failure was not reporting an internal watchdog timer expired. When I had the TFT display hooked up or running my S3 TFT Feather, the board disconnects from USB, the Neopixel turns solid white, and no reconnection of the board to USB. However the TFT shows the co...

brazen hatch
#

imma do it on a s3 tft feather too

#

the one with the replaced screen

short tendon
tulip sleet
brazen hatch
#

Got a watchdog timeout while wiping storage.. huh...

#

I think the idf isn't feeling so good..

#

latest artifact

tulip sleet
#

idf has just updated to 5.2.2 from 5.2.1, but nothing in the release notes that looks pertinent. Nevertheless Scott will be updating and we'll re-test

brazen hatch
#

It's mainly just bug fixes.
I could go and try to merge it downstream for testing..

short tendon
tulip sleet
brazen hatch
# tulip sleet it is in or will be in an upcoming PR from Scott

Nothing has been done in the CP side yet, since last I checked at least.
I was thinking of just merging all the commits into an idf fork branch and pointing my local CP clone to use that.
Just a yolo merge. It wouldn't be the first time i'd done that.

manic glacierBOT
tulip sleet
brazen hatch
#

Now for some reason that I got pwm running it's not crashing..

tulip sleet
#

you stay up awful late 🙂

brazen hatch
#

it's 1 am, we have time till 5

#

😇

tulip sleet
#

do you usually stay up until 5

brazen hatch
#

yep, I can focus well at night

brazen hatch
#

Yep, it's one of those bugs.

tulip sleet
#

what does it go to from 500 Hz?

#

or does it start to wander?

brazen hatch
#

192 pretty much

#

tho my oscilloscope is pretty inaccurate

#

while running it says 480-520

#

it doesn't wander

tulip sleet
#

I wonder if it was locked to USB and then lost that. Could you hook it up to a 5v wall wart and see if the same thing happens (500 Hz -> 192)

#

prob not that problem, but SAMD chips can lock to USB clock. That's how we get good timing on crystal-less designs

#

another thing I think I see is that sometimes it disconnects from USB and then reconnects, and then disconnects for good

brazen hatch
#

plugged it to thinkpad, expensive gan charger, powerbank, all the same

tulip sleet
#

thanks!

brazen hatch
#

all 192hz

#

Imma go write an essay on gh now

tulip sleet
#

that would cause a lotta problems for many things.

brazen hatch
#

wait lemme try lipo too

#

exactly the same

tulip sleet
#

i am looking through esp-idf issues

tulip sleet
# brazen hatch exactly the same

I commented out all the calibration code, so the only stuff left is

adc_oneshot_new_unit()
adc_oneshot_config_channel()
adc_oneshot_read()  // twice
adc_oneshot_del_unit()

still fails

brazen hatch
#

In which case a hack would be to read core freq and set it back in CP

#

after every adc read

tulip sleet
#

i don't think it would do that, that would affect all kinds of other stuff. It hink it is a bug or a hw problem

brazen hatch
#

I mean, a clock change for just a few cycles..

tulip sleet
#

i still don't think it would do that. the peripheral has its own clocking from the master clock

#

it wouldn't change the master clock

#

but i am planning to read their code

brazen hatch
#

I will too, after I finish my gh essay.

#

I don't know xtensa asm, so if it comes to that it will take a while

manic glacierBOT
#
bill88t
[adafruit/circuitpython] New comment on issue #9291: ESP32-S3 ADC use causes crashes when WiFi in use

ESP32-S3 TFT Feather example:

from time import sleep
import board, analogio, digitalio
from neopixel_write import neopixel_write
import pwmio

state_a = bytearray([0, 8, 0])
state_b = bytearray([0, 0, 8])

adc = analogio.AnalogIn(board.D10)
pwm = pwmio.PWMOut(board.A4)
pwm.duty_cycle = 32767 # 50%
nx = digitalio.DigitalInOut(board.NEOPIXEL)
nxp = digitalio.DigitalInOut(board.NEOPIXEL_POWER)
nx.switch_to_output()
nxp.switch_to_output()
nxp.value = 1

print("Ripping!")
...
manic glacierBOT
manic glacierBOT
manic glacierBOT
spare jacinth
#

did i somehow mess my library up, or is this to blame on circup?```
$ circup install toml
Downloading latest bundles for adafruit/CircuitPython_Community_Bundle (20240608).
py:
There was a problem downloading that platform bundle. Skipping and using existing download if available.
Searching for dependencies for: ['toml']
Ready to install: ['toml']

'toml' is already installed.

$ circup update
Downloading latest bundles for adafruit/CircuitPython_Community_Bundle (20240608).
py:
There was a problem downloading that platform bundle. Skipping and using existing download if available.
Found 1 module[s] needing update.
Please indicate which module[s] you wish to update:

'toml' is a Major Version update and may contain breaking changes. Do you want to update? [y/N]: y
Updated toml
Checking 1 updated module[s] for missing requirements.
Downloading latest bundles for adafruit/CircuitPython_Community_Bundle (20240608).
py:
There was a problem downloading that platform bundle. Skipping and using existing download if available.
Ready to install: []

#

it installed 0.1.7 and latest update (yesterday) is 0.1.9

lone axle
# spare jacinth did i somehow mess my library up, or is this to blame on `circup`?``` $ circup i...

it appears that the automated build of the Community Bundle failed last night: https://github.com/adafruit/CircuitPython_Community_Bundle/actions those actions are responsible for building and uploading the bundle zip to the release page which is where circuip is trying to download it from I believe.

Since the actions failed, there is no 20240608 bundle so it fails to download. If you have an older copy cached locally already it looks like it will use that, but if you didn't have that I guess it would just fail entirely.

#

we should try to add some fallback logic if/when it finds that a bundle download link is dead, maybe it could work backwards through the most recent releases to find one that works? or perhaps that's too complex for handling something that ought to get fixed in the actions for the bundle.

spare jacinth
#

could it be that i created a tag several days ago, then realized it had a bug and removed it (circup didnt get to publish it afaik) and i broke something?

#

i made another release a while ago (0.1.10) which has no code changes, in hopes to cause another packaging of the bundle with the daily update 😅

lone axle
#

possibly, I'm not sure how / if it will handle a tag that got released. I'm not entirely sure though, I'll dig into the actions log here in a bit.

spare jacinth
#

hah, think i got it

# 
$ mpy-cross _toml.py
Traceback (most recent call last):
  File "_toml.py", line 95
SyntaxError: raw f-strings are not supported
#

this should do :)```diff
--- a/toml/_toml.py
+++ b/toml/_toml.py
@@ -92,7 +92,7 @@ class Tokens:
return char, 1 + width

     # TODO: should this raise instead?

  •    warnings.warn(rf"Unknown/invalid escape sequence '\{escaped}'")

  •    warnings.warn(f"Unknown/invalid escape sequence '\\{escaped}'")
   return "\\" + escaped, 1
#

if anyone can manually trigger an update, instead of waiting for the daily(?) one, we should be back to business 😎

lone axle
#

I will try, but I don't know if they manual re-trigger will work. I've never done it for a full bundle release.

tulip sleet
#

@spare jacinth did you re-release toml with a fix?

spare jacinth
#

yeah, i just did

tulip sleet
#

@lone axle I think the tag will need to be removed before any re-run, becuase otherwise it will pick up the same commit.

#

.... could just delete the broken tag and release and wait for tonight if a re-run doesn't work

lone axle
spare jacinth
#

and sorry for the inconveniences! 😅

spare jacinth
lone axle
#

I only re-ran the build action, not the release one. But it seems to have failed the build anyhow.

#

I'm not sure it really took the new one. I'm leaning toward just waiting for the automated one to kick in.

tulip sleet
#

I think you'll have to delete the tag and the release and then re-run the scheuled action, which I found very tricky to do in the past

lone axle
#

I'm not sure if deleting the tag is the same as deleting the release, I thought it was, but am not sure. If we delete the release then I don't think the actions can succeed because they assume the release exists to upload files to I believe.

tulip sleet
#

it is "trigger" and not a "re-run" or something like that.

#

I'd say just delete the release first, then the tag, and let's wait for tonight

lone axle
#

Ahh, it was definitely a "re-run" that I tried. anyhow though I'm scared to mess with it too much since we know that the automation will try again.

tulip sleet
#

(and then try circup)

#

do you have privs to delete the release and tag?

#

I'll just do it...

#

done

lone axle
#

I do, but I'm not confident enough to do it.

#

I'm unclear on the difference between tag and release. I thought they were the same, but looking closer now it appears technically they are different.

tulip sleet
#

it makes a tag first (just a labelled) commit, then it builds a release at that tagged commit

lone axle
spare jacinth
#

tag is a git thing, like a commit or a branch, that you can git checkkout into
and a release is a zip with the code from that version

... i think

lone axle
#

Ahh

#

tag is a labelled commit. That is the piece I was missing.

tulip sleet
#

I cleaned out the local bundle storage and installed something with circup, and it downloaded the bundles OK. So I think we are all set

#

local bundle store is in ~/.local/share/circup. I thought it was in .cache/circup, but that just seems to be logs

spare jacinth
#

it updated to 0.1.8 now, no idea why it said it was a major release

#

hopefully 0.1.10 will be available tomorrow 🙂

tulip sleet
#

personally I would just as soon remove the drivers/helpers distinction. It takes a lot of guessing sometimes

lone axle
#

Yeah, they both seem rather broad. We have so many libraries in each one it's really practical to just browse them anyway. I'm not sure the extra organization that the different folders brings is really that big of a help anywhere.

I do believe there are at least a few things that are hard-coded to look for those directories though, so if we make that change we'll have some updating of various other bits that needs to be done at the same time

tulip sleet
#

i'll be afk now

manic glacierBOT
#
bill88t
[adafruit/circuitpython] New comment on issue #9291: ESP32-S3 ADC use causes crashes when WiFi in use

I went through all the files containing the string rtc_spinlock and extended the range of the critical code.
I also added it in a lot of places.

For the moment I assume all the spinlocks work as intended, across cores.
All of the test were performed with clean builds.

Nothing fixed it.
Not even switching to portENTER_CRITICAL_SAFE.
No wifi code explicitly spinlocks rtc, but instead leaves components/esp_hw_support/adc_share_hw_ctrl.c to do the adc stuff.

I don't think I can...

brazen hatch
#

I went to test if there would be any performance difference with the unicore flag.
My os? Doesn't care.
Division spam? Also doesn't care.

from time import monotonic
from os import urandom

st = monotonic()
samples = 0
counts = []

try:
    while len(counts) < 60:
        if monotonic()-st > 1:
            print("Samples taken:", samples)
            counts.append(samples)
            samples = 0
            st = monotonic()
        else:
            for i in range(10):
                (1045566123 + i) / 993999999955588
            samples += 1
except KeyboardInterrupt:
    pass
except:
    pass

print("Total:", sum(counts)/60)

I let this rip on each build.
Current master total score: 14095.3
Current master with unicore total score: 14333.3
Which is a 1% difference. Margin of error stuff.

I tested all system functions, nothing seems affected.

manic glacierBOT
#
bill88t
[adafruit/circuitpython] New comment on issue #9291: ESP32-S3 ADC use causes crashes when WiFi in use

CONFIG_FREERTOS_UNICORE has literally 0 downsides as far as I can tell.

Measured a 1.1% perf difference on a division benchmark, with the unicore build being the faster one.
There was no other differences across the builds.
(Other than that indeed, the crashes went away.)

So I think, at least temporarily, we should go with that, and open an upstream ESP-IDF issue to have it resolved properly.
(I can go do it if you want me to.)

brazen hatch
#

Not opening PR, this was Dan's find.

manic glacierBOT
manic glacierBOT
manic glacierBOT
#
dhalbert
[adafruit/circuitpython] New comment on pull request #9315: [DO NOT MERGE, EXPERIMENTAL] ADC `samples` keyword and persistent pin init.

I might have done this for #8690. If you want to complete it to check AnalogIn off in that issue, that would be great. Look for how finalizers are done in, say, #9165. Note that finalizer is spelled finaliser in function names due to the British/Australian spelling difference. The basic thing to do is to say it has a finalizer when you construct it, and to make __del__() call deinit(). You also need to remove any global init/reset.

tulip sleet
brazen hatch
#

Now it's been running for an hour, trying to drain it's battery.

manic glacierBOT
sleek relic
# manic glacier

Double negative? Double word? Try:
However, not all ports are fully functional. Some have limited functionality and known serious bugs. …

manic glacierBOT
#
aroidzap
[adafruit/circuitpython] Issue opened: #9319 How to set USB HID name to a composite HID device

CircuitPython version

Adafruit CircuitPython 9.0.5 on 2024-05-22; Waveshare RP2040-Zero with rp2040

Code/REPL

gamepad_a = usb_hid.Device(
    report_descriptor=GAMEPAD_REPORT_DESCRIPTOR_A,
    usage_page=0x01,           # Generic Desktop Control
    usage=0x05,                # Gamepad
    report_ids=(4,),           # Descriptor uses report ID 4.
    # in_report_lengths=(6,),    # This gamepad sends 6 bytes in its report.
    in_report_lengths=(3,),  ...
manic glacierBOT
manic glacierBOT
#
bill88t
[adafruit/circuitpython] New comment on pull request #9315: ADC `samples` keyword, only initialize once on `espressif`, add finalisers to `AnalogIn` objects

Updated the title an opening comment.

Also, @dhalbert turns out it's still possible to trigger a watchdog safemode with just CONFIG_FREERTOS_UNICORE.
You just have to be unlucky enough to .value during a dhcp leash renewal.
Well, with this PR this also goes away, since the object gets initialized only once.

Do you maybe want me to also include CONFIG_FREERTOS_UNICORE in this PR too?
Or are you going to PR it seperately?

manic glacierBOT
#
bill88t
[adafruit/circuitpython] Issue opened: #9320 `microcontroller.cpu.temperature` watchdog safemode on ESP32-S3 when wifi connected

CircuitPython version

Current master
CONFIG_FREERTOS_UNICORE=y
adc_paranoia (the name of the adc multi-sampling branch)
and pretty much every other build I tested

Code/REPL

# Assuming wifi connected via settings.toml
import microcontroller
while True:
    a = microcontroller.cpu.temperature
# `a =` is there to not slow down with printing to serial
# It can take a lot longer than the ADC .value bug to proc.

Behavior

Auto-reload is off.
Ru...

spare jacinth
#

still broken ?

$ circup update
Found device at E:\, running CircuitPython 9.0.0-alpha.6-4-g8e2235e34a-dirty.
A newer version of CircuitPython (9.0.5) is available.
Get it here: https://circuitpython.org/board/weact_esp32_s3_n16r8
Downloading latest bundles for adafruit/CircuitPython_Community_Bundle (20240609).
py:
There was a problem downloading that platform bundle. Skipping and using existing download if available.
Found 1 module[s] needing update.
Please indicate which module[s] you wish to update:

'toml' is a Major Version update and may contain breaking changes. Do you want to update? [y/N]: y
Updated toml
Checking 1 updated module[s] for missing requirements.
Downloading latest bundles for adafruit/CircuitPython_Community_Bundle (20240609).
py:
There was a problem downloading that platform bundle. Skipping and using existing download if available.
Ready to install: []
#

seems to have failed on the very same failed as yesterday (toml/_toml.py) but i can "compile" it locally with mpy-cross 🤔

#

and the bundle does point to my latest commit, i have no idea what's going on here

manic glacierBOT
manic glacierBOT
#
todbot
[adafruit/circuitpython] New comment on issue #9319: How to set USB HID name to a composite HID device

I believe that in order to do this, you would need to create multiple UBS HID interfaces, as the interface name is part of the USB Config Descriptor, and not part of the HID Descriptor. I believe you could do this in C with TinyUSB as a composite device, but CircuitPython creates only one HID interface that all HID devices go into (keyboard, mouse, any custom ones you create like gamepads, etc)

lone axle
#

maybe it's due to the bundle build

spare jacinth
#

...and i have no clue why my lib is breaking it 😅

tulip sleet
spare jacinth
#

i dont have such file

#

but i guess i could create one

manic glacierBOT
tulip sleet
#

maybe you have setup.py instead of pyproject.toml

spare jacinth
#

neither

#

i dont have any "configuration" file

#

which is maybe a wrong setup (?)

tulip sleet
#

For immediate fixes, try running mpy-cross on your library to see what's breaking.

#

but you really need to set it up as described above

spare jacinth
#

i did yesterday and removed the rf"" which was unsupported

#

and then, it supposedly builds my 3 files

tulip sleet
#

there is something else it is upset about

spare jacinth
# 
$ ls toml
__init__.py  _dotty.py  _toml.py
$ mpy-cross toml/__init__.py && mpy-cross toml/_dotty.py && mpy-cross toml/_toml.py 
$
#

yet the github action shows an error on _toml.py

tulip sleet
#

It is fishy it says Automated update by Adabot (adafruit/adabot@20240604) but it happened 12 hours ago. The 20240604 tag is days ago

spare jacinth
#

shall i nuke my repo off [github/bundle] and then add it again?

tulip sleet
#

yes, let's try that. Remove it from the community bundle with a PR. I'll approve that, so that tonights build is OK. In the meantime you can then get it into standardized form.

spare jacinth
#

the link you sent above doesnt mention the pyproject.toml file

#

not sure if it is missing on the docs, or it is not required

#

and it also says that setup is (mainly?) for pypi deployment, which i dont really want to do, but i guess i may still need to have an empty/minimal one

tulip sleet
#

it is going to be created when you create the cookie-cutter template.

#

if you add it to pypi then people can use this library both in CircuitPython and CPython

#

@lone axle I deleted the community release and the tag, as i did yesterday

spare jacinth
#

so i dont feel like maintaining a PyPi thing that really shouldnt get used because there's a better one 🤣

lone axle
#

I don't believe it's required to release on pypi to have it in the bundle.

tulip sleet
#

you can suppress it from pypi if you want. But use the rest of the structure. I don't know why it's breaking but if you use the standard release actions it will build .mpy files during a release, and do a bunch of other standard things so we'll know if that's failing

spare jacinth
#

im on it

tulip sleet
#

thanks!

spare jacinth
#

fixed some tool yelling at licenses, then some pylint warns, but it still fails to build online (yet works fine locally, maybe i have another version of the cross-compiler)

#

hum, maybe i do```
$ mpy-cross --version
MicroPython v1.23.0 on 2024-06-01; mpy-cross emitting mpy v6.3

#

a fresh build of it doesnt fail either... ```
$ ./build/mpy-cross --version
CircuitPython 9.1.0-beta.1-17-g9ab8831d38 on 2024-06-09; mpy-cross emitting mpy v6.1
$ ./build/mpy-cross ~/circuitpython_toml/toml/_toml.py
$

#

im calling it a day for now, honestly

onyx hinge
#

@spare jacinth the f"{whatever!r}" syntax is not supported in CircuitPython 8.x.

spare jacinth
#

so i needed to run an older mpy-cross?

onyx hinge
# 
f"{0!r}"
jepler@bert:~/src/circuitpython_toml$ ./build_deps/mpy-cross-9.x banger.py 
jepler@bert:~/src/circuitpython_toml$ ./build_deps/mpy-cross-8.x banger.py 
Traceback (most recent call last):
  File "banger.py", line 1
SyntaxError: invalid syntax
#

(build_deps/mpy-cross-* are downloaded when circuitpython-build-tools runs, you may have your own preferred ways of getting mpy-cross)

spare jacinth
#

apparently fixed now, today's bundle update will tell

#

do i need to make tag/releases each time i make changes, or will it simply look for new commits?

#

i've been doing them, but would love not to, if they aren't needed 🤣

lone axle
spare jacinth
#

released just in case

onyx hinge
#

bundles use github releases (or at any rate git tags), not branch tips

#

I'm actually not sure whether it fetches git tags or requests github releases; every release has a tag but the reverse is not necessarily true. creating a release via the github UI is probably the "best" way to mark a new revision of your code as being bundle-ready

spare jacinth
#

that's what i always do
new release -(select tag dropdown)-> create tag on release

spare jacinth
#

Seems like i didnt break the bundle this time, thx for the help yesterday @onyx hinge 😁

#

Will confirm using circup when i get back from work, but i can at least see a green tick on the action instead of an error 😛

orchid basinBOT
spare jacinth
#

yay, it did work, however it seems like naming the release and tag as "cookiecutter" was not a great idea, will get back to semver next time i edit anything 🙃

tulip sleet
spare jacinth
#

done

tulip sleet
spare jacinth
#

done aswell

onyx hinge
#

@spare jacinth glad to hear it and happy I could help a tiny bit.

orchid basinBOT
onyx hinge
#

<@&356864093652516868> our weekly meeting is coming up in about 70 minutes, so take the time to add your notes to the document [checked pinned messages for link] & join us in the voice chat at 2PM ET / 10AM PT! I look forward to hearing what everyone is working on.

manic glacierBOT
manic glacierBOT
onyx hinge
#

I'll be joining you in voice in just 3 minutes! thanks for your patience

manic glacierBOT
lone axle
#

https://www.youtube.com/watch?v=82Rb3HbkD_E

YouTube
Kevin McAleer
Is MicroPython better than Basic? Let's reprogram 80s basic retro g...

Let's re-program some 80s Retro games into MicroPython and run them on a Microcontroller.

Sponsored by PCBWay: https://www.pcbway.com

PCBWay, your ultimate destination for PCB manufacturing and assembly. Whether you're a hobbyist, a startup, or a seasoned professional, PCBWay has got you covered.

💁‍♂️ For more information on Kevs Robots, tuto...

▶ Play video
manic glacierBOT
lone axle
#

Thanks for hosting Jeff. Have a great week everyone!

zenith drift
#

hello all, it's good to be here, I am very new to all things ESP / Coding / etc and am looking for a little advice on how best to get started, I would like a rock-solid foundation to build up as I would like to get to the stage where I can design and build my projects, as I see so many cool projects on the internet / YT etc. I have bought a little adafruit circuit playground express, but if thats too much too soon, i am happy to put that to oneside and start with something else. thanks in advance for any tips or ideas.

stuck elbow
#

I think playground express is a very good starting board, you don't have to use all of its features right away.

#

Start with blinking a led

manic glacierBOT
onyx hinge
#

Here is the notes document for next Monday’s CircuitPython Weekly Meeting. It is at the normal time of 11am Pacific / 2pm US Eastern here on Discord. Add your hug reports and status updates to the document before the meeting. If you are unable to attend but would still like to contribute, feel free to add your notes and we’ll read them off during the meeting. Hope to see you there! <@&356864093652516868>
https://docs.google.com/document/d/1QF9nIA7XxxIK3zp6KQ7hVN8nJo3L5HY0ngkQVftxuyw/edit?usp=sharing

Google Docs
CircuitPython Weekly Meeting for June 17, 2024

CircuitPython Weekly Meeting for June 17, 2024 Welcome to the CircuitPython Weekly meeting notes! Feel free to add your Hug Reports and Status Updates before the meeting, alphabetically by your username. During the meeting, we go through them in order. If you can’t make the meeting and would s...

onyx hinge
orchid basinBOT
short tendon
#

Is there any reason radio.ap_info rasies a NotImplementedError on a Pico W?

tulip sleet
#

I guess @onyx hinge didn't have the info available?

slender iron
#

@willow totem you still up?

short tendon
#

and on that topic, @tulip sleet do you see any issue moving the properties like ssid on adafruit_esp32spi to a ap_info class to be the same as native?

onyx hinge
#

I don't specificially remember why I didn't implement it; I'd be happy to see it implemented. an issue can be filed since I'm not likely to get to it right away @short tendon

tulip sleet
#

@short tendon might be worth looking at the PR to see if there was any mention of that

short tendon
tulip sleet
#

yes, the one that implemented wifi

short tendon
#

And I'm totally happy to update examples for that other change

tulip sleet
#

i think you know that already 🙂

short tendon
#

Sure do!

manic glacierBOT
short tendon
onyx hinge
manic glacierBOT
orchid basinBOT
short tendon
manic glacierBOT
#
BoLiu97
[adafruit/circuitpython] Issue opened: #9321 Issue with PDMin and usb_cdc sending data to PC

CircuitPython version

Adafruit CircuitPython 9.0.5 on 2024-05-22; Adafruit Feather M4 Express with samd51j19
Board ID:feather_m4_express

Code/REPL

import time
import array
import math
import board
import audiobusio
import usb_cdc
import struct

# Main program
mic = audiobusio.PDMIn(board.TX, board.D12, sample_rate=24000, bit_depth=16)
samples = array.array('H', [0] * 160)

print("Program started. Collecting and transmitting data...")
while Tru...
manic glacierBOT
#
dtcooper
[adafruit/circuitpython] New comment on pull request #9322: Update `tud_hid_set_report_cb()` report handling

I confirm it works on the Raspberry Pi Pico!

dave@tomato:~$ hidapitester --usagePage 0xff00 --usage 1 --open -l 64 --send-output 2,3,4,5 --timeout 1000 --read-input 1
Opening device, vid/pid:0x0000/0x0000, usagePage/usage: FF00/1
Device opened
Writing output report of 64-bytes...wrote 64 bytes:
 02 03 04 05 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00
 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00...
manic glacierBOT
manic glacierBOT
#
BoLiu97
[adafruit/circuitpython] New comment on issue #9321: Issue with PDMin and usb_cdc sending data to PC
  1. By using the code above on circuitpython, I didn't receive anything from my PC side.
    For example

WARNING:root:Expected 320 bytes, received 0 bytes.

2,3. By using
mic.record(samples, len(samples)) data = struct.pack('<' + 'H' * len(samples), *samples) print('something') usb_cdc.data.write(data)
or
mic.record(samples, len(samples)) data = struct.pack('<' + 'H' * len(samples), *samples) usb_cdc.data.write(data) time.sleep(10)
I was a...

#
BoLiu97
[adafruit/circuitpython] New comment on issue #9321: Issue with PDMin and usb_cdc sending data to PC

However, I am unable to send data through the code above. By adding a timesleep, I am able to send but having a different sample rate.

  • Could you elaborate? What exactly fails with the code above?
  • Where did you add the time.sleep() and for how long
  • What do you mean by "having a different sample"?
  1. By using the code above on circuitpython, I didn't receive anything from my PC side.
    For example

WARNING:root:Expected 320 bytes, received 0 bytes.

2,3. By us...

misty garnet
#

I someone actively working on the SPI portion of circuitpython_register? Are there any gotchas why it's not a straightforward implementation relative the I2C version?

onyx hinge
# misty garnet I someone actively working on the SPI portion of circuitpython_register? Are the...

I don't know that anyone's actively working on https://github.com/adafruit/Adafruit_CircuitPython_Register_SPI/issues/1. I think the short reason why is: adafruit makes a lot more sensors/devices with i2c than spi, so it hasn't been an important goal for them. Secondarily, I'm not sure there's as much standardization around a "register model" in SPI. The most common SPI devices Adafruit sells are TFTs and those are covered by displayio. Second most common might be SD cards and those are covered by sdcardio (or sdioio, where available).

misty garnet
#

thanks for that, it looks like it should be pretty straightforward though I'm not a big fan of the several levels of abstraction that seem to be in place busio->bus_device->devicedriver->i2c_register, not to mention it seems odd to have a bcd_alarm and datetime builtin

#

my use case is the TMC4361a SPI stepper controller

manic glacierBOT
#
dhalbert
[adafruit/circuitpython] New comment on issue #9321: Issue with PDMin and usb_cdc sending data to PC

Instead of trying to read 320 bytes at once with a 10-second timeout, try reading much smaller chunks, or read with timeout=0.

I am thinking that the reader on the PC side might get out of sync and miss some of the initial bytes. That would make it be out of sync from then on, waiting and waiting, when the microcontroller side has already sent everything. To test this, I would change the microcontroller side to send some known test data, and not do PDMIn for now. Instead, send some packe...

manic glacierBOT
#
romkey
[adafruit/circuitpython] New comment on issue #8918: Input/Output: MacOS Sonoma 14.4b3 CIRCUITPY Slow Write Performance

Just tested the first developer beta of macOS 15 Sequoia with an Unexpected Maker Bling board and it's still slow. From the command line, roughly one second per file (tried with three different sets of files) for modestly sized files (like Python source code). Larger files took > 1s/file

The good news is it does work, so they didn't totally break it this time (yet)! Ejecting is also still very slow.

With 24 files:
<img width="545" alt="Screenshot 2024-06-11 at 11 12 46 AM" src="https:...

#
dhalbert
[adafruit/circuitpython] New comment on issue #8918: Input/Output: MacOS Sonoma 14.4b3 CIRCUITPY Slow Write Performance

@romkey Thank you for testing! Do you alreadyhave a Feedaback issue open with Apple about this? I would like to come up with a simple portable test that shows what @eightycc 's OP in this issue shows: slow on 1GB or smaller, much faster above. I tried a multi-volume USB stick but that failed in various ways. It seems difficult to create a disk image of a small size natively on macOS. I just want to make it as easy as possible for them to reproduce the issue.

manic glacierBOT
manic glacierBOT
#
bendavis78
[adafruit/circuitpython] New comment on pull request #8830: shared-module/usb_hid: allow HID to wake sleeping host computer

I see, thanks! I tested on a Mac and a keypress does in fact wake it up. Side-note: I'm using a Pico W for this project.

My current problem is with waking a Linux host. I've tried the solution explained in the SO post:

$ lsusb | grep Adafruit
Bus 003 Device 018: ID 239a:8120 Adafruit Pico W
$ cat /etc/udev/rules.d/10-wakeup.rules
ACTION=="add", SUBSYSTEM=="usb", ATTRS{idVendor}=="239a", ATTRS{idProduct}=="8120" RUN+="/bin/sh -c 'echo enabled > /sys/bus/usb/devices/usb3/power/wa...
manic glacierBOT
manic glacierBOT
manic glacierBOT
#
dhalbert
[adafruit/circuitpython] New comment on issue #9291: ESP32-S3 ADC use causes crashes when WiFi in use

Here are some other apparently related issues. These are about running the ADC tight loops, it appears, that cause watchdog errors. The original test in this issue does 20ms waits between reads, but still has problems.
https://github.com/espressif/esp-idf/issues/12466 (as mentioned by @jepler)
https://github.com/espressif/esp-idf/issues/8753
https://github.com/espressif/arduino-esp32/issues/6549 (same author as 8753)

manic glacierBOT
manic glacierBOT
#
bendavis78
[adafruit/circuitpython] New comment on pull request #8830: shared-module/usb_hid: allow HID to wake sleeping host computer

Update: It looks like it was just a matter of getting the devpath correct in the udev rule. I figured out a more rubust way to set the power/wakeup attribute (since the devpath can change between boots):

ACTION=="add", \
SUBSYSTEM=="usb", \
ATTRS{idVendor}=="239a", ATTRS{idProduct}=="8120", \
ATTR{power/wakeup}="enabled", \
RUN+="/bin/sh -c 'echo set attributes for %p >> /var/log/wakeup-device.log'"

The idVendor and idProduct are what I found in lsusb. Not sure if ...

manic glacierBOT
lone axle
#

Is it possible to increase the range that the BLE workflow will work for?

manic glacierBOT
#
dhalbert
[adafruit/circuitpython] New review comment on pull request #9325: Improve ESP BLE and turn on BLE workflow

The _api_password + 1 was pretty mysterious until I looked at the other place the password was fetched.

    // Fetch the web api password the same way it's fetched in web_workflow.c.
    char _api_password[64];
    const size_t api_password_len = sizeof(_api_password) - 1;
    os_getenv_err_t result = common_hal_os_getenv_str("CIRCUITPY_WEB_API_PASSWORD", _api_password + 1, api_password_len);
    wifi_workflow_active = result == GETENV_OK;
manic glacierBOT
manic glacierBOT
#
MatthewJeffson
[adafruit/circuitpython] Pull request opened: #9326 Add: add seeed_xiao_esp32c6

Add a new board Seeed Studio XIAO ESP32C6 from Seeed Studio.
Purchase: https://www.seeedstudio.com/Seeed-Studio-XIAO-ESP32C6-p-5884.html
Pinout: https://wiki.seeedstudio.com/xiao_esp32c6_getting_started/#hardware-overview
Schematic: https://files.seeedstudio.com/wiki/SeeedStudio-XIAO-ESP32C6/XIAO-ESP32-C6_v1.0_SCH_PDF_24028.pdf

Already done testing LED, digital input, Wi-Fi, analog reading, and IIC, but the UART seems a little trouble. More info and .bin file [here](https://github.com/Z...

manic glacierBOT
#
gitcnd
[adafruit/circuitpython] New comment on pull request #9238: Add support for sunton_esp32_2432S024C board

@lboue I'm planning to build a display controller using this in the next week or so, at which point I'll check all those other features are working properly, pick all the correct naming, maybe write some examples to make it easy to get started, then do the submission.

If you need it now - what I've done so far is good enough to work from. I can post the firmware build if you just want to try it out without the hassle of pulling this and building yourself.

jade root
#

Hello.
Does anyone have a file transfer client library such as ftp/tftp/scp that works with blinka?

lone axle
jade root
lone axle
# jade root Yes, that's right.

In that context you should have access to existing "Normal" CPython libraries and built-in modules. It seems that there is ftplib builtin for interacting with FTP Servers, I've never used it so I can't say much more specific about it, but you should be able to use that from your Python code running on the RPi even if / when that code is using the Adafruit_Blinka library to also interact with other hardware (or whatever it's doing)

#

This would be something you can only use in that CPython context on a Raspberry pi or other linux computer though. CircuitPython microcontrollers will not have access to this built-in module nor are they likely to work with existing libraries made for normal python.

jade root
manic glacierBOT
manic glacierBOT
manic glacierBOT
manic glacierBOT
slender iron
lone axle
# slender iron As in distance? There is a slower phy that should go further but I haven't tried...

Yeah it mentions "This public broadcast is done at a lower transmit level so the devices must be closer. " in the docs here https://docs.circuitpython.org/en/latest/docs/workflows.html#ble. I've noticed that I don't see my CPB in the list of results from discover if it's on my desk but if it put it down next to my PC cast near the antenna's (which I assume one of is bluetooth, but not certain) then it does show up in the list of devices discovered.

But in my case that is out of reach, so I have to get up and reach over the desk to get it to press the reset button or do anything else.

#

I guess part of my problem though might be that I haven't got the pairing / bonding process working. Maybe after that it switches back to higher power that is usable from further away?

slender iron
#

right, once it is bonded it broadcasts stronger

#

it is intentional that it has to be close

#

though we could hack it for your testing if you want

lone axle
#

When you press reset button during the blue blinks in order to enable BLE workflow does it ever "timeout" and disable itself? I was finding that sometimes it was disappearing and would come back if I reset twice again. I wasn't sure if that was intended behavior or perhaps just BLE flakiness on my computers part (I was seeing some other weird behaviors with the bluetooth and in the settings window)

manic glacierBOT
#
bill88t
[adafruit/circuitpython] New comment on pull request #9315: ADC `samples` keyword, only initialize once on `espressif`, add finalisers to `AnalogIn` objects

While I'm at it, should I maybe add a property for configuring sample size after the object has been created?

As is, the sample size can only be set during object creation like a so:
adc = analogio.AnalogIn(board.ADC_PIN, samples=5000)

I kinda think like adding a adc.samples property would be helpful both for visibility and ease of use.

lone axle
#

Also one other question, are the mobile Glider app and this Circuitpython library: https://github.com/adafruit/Adafruit_CircuitPython_BLE_File_Transfer/ the only existing implementations that interact with BLE workflow? I don't have too much experience with BLE generally so I'm trying to find any other existing implementations to look at and learn from.

slender iron
#

I don't think it should timeout

#

There is a js implementation too for the browser

lone axle
#

okay, thank you. Nice I'll look into that one.

slender iron
#

kk

manic glacierBOT
manic glacierBOT
manic glacierBOT
onyx hinge
#

we want mpy-cross to be most broadly compatible so should we bump to 12? or is 12 as old in dog years as say ubuntu 20.04 ?

#

this will also rename the binary so that the name contains macos-12 instead of macos-11 which is likely to cause problems for other sw. 😕 I don't know enough about the mac development environment to know how forwards & backwards compatible binaries are.

#

(or if the -12 should come OUT OF the binary name as long as we have the pain of change)

manic glacierBOT
#
jepler
[adafruit/circuitpython] Pull request opened: #9329 Update macos version in github actions

Move from 11 (deprecated) to 12 (the oldest non-deprecated)

Because the artifacts formerly had the macos version number in the
mpy-cross binary filename, we have to change the binary name anyway.
Rename it without the macos version (e.g., mpy-cross-macos-x64 instead of
mpy-cros-macos-11-x64)

https://github.blog/changelog/2024-05-20-actions-upcoming-changes-to-github-hosted-macos-runners/

Closes: #9328

tulip sleet
onyx hinge
#

many are just circuitpython forks so that's not terrible

#

I filed several bugs with various non-fork repos about this.

#

no idea what's going on with GH actions, some of the PR jobs just fell over with no logs

image.png
short tendon
manic glacierBOT
#
bill88t
[adafruit/circuitpython] New comment on pull request #9325: Improve ESP BLE and turn on BLE workflow

I did a little more testing.

Please, please, please, please, do not enable this for boards without psram.
ESP32-S3 doesn't have the memory for this..

Current main branch: 195k usable
This PR: 125k usable

ble takes 70k. 35% of all memory...
It will break a lot of applications.

If ble has to stay, can we please at least have a CIRCUITPY_NOBLE flag or something like this so that all the memory doesn't get allocated?
Or at least can we check as to if something is paired befor...

short tendon
brazen hatch
#

I have never gone from liking a pr to being scared of it so quickly.

#

We absolutely do need a way to turn ble off.

#

I'm actually hopping on S2 rn

#

it's probably going to be not-fine TM, since it has even less ram

#

oh wait it doesn't have ble

#

I don't have a C3 with 8MB flash, so I can't test there either

#

it has 120k usable

#

so -70k = 50k usable?

#

calling 50k usable is a bit of a stretch..

#

If we did have a way to not have this brick in ram unless needed, it would be fine on S3 probably.
C3, no, I don't think so.

#

Wait does og esp32 have ble too? Huh, well the goes into the same camp as S3.

stuck elbow
#

samd21 has less

brazen hatch
#

💀

stuck elbow
#

but yeah, it seems like there will be a need for images with optional functionality, also for the pico w

tulip sleet
#

@brazen hatch Please add comments to the PR or open an issue. Thanks.

brazen hatch
#

did

tulip sleet
#

For your AnalogIn PR, I will not be merging that in for 9.1.0, in the interest of not generating new churn.

#

so don't rush

brazen hatch
brazen hatch
manic glacierBOT
manic glacierBOT
#
jepler
[adafruit/circuitpython] Issue opened: #9330 ship mpy-cross similar to micropython-uncrustify

Recently micropython dealt with the difficulty of getting the right uncrustify version by creating a micropython-uncrustify package. This looks like it'll install on windows, linux & mac.

we could do mpy-cross the "same way", probably having an mpy-cross- package so that different mpy-cross binaries are co-installable.

then, use this in circuitpython-build-tools, get rid of all the hard coding of download paths in various tools, etc.

onyx hinge
#

just a random brainstorm ..

manic glacierBOT
manic glacierBOT
onyx hinge
#

opening a discussion might be appropriate, but I'm not going to do it in haste

manic glacierBOT
glossy pecan
#

Finding nearest RP2040 board cousin for porting my board to CP, based on ext. Flash type

manic glacierBOT
manic glacierBOT
#
bablokb
[adafruit/circuitpython] New comment on issue #8754: Espressif ADC is not accurate, has a deadzone and differs per cpu

I also stumbled across this issue. IMHO, there is a misunderstanding and the CP-documentation is wrong about what VREF is. It is not the maximum measurable voltage as stated by the docs. For many MCU this is the case, but not for all.

ESP32 is one case, the VREF is in the range 1000mV-1200mV, the measurable voltage range is documented in the Arduino docs, see table in https://docs.espressif.com/projects/arduino-esp32/en/latest/api/adc.html.

I also think there should be consistency acr...

manic glacierBOT
tulip sleet
#

@slender iron do you want to keep debugging in the open BLE PR 9325, or should we merge this and then work on the remaining BLE bugs? If we merge now, then we can merge the sleep PR and people can test that. We'll leave 9.1.0 issues open.

slender iron
#

and you can help with the bugs

tulip sleet
#

that's exactly what I thought. I'll do that now.

slender iron
#

at my desk now through lunch then I'm out

#

I'll rebase the second PR asap

manic glacierBOT
slender iron
manic glacierBOT
slender iron
#

moving code from flash to ram as-needed would be super helpful but also really tricky

#

code is usually assumed to be at a fixed address

#

its the same thing that'd help rp2040 usb_host and picodvi

#

@willow totem thanks for testing my BLE PR. Sorry it doesn't fix your issue

manic glacierBOT
onyx hinge
#

@tulip sleet any idea what's up with this failure? It's affecting my 9.0.x PR, which doesn't touch silabs. https://github.com/adafruit/circuitpython/pull/9329 https://github.com/adafruit/circuitpython/actions/runs/9489365651/job/26153337808?pr=9329

Traceback (most recent call last):
  File "/home/runner/work/circuitpython/circuitpython/ports/silabs/tools/slc_cli_linux/slc", line 21, in <module>
    from lib.slc_run import _entryPoint
  File "/home/runner/work/circuitpython/circuitpython/ports/silabs/tools/slc_cli_linux/lib/slc_run.py", line 24, in <module>
    from slc_common import (
  File "/home/runner/work/circuitpython/circuitpython/ports/silabs/tools/slc_cli_linux/lib/slc_common.py", line 9, in <module>
    from distutils.version import StrictVersion
ModuleNotFoundError: No module named 'distutils'
make[1]: *** [Makefile:170: slc-generate] Error 1```
slender iron
#

disutils was removed in python 3.12 right?

tulip sleet
#

yes, that's right

onyx hinge
#

oh did "python-version: 3.x" change recently to be 3.12?

tulip sleet
#

yes, but not yesterday or anything

#

it was a while ago

#

i'll look at the beta.3 build

onyx hinge
# 
Python 3.12.4 (main, Jun  7 2024, 19:06:40) [GCC 12.2.0] on linux
Type "help", "copyright", "credits" or "license" for more information.
>>> from distutils.version import StrictVersion
```huh that import works in the python 3.12 docker image, oddly enough.
tulip sleet
#

beta.3 used 3.12.3. They might have updated the build p[latform and distuilts is no longer in the system image. You could try adding it requirements-dev.txt

onyx hinge
#

@tulip sleet do you want me to try that in my PR? since it'll demonstrate the problem, I think

#

2024-05-22T14:56:34.000Z 637.2 kB adafruit-circuitpython-explorerkit_xg24_brd2703a-en_US-9.0.5.bin

#

last build of silabs explorerkit for 9.0.x was weeks ago

tulip sleet
# 
Runner Image
  Image: ubuntu-22.04
  Version: 20240516.1.0
  Included Software: https://github.com/actions/runner-images/blob/ubuntu22/20240516.1/images/ubuntu/Ubuntu2204-Readme.md
  Image Release: https://github.com/actions/runner-images/releases/tag/ubuntu22%2F20240516.1

in 9.1.0-beta.3 build

# 
Runner Image
  Image: ubuntu-22.04
  Version: 20240609.1.0
  Included Software: https://github.com/actions/runner-images/blob/ubuntu22/20240609.1/images/ubuntu/Ubuntu2204-Readme.md
  Image Release: https://github.com/actions/runner-images/releases/tag/ubuntu22%2F20240609.1
#

in the failed build ^

#

I don't see distuils explicitly, but various tools were updated and maybe distutils is no longer installed as a dependency somewhere

tulip sleet
onyx hinge
#

I'll give that a try

manic glacierBOT
manic glacierBOT
manic glacierBOT
tulip sleet
#

PR9325, not 3925

manic glacierBOT
tulip sleet
#

@slender iron the Unicode for the BLE heart rate monitor I think may be related to the other characteristic errors. It is not the HRM itself: it's trying get the manufacturer string from the DeviceInfoSerivce.

#

... it's getting the manufacturer characteristic, but the value is a long bunch of garbage.

willow totem
#

hey scott, do you mean with the circuitpython client code that you want to recreate with something instead of bluetooth web dashboard? I think it just subscribes (excuse my terminology) to the notify Accelerometer characteristic, and immediately then changes the notification period to 500ms. I half read recently there was a much lower recommended threshold, like 8-2xx ms, but can't remember if that was nrf related or where it was. No idea what the code would look like for circuitpython client BLE stuff, but might have a go anyway as the examples were fairly informative. I'm not able to get on device right now, probably tomorrow night or over the weekend.

orchid basinBOT
alpine rapids
#

Hi! I've been very slowly working on getting this ESP32 added, got a question about the onboard neopixel: The color order is GRB for whatever reason, and I added what I thought was the appropriate way to mitigate that here: (direct link to the line https://github.com/adafruit/circuitpython/pull/9274/files#diff-5afc6bbdd15add58d62e8e910fa425b7a0685267148ab24980173d54acf15da4R14) but when I try setting the color in a sample in code.py, the color order is still off. Does that line that I linked to only affect the status light that CircuitPython uses to indicate statuses (i.e., blinking green when the code finishes without error, or yellow for safe mode) (answered my own question, yes).

Is there any way to make the color order behave as expected for code in code.py, or should I just document what the correct color ordering is in the description of the board?

tulip sleet
manic glacierBOT
orchid basinBOT
tulip sleet
#

@slender iron are you really on, or are you busy packing, etc.?

#

i have a diagnosis of the Characteristic read problem (maybe you already figured this out)

#

Reading the characteristic works in 9.0.5 and earlier. common_hal_bleio_characteristic_get_value() hasn't changed. It has this code.

        CHECK_NIMBLE_ERROR(ble_gattc_read(conn_handle, self->handle, _read_cb, &read_info));
        int error_code;
        xTaskNotifyWait(0, 0, (uint32_t *)&error_code, 200);
        CHECK_BLE_ERROR(error_code);
        return read_info.len;

It does the ble_gattc_read(), waits 200 ticks, and then returns. in main, 200 is not enough time, 2000 works. We changed the ms per tick from 10 to 1 in https://github.com/adafruit/circuitpython/pull/9134. So any ticks-based wait needs to be *10. And it shouldn't be ticks anyway, it should be ms. I will start with this and keep debugging the other problems.

orchid basinBOT
manic glacierBOT
lone axle
#

Is there any trick to, or information about the Service Discovery process during BLE workflow connection? I have tried a few different ways to use bleak python library to make a connection to a CPB that has been reset twice to enable BLE workflow but thus far my script is always crashing when I try to connect. It seems that upon connection bleak is attempting to run service discovery, but the device is immediately disconnecting instead of allowing that to happen bleak.exc.BleakError: failed to discover services, device disconnected

manic glacierBOT
tulip sleet
lone axle
tulip sleet
lone axle
#

I tried 8.2.10 as well but same results. I can get successful connection to the same device with code circuit python org and glider so I think it's okay on the device side just bleak doesn't seem to play nice

#

Going to bed for now

tulip sleet
#

goodnight! The bleak version is important. I haven't tested with later than the version pinned by blinka bleio

#

it's quite out of date

#

and might not work on newer OS versions

manic glacierBOT
#
ashirvadz
[adafruit/circuitpython] Issue opened: #9333 Setting the name property on an adafruit_ble.BLERadio instance doesn't work on ESP32-S3

CircuitPython version

Adafruit CircuitPython 9.1.0-beta.3 on 2024-05-22; FeatherS3 with ESP32S3

Code/REPL

import time
from adafruit_ble import BLERadio
from adafruit_ble.advertising.standard import ProvideServicesAdvertisement
from adafruit_ble.services.nordic import UARTService
from adafruit_ble.uuid import StandardUUID
from adafruit_ble.services import Service
from adafruit_ble.characteristics import Characteristic
from adafruit_ble.characteristics...
manic glacierBOT
manic glacierBOT
lone axle
tulip sleet
#

Bleak has to keep up with a lot of churn in host OS's, especially on macOS, but Windows has been a problem too.

lone axle
#

I am on Linux (Ubuntu)

tulip sleet
lone axle
#

I will go back to something in the 0.20 range when I get back to it.

Have you ever used bleak to connect with the BLE workflow? I wondering if there may be some incompatibility between them (or at least with the way I am trying to connect)

tulip sleet
lone axle
#

You mentioned some things about timeouts in that PR as well. When it does fail for me it's almost instantaneous upon trying to connect I think. I'll try to validate that with additional logging and see if there is anywhere I can try increasing it.

tulip sleet
#

I am testing the Espressif BLE stuff now and having to do hard resets sometimes after an error. The persistence needed for the workflow may be causing it to stay in a bad state sometimes, but I'm not sure about that.

#

I am not testing workflow yet at all. There are other things to fix.

lone axle
#

I do think I've seen issues with the persistence of it as well. I have not pinned it down to a specific pattern just yet, but it does keep "disappearing" from the scan list and I can get it to re-appear by another 2x reset of the MCU, and a few times I've had to toggle off / on the BLE on my computer as well because it refused to even find it in the scan anymore, but did after that power cycle of BLE.

manic glacierBOT
jaunty juniper
#

@onyx hinge thanks for the heads up on the mpy-cross thing

#

By the way, I have been away from CP things for a few months, for various reasons, but I plan to come back and get up to date on things soon hopefully

onyx hinge
manic glacierBOT
#
RetiredWizard
[adafruit/circuitpython] New comment on issue #6934: Add support for the Wemos D1 R32 Development board.

which binary firmware do you suggest to flash on this board?

it is still not listed in the downloads page.

EDIT: the board seems to have wifi issues with the Esp32 firmwares i've tried (not from circuitpython) . Not sure if it is fixable at firmware-level.

According to @mirkokral above:

UPDATE: Looks like if i use the Adafruit Feather HUZZAH32 board it works like a charm. Should i close or not?

manic glacierBOT
#
mikeysklar
[adafruit/circuitpython] Issue opened: #9334 storage.remount disable_concurrent_write_protection

CircuitPython version

CircuitPython 9.0.5
Feather RP2040

Code/REPL

import storage

# Remount the filesystem as writable with concurrent write protection disabled
storage.remount("/", readonly=False, disable_concurrent_write_protection=True)

Behavior

Still seeing drive mounted as read-only when laptop is rebooted, but Feather is powered from USB hub.

Description

In this case the desired behavior is to override the write_protection so de...

#
jepler
[adafruit/circuitpython] New comment on issue #2202: Add Native SSL Library

It's now possible to use the core ssl module with sockets implemented in Python (#8954); this actually works with wiznet. However, nobody's checked whether this actually works with airlift; it might require changes in the airlift socket implementation to correctly align with the standard socket. If it doesn't, the esp32spi github repo is probably the right place to open issues.

Additionally, few or no builds for board with an airlift include the core ssl module. If there are combos for whi...

orchid basinBOT
manic glacierBOT
orchid basinBOT
manic glacierBOT
#
mikeysklar
[adafruit/circuitpython] New comment on issue #9334: storage.remount disable_concurrent_write_protection

same same w CircuitPython 9.1.0-beta.3

initial state (rw on CIRCUITPY)

sklarm@grazie CIRCUITPY % cat boot_out.txt
Adafruit CircuitPython 9.1.0-beta.3 on 2024-05-22; Adafruit Feather RP2040 with rp2040
Board ID:adafruit_feather_rp2040
UID:4545373038057E2C
boot.py output:

sklarm@grazie CIRCUITPY % mount | grep CIRCUITPY
/dev/disk4s1 on /Volumes/CIRCUITPY (msdos, local, nodev, nosuid, noowners)

disconnect laptop (or reboot or power off):

sklarm@grazie CIRCUITPY %...
#
mikeysklar
[adafruit/circuitpython] New comment on issue #9334: storage.remount disable_concurrent_write_protection

boot_out.txt

Adafruit CircuitPython 9.1.0-beta.3 on 2024-05-22; Adafruit Feather RP2040 with rp2040
Board ID:adafruit_feather_rp2040
UID:4545373038057E2C
boot.py output:

boot.py

import storage

# Remount the filesystem as writable with concurrent write protection disabled
storage.remount("/", readonly=False, disable_concurrent_write_protection=True)

code.py

# SPDX-FileCopyrightText: 2021 Kattni Rembor for Adafruit Industries
# SPDX-License-Identifier...
orchid basinBOT
manic glacierBOT
#
bill88t
[adafruit/circuitpython] Issue opened: #9335 ESP32-C2 constant crash loop when web-workflow is enabled

CircuitPython version

Adafruit CircuitPython 9.1.0-beta.3-29-ged5591c5cb on 2024-06-14; NodeMcu-ESP32-C2 with ESP32C2
# Aka latest as of the time of writing

Code/REPL

# Wifi creds & web-workflow passwd set in settings.toml
# No code.py

Behavior

Crashes into safemode, then crashes again since wifi-autoconnect and web-workflow run even in safemode.
The board cannot be stopped, since neither wifi-autoconnect nor web-workflow check for keyboard in...

manic glacierBOT
manic glacierBOT
manic glacierBOT
lone axle
#

Would it be possible / super difficult to have a way to specify the brightness of the startup RGB led blinking? Specifically the Red, Yellow Blinks, Blue blinks / solid on the nrf devices is what I am interested in making dimmer.

tulip sleet
lone axle
#

I don't know if having it be dynamically setable like that is feasible, or worth the space it might take though.

tulip sleet
#

but you are talking about pre boot.py.

#

how's the brightness in the UF2 bootloader? Maybe we should match that

lone axle
#

it's difficult to compare for certain since the bootloader is more steady instead of blinks from on to off. But to me the yellow and blue blinks during startup are entering uncomfortable territory.

shadow vigil
tulip sleet
tulip sleet
manic glacierBOT
#
gitcnd
[adafruit/circuitpython] Issue opened: #9336 is /cp/serial/ working?

CircuitPython version

Adafruit CircuitPython 9.1.0-beta.1-24-g24f2e4f994 on 2024-05-18; S2Mini with ESP32S2-S2FN4R2
and
Adafruit CircuitPython 9.1.0-beta.3-28-g03e42a8c0c-dirty on 2024-06-15; AButton V3 with ESP32
and
Adafruit CircuitPython 9.0.5 on 2024-05-22; S2Mini with ESP32S2-S2FN4R2

Code/REPL

Connecting to my cpy boards delivers the "welcome!" screen.  They include a "serial terminal" [sic] link which is a websocket (not serial) connection to the ...
lone sandalBOT
manic glacierBOT
#
jepler
[adafruit/circuitpython] New comment on pull request #9329: Update macos version in github actions

I don't think there's anything we need to take care to do at the same moment. Until changed, for example, circuitpython_build_tools will always use mpy-cross from 9.0.0-alpha.2 (!) for creating its 9.x mpy files.

If we can drop 8.x from circuitpython_build_tools at the same time as we add 9.1 or 10, we can change it then to simply use the new path. If we need to keep 8.x support we'll need to add some version dependent checking for the mpy cross location. but I think this can all be handle...

manic glacierBOT
#
dwetchells
[adafruit/circuitpython] New comment on issue #9211: Walkmp3rson learn guide does not work with CP 9.x Freezes and displays USB not recognizable on PC Windows

Hey Dan, Still testing. I have JP project working on a nRF52840 with the sound files from JP project. Still no luck with using an RP2040. Waiting on some other feathers to try when they come in. Thats what I get when I use the US Postal system. but it gets here. UPS has lost to many of my orders.
Thank again. I'll try changing the bit rate down. I would not have thought it would stop the directory display of the SD card but what do I know.
Have a super week..

Dale
______________...

manic glacierBOT
#
Timeline8
[adafruit/circuitpython] New comment on issue #9291: ESP32-S3 ADC use causes crashes when WiFi in use

Appreciate everyone's help, but I am still experience problems. Back to the Wavershare S3 Zero board, I downloaded adafruit-circuitpython-waveshare_esp32_s3_zero-en_US-20240613-main-PR9325-03e42a8.uf2 and installed it. Did it a couple times due to still having problems. The boot_out.txt reads:

Adafruit CircuitPython 9.1.0-beta.3-28-g03e42a8c0c on 2024-06-13; Waveshare ESP32-S3-Zero with ESP32S3
Board ID:waveshare_esp32_s3_zero
UID:437BAD9541C4

There is a more recent version by o...

manic glacierBOT
#
Timeline8
[adafruit/circuitpython] New comment on issue #9291: ESP32-S3 ADC use causes crashes when WiFi in use

D7 & D8 so IO7 & IO8 which go to GPIO7 & GPIO8 of the S3 (fortunately Waveshare kept numbers the same on the board vs the ESP32) which according to the datasheet for the ESP32-S3 is ADC1

(cut & pasted)

ADC1_CH6 GPIO7
ADC1_CH7 GPIO8

My full code that I running at this moment where I am seeing this is...

`import gc
import time
import board
import neopixel
from random import randint
import busio
import displayio
from fourwire import FourWire
from adafruit_st7789 import ST7...

#
dhalbert
[adafruit/circuitpython] Issue opened: #9338 Espressif BLE: Heart Rate not working

Testing on main (after https://github.com/adafruit/circuitpython/pull/9325):

Run ble_heart_rate_simpletest on a Metro ESP32-S3. Does not work on 9.0.5 either.

Fixes and diagnoses already done:

  1. xTaskNotifyWait(0, 0, (uint32_t *)&error_code, 200); used several times. 200 is ticks. We changed ticks from 1 per 10msecs to 1 per msec in #9134. I changed to use pdMS_TO_TICKS() to make it tick-length independent.
  2. xTaskNotifyWait() above may return error_code BLE_HS_EAGAIN, whi...
manic glacierBOT
manic glacierBOT
manic glacierBOT
manic glacierBOT
manic glacierBOT
#
wraetz-dev
[adafruit/circuitpython] Issue opened: #9339 Waveshare ESP32-S3-SIM7670G Support for CircuitPython

Waveshare has a ESP32-S3R2 based board that has a SIM7670G-4G Cellular modem. It seems to share many elements with Waveshare's ESP32-S3-Pico board. They both have 2MB PSRAM and 16MB of flash memory. Wiki page for the ESP32-S3-SIM7670G is here.
The pin mapping appears different from ESP32-S3-Pico but I can't identify the relevant pins from the [pinout diagram](https://www.waveshare.com/img/devkit/ESP32-S3-SIM7670G-4G/ESP32-S3-SIM7670G-4G-...

manic glacierBOT
short tendon
#

Question on where to put some request based methods:

For HTTP_AIO, a method is needed that will send base64 encoded data to a AIO webhook endpoint.

To do this, requests will need to be able to handle a file object (easy).

But we also need a wrapper that will do the encoding, let's call it Base64FileEncoderStream. My question is where does that class live? I can imagine people wanting to use it outside of HTTP_AIO, but it feels strange to put it in requests...

manic glacierBOT
lone sandalBOT
manic glacierBOT
#
vospascal
[adafruit/circuitpython] New comment on pull request #8989: usb_hid.set_interface_name

thanks a lot Dan i was already almost done making custom builds in github ci..
how can i not have seen that ...

my solution

from hid_gamepad.simple.boot import gamepad_descriptor
import usb_hid
import supervisor

# Define custom VID, PID, product name, and manufacturer
CUSTOM_VID = 0xDDFD  # Non-registered VID
CUSTOM_PID = 0x5050  # Non-registered PID
PRODUCT_NAME = "SUNSHINE"
MANUFACTURER_NAME = "Good mood"

# Set USB identification
supervisor.set_usb_identification(vid=...
manic glacierBOT
#
Sola85
[adafruit/circuitpython] Issue opened: #9341 ESP32S3: displayio causes "CircuitPython core code crash"

CircuitPython version

Main branch (see commit ids below) on espressif_esp32s3_devkitm_1_n8

Code/REPL

import busio, board
import displayio
displayio.release_displays()
# The issue is reproducable whether there is something connected to these pins or not. The exact pins used here also don't matter.
spi = busio.SPI(board.IO5, board.IO4, None)
display_bus = displayio.FourWire(spi, command=board.IO3, chip_select=board.IO2, reset=board.IO1)

print("\nDone ...
manic glacierBOT
lone axle
manic glacierBOT
short tendon
onyx hinge
manic glacierBOT
short tendon
onyx hinge
#

they don't seem super related to me, other than being related to streams

stuck elbow
#

I think the general preference is to have separate small libraries, because then it's easier to fit only the needed ones on resource constrained boards

short tendon
#

Makes sense. Although, you can always remove the mpy files you don't need.

Once I get everything working will make PRs and such and see what people think

manic glacierBOT
tulip sleet
manic glacierBOT
manic glacierBOT
manic glacierBOT
clear halo
willow totem
#

What are the first things to disable to save flash compile size? I'm trying to get a debug build for the C6 4mb flash with BLE, but even disabling DISPLAYIO and FRAMEBUFFERIO wasn't enough. Are there any other large obvious ones?

tulip sleet
willow totem
#

thanks Dan!

empty salmon
#

I had not thought about ULAB. While I don't currently have a space issue I am curious to determine the size different. Thanks.

willow totem
#

for future fun followers, the build flags I've gone for: (I'm aware one framebuffer is wrong)
CIRCUITPY_DISPLAY_LIMIT=0 CIRCUITPY_DISPLAYIO=0 CIRCUITPY_FRAMEBUFFERIO=0 CIRCUITPY_FRAMEBUFFER=0 CIRCUITPY_ULAB=0 CIRCUITPY_SYNTHIO=0

manic glacierBOT
manic glacierBOT
#
b-blake
[adafruit/circuitpython] New comment on issue #9334: storage.remount disable_concurrent_write_protection

@dhalbert;

Is the remount in boot.py?
Could you check boot_out.txt and see that no errors were written there?
There is no remount in boot.py
There are no errors in boot_out.txt

This is the output of CircUp.exe this morning from my Windows 10 PC after having it off all night. It was the first thing run at boot-up. If you want specifics about any device. please ask.

` Volume in drive C is Windows SSD
Volume Serial Number is B984-5BB6

Directory of C:\Users\Bruce\AppData\Lo...

manic glacierBOT
#
timchinowsky
[adafruit/circuitpython] Pull request opened: #9343 Fix delays and rounding in samd PWM

Blocking delays are observed when setting duty_cycle in the samd port of CircuitPython (see https://github.com/adafruit/circuitpython/issues/7653), behavior which interferes with some applications of PWM, and which differs from from that of other ports. Investigation of this issue revealed another problem: In some cases, the API will round a non-zero duty cycle down to zero, causing the PWM signal to stop cycling at all. The changes proposed in this PR address both of these issues.

De...

onyx hinge
#

@slender iron I may be a bit late to the meeting please read my hug reports if I'm not there in time (but I most likely will be)

slender iron
#

kk

#

<@&356864093652516868> We'll have our weekly meeting in about 5 minutes from now in this text channel and in the circuitpython voice channel. Please take the time to add your notes in advance to the document: https://docs.google.com/document/d/1QF9nIA7XxxIK3zp6KQ7hVN8nJo3L5HY0ngkQVftxuyw/edit -- I look forward to everyone's updates!

Google Docs
CircuitPython Weekly Meeting for June 17, 2024

CircuitPython Weekly Meeting for June 17, 2024 Welcome to the CircuitPython Weekly meeting notes! Feel free to add your Hug Reports and Status Updates before the meeting, alphabetically by your username. During the meeting, we go through them in order. If you can’t make the meeting and would s...

lone axle
gilded cradle
#

Thanks

wraith crow
#

Thanks All 😁

onyx hinge
lone axle
#

Thanks for hosting Scott, have a great week everyone!

tulip sleet
#

thanks @slender iron !

slender iron
#

thats for the web workflow

#

not ble

onyx hinge
#

oh, hm

slender iron
#

I don't have a specific change in mind. I just know the library got most of its use as the protocol was being developed

onyx hinge
#

yeah nevermind

lone axle
#

Thank you anyhow. I'll dig and compare the web and mobile app implementations with the library. I was able to use both web and glider successfully so they've got whatever is needed to get it working.

#

Once I get it worked out I'm hoping to add more specific info to the documentation. Today it's mostly just a pointer to the file transfer library. The web workflow implementation has lots more specific documentation available.

onyx hinge
#

now I'm sweating it that I can't find the bluetooth listDir implementation in that repo. where is it?

lone axle
onyx hinge
#

ah thanks, it was going to bother me

slender iron
lone axle
slender iron
#

It is separate because my intention was for it to be more general than circuitpython

#

You could change the CP link to go to that bullet point

slender iron
#

Here is the notes document for next Monday’s CircuitPython Weekly Meeting. It is at the normal time of 11am Pacific / 2pm US Eastern here on Discord. Add your hug reports and status updates to the document before the meeting. If you are unable to attend but would still like to contribute, feel free to add your notes and we’ll read them off during the meeting. Hope to see you there! <@&356864093652516868> https://docs.google.com/document/d/144fIOxqiH-xJLrzGLr7FSjrF0y5hylkiNM5iMPc7vpU/edit?usp=sharing

Google Docs
CircuitPython Weekly Meeting for June 24th, 2024

CircuitPython Weekly Meeting for June 24th, 2024 Here is the notes document for next Monday’s CircuitPython Weekly Meeting. It is at the normal time of 11am Pacific / 2pm US Eastern here on Discord. Add your hug reports and status updates to the document before the meeting. If you are unable t...

slender iron
lone axle
#

I got a bit further along, watching BLE traffic in wireshark I saw that code.circuitpython.org BLE workflow implementation sends the listdir command "header" in one write, and then in a subsequent write packet sends the path that it wants to list.

The ble_file_transfer library seems to write only a single packet with the path appended to the end of the header section.

After changing to use two subsequent write operations I can see in wireshark now that the device is sending back all of the information about files in the directory, however the code inside blinka_bleio / ble_file_transfer library is still seemingly stuck in an infinite loop trying to read. It doesn't seem to be actually "seeing" the data that I know (from wireshark) is being sent.

#

the basic version characteristic read succeeds and gets proper data, but the more complex transfer characteristic read is just looping forever never seeing any data. My best guess is it's something with bleak, blinka_bleio causing trouble. I've tried a few versions of bleak but no luck so far.

I did actually get the MCU to MCU one working with two nrf52840s though so I am no longer having the same infinite loop issue on the microcontroller side. It's successfully completing the listdir.

I'm going to try removing blinka_bleio and ble_file_transfer library from the mix for now and just try to get some code that uses bleak to read and write characteristics directly. I manually prepared the list dir command and am able to see that I got back some real data, I'll just need to parse it more manually without the helper library for now.

slender iron
#

@lone axle are you doing a read instead of subscribing to notifications?

lone axle
lone axle
#

I changed the manual code to subscribe for notifications and use a callback to receive them. Thanks for the pointer.

#

Hmm, I seem to be getting duplicated callbacks, every two callbacks to the notification subscription contain the same data.

sleek relic
lone axle
#

I thought maybe it could be related to using two different write commands, so I changed that back to a single write and I still get the duplicate callbacks. But it is actually sending data in the callbacks so now that is working successfully in with the single write.

I think I've got the basic bits needed to make it work with bleak directly so I'm going to head that direction for now.

It's possible there is something in the adafruit_ble, blinka_bleio, or adafruit_ble_file_transfer layers that is leading some different behavior when using adafruit_ble_file_transfer on a PC vs. MCU. I don't have a clear idea where to look within those libraries to start tracking down the difference.

slender iron
#

I was using the library on mcu only

#

not with bleak

lethal grove
#

Hey devs!

Unfortunately, I've had to migrate away from CircuitPython for my current project (interpreter too heavy), but I was hoping someone could help me ID where in the source code certain file system behavior lives?

I'd really like to implement a refresh-on-modification feature. Specifically, file system modification causes (soft) reset. And I'd like to do it close to bare metal. I think CircuitPython handles it well and I'd love to use it as a reference. Sadly, my lack of experience is making that functionality hard to find in the source code. Anyone have a pointer they'd be willing to share?

Thanks a million,

  • Finn
onyx hinge
# lethal grove Hey devs! Unfortunately, I've had to migrate away from CircuitPython for my cu...

the function autoreload_trigger (in supervisor/shared/reload.c) may let you start to get a handle on how the functionality works. basically, it's called from shared/usb/usb_msc_flash.c when the drive content is changed; it has a timer that is intended to wait for the drive activity to quiesce, then it initiates the reload reload_initiate(RUN_REASON_AUTO_RELOAD). Elsewhere in the code this occurs as a Python exception that is ultimately handled by tearing down and restarting the interpreter (top level main.c)

lethal grove
spare jacinth
lethal grove
spare jacinth
#

im not sure, but definitely stick to a single one

#

probably fine to ask on either one

lethal grove
manic glacierBOT
slender iron
#

since it is about circuitpython internals

slender iron
#

@tulip sleet S3 is waking up immediately. I'm looking into it

#

may be correct because I'm not requesting a pull

#

false alarm. works once I set pull

manic glacierBOT
lone axle
#

I managed to read and parse the full listdir response 🎉. The data buffering ended up different from adafruit_ble_file_transfer, and there is probably room for improvement. But I was able to take some of the struct un-packing and logic in the reading / parsing loop from adafruit_ble_file_transfer and put into this script that is using bleak directly.

manic glacierBOT
#
tannewt
[adafruit/circuitpython] New comment on pull request #9342: esp: make cpu frequency settable

(Personally I would prefer option 1, as the code would be more maintainable than with option 2)

Option 1 sounds good to me.

Question 2

I think it makes sense to inform the user about the possible frequencies that can be set, as there are not many valid ones (e.g. 20, 40, 80, 160, 240 MHz for the original ESP32). For now I created a (rather long) new error message for each esp variant containing the valid frequencies, which is consistent with the settable frequency for `po...

manic glacierBOT
manic glacierBOT
manic glacierBOT
#
bill88t
[adafruit/circuitpython] New comment on pull request #9344: Only start BLE for workflow when desired

I have good news and bad news.

Good news first, from a quick test, this seems to bring back the ram usage back to normal.

It won't match 9.0.x though because of the BLE related code in IRAM.

I entered an optimization frenzy and now I'm at a net-negative ram usage.

Bad news, brings back #9291.. somewhat.

It no longer produces a crash.
The IO clock or whatever it's called is just slowing down instead of the system clock.
This means, the neopixel goes white (wrong frequency), ...

brazen hatch
#

I have an irk that's it's also affecting esp32 (og)

#

I need to test a bit more

#

I hopped between builds a couple times now

manic glacierBOT
#
gitcnd
[adafruit/circuitpython] New comment on issue #9336: is /cp/serial/ working? (web workflow)

Ahh - you use the IP address, right?

Try again using the DNS name:

http://cpy-eaf90c.local/cp/serial/

I notice it works with the IP address, but not with the name.

I'm using Chrome on Windows; windows does properly resolve mdns so the web page at http://cpy-eaf90c.local/ loads and works fine, but not the /cp/serial/ part.

ymmv with ubuntu; I never figured out how to get that understanding mdns (at least not in WSL anyhow)

#
dhalbert
[adafruit/circuitpython] New comment on issue #9336: is /cp/serial/ working over mDNS names? (web workflow)

It works fine for me with both IP and cpy-nnnn.local names. I'm assuming your browsers are up to date. I've tested with all three operating systems. My LAN is running on a Verizon FIOS router.

Have you set the DNS server for your OS's to be your router (e.g., 192.168.x.x) or have you set DNS to directly go somewhere else, such as 8.8.8.8 (Google)?

Are you running a code.py that is doing something? If so, try removing code.py and see if it makes any difference.

manic glacierBOT
#
gitcnd
[adafruit/circuitpython] Issue opened: #9345 terminal of code editor malfunctions when using mDNS names on esp32* (at least)

CircuitPython version

Adafruit CircuitPython 9.1.0-beta.1-24-g24f2e4f994 on 2024-05-18; S2Mini with ESP32S2-S2FN4R2
and
Adafruit CircuitPython 9.1.0-beta.3-28-g03e42a8c0c-dirty on 2024-06-15; AButton V3 with ESP32
and
Adafruit CircuitPython 9.0.5 on 2024-05-22; S2Mini with ESP32S2-S2FN4R2

Code/REPL

Same as issue https://github.com/adafruit/circuitpython/issues/9336

Behavior

Using the name (e.g. http://cpy-eaf90c.local/code/ ) instead of IP ad...

manic glacierBOT
manic glacierBOT
manic glacierBOT
#
gitcnd
[adafruit/circuitpython] New comment on issue #9336: is /cp/serial/ working over mDNS names? (web workflow)

code.py is default.

http://cpy-eaf90c.local/ loads and works fine.

http://cpy-eaf90c.local/cp/serial/ does not work properly.

Nor does the serial console from http://cpy-eaf90c.local/code/

both works OK as an IP address only.

Regardless of whether or not my network settings are involved, since http://cpy-eaf90c.local/ works fine, there must be a problem that needs fixing, even if there was something I could do to work around it. Fixing the cause is always better than covering...

#
dhalbert
[adafruit/circuitpython] New comment on issue #9336: is /cp/serial/ working over mDNS names? (web workflow)

I cannot replicate the problem locally, so I want to try to figure out what's different about your setup.

Do you have any extensions installed? Could you try Chrome with no extensions? See https://www.howtogeek.com/779557/how-to-start-google-chrome-without-extensions-on-windows/

What are your DNS servers?
What version of Windows, and what version of Chrome?
Could you try Firefox, and Opera?

manic glacierBOT
#
lewinski
[adafruit/circuitpython] Pull request opened: #9346 Fix display power on Adafruit ESP32-S3 Reverse TFT Feather

I got this board over the weekend and was unable to get anything out of the display in CircuitPython (8.2.7, 9.0.5, and 9.1.0-beta.3) but I knew the display worked fine because both the UF2 boot loader and Arduino example code for the board could display things. I ended up figuring out that the Arduino board definition was forcing the I2C power pin high, but the CircuitPython was forcing the NeoPixel power pin high. I confirmed this using the [pin assignment diagram](https://learn.adafruit.co...

manic glacierBOT
manic glacierBOT
#
applecuckoo
[adafruit/circuitpython] Pull request opened: #9347 Revise Arduino Nano 33 BLE definition to include pins for the Nano 33 BLE Sense Rev2

This is a quick addition to define a couple of interrupt pins only available on the second revision of the Arduino Nano 33 BLE Sense. In a way, this is to close my own issue, #7991.

I figured that since there weren't any pin-related breaking changes between the two revisions, it was better to add to the existing definition.
This is only based on pin diagrams since I don't have the actual hardware to test this.

manic glacierBOT
#
gitcnd
[adafruit/circuitpython] Issue opened: #9348 ping does not return a time (always says 0.0 or None)

CircuitPython version

Adafruit CircuitPython 9.1.0-beta.3 on 2024-05-22; Ai Thinker ESP32-CAM with ESP32

Code/REPL

>>> import time
>>> import wifi
>>> import ipaddress
>>> ip1 = ipaddress.ip_address("196.1.56.26") # this is deliberately a very slow server in S.A.
>>> print(wifi.radio.ping(ip1))
None
>>> print(wifi.radio.ping(ip1,timeout=2))
0.0
>>> ip1 = ipaddress.ip_address("1.1.1.1")
>>> print(wifi.radio.ping(ip1))
0.0
>>> print(wifi.radio.ping...
manic glacierBOT
#
timchinowsky
[adafruit/circuitpython] Pull request opened: #9349 Tools for setup-free CircuitPython builds

Setting up tooling for CircuitPython builds, particularly for ESP32, can be confusing. This PR contains tools which enable setup-free containerized building of any CP port with the one-liner sudo tools/docker-build/build .

There is doubtless a lot of room for improvement in these scripts, in particular in the way file ownership is fixed up at the end of build, but they seem to work, and I thought I'd share in case you're interested.

manic glacierBOT
#
lewinski
[adafruit/circuitpython] New comment on pull request #9346: Fix display power on Adafruit ESP32-S3 Reverse TFT Feather

I think the S2 Reverse needs the same change but I don't have one to test on. Let me know if you want the change proposed anyways.

I've been using an S2 reverse in a project for a year or so (running 8.0.5) and the display has always worked fine so maybe not an S2 issue 😁

That's weird, because the board definition for the S2 Reverse definitely has the wrong pin being forced high next to the "make sure the TFT is on" comment. That leads me to wonder if maybe different builds of t...

#
lewinski
[adafruit/circuitpython] New comment on pull request #9346: Fix display power on Adafruit ESP32-S3 Reverse TFT Feather

I'm thinking that maybe CircuitPython should be putting both 7 and 21 high, which is what both TinyUF2 and Arduino are doing.

I'll push up a revised change later.

#
dhalbert
[adafruit/circuitpython] Pull request opened: #9350 espressif: fix returning ping time and pinging too often
  • Fixes #9348
  • Fixes #7940

The ping time gets lost if it's not retrieved during a ping callback, because the storage used is freed quite early. Set up a callback to retrieve the time instead of fetching it directly.

I fixed #7940 while I was in the code, by waiting a couple of seconds and trying again if esp_ping_new_session() fails. Interestingly, I had to wait at least two seconds, not the one second that is the task timeout.

@gitcnd

#
dhalbert
[adafruit/circuitpython] New comment on pull request #9346: Fix display power on Adafruit ESP32-S3 Reverse TFT Feather

I'm thinking that maybe CircuitPython should be putting both 7 and 21 high, which is what both TinyUF2 and Arduino are doing.

The NeoPixel power is handled by CIRCUITPY_STATUS_LED_POWER, defined in mpconfigboard.h and checked in supervisor/shared/status_leds.c. So you don't need to add that to board.c.

However, the pin is alos wrong for ESP32-S2 Reverse TFT. Could you fix that as well as part of this PR? Thanks.

The non-reverse S2 and S3 TFT boards look OK to me.

lone sandalBOT
manic glacierBOT
#
jepler
[adafruit/circuitpython] New review comment on pull request #9350: espressif: fix returning ping time and pinging too often

(sorry if this appears twice; github seems to have eaten my comment)

I'm not super confident about this change.

  • I don't relish the idea of explaining to the user: yes, ping takes 2 seconds in time.monotonic but claims the ping response time is 0.1 seconds.
  • esp_ping_new_session might fail for other reasons (e.g., all sockets are in use) and now ping will wait 10 seconds before returning failure. before, it would fail prompt.y
#
dhalbert
[adafruit/circuitpython] Pull request review submitted: #9349 Tools for setup-free CircuitPython builds

Thanks for doing this. It's a one-stop shop for someone who doesn't want to set up a build enviornment.

I don't know how docker works. Does the Dockerfile run on every invocation, or are you just including it to show how you built timchinowsky/build_circuitpython?

Rather than put this into the circuitpython repo, I am thinking that it could be left as a repo that you control, since you control the docker image anyway. Otherwise we become ultimately responsible for maintaining it. We...

manic glacierBOT
#
timchinowsky
[adafruit/circuitpython] New review comment on pull request #9349: Tools for setup-free CircuitPython builds

This gets the requirements files from the tip of main, but I think you want to get them from the repo?

You mean the repo that CP is being built from? Can't do that because the image is given access to that repo at runtime, but the requirements need to be available at image build time so that the python built into the image can be configured. If you wanted to do it the other way, you'd need to have python pull in the requirements every time it is asked to build CP.

#
timchinowsky
[adafruit/circuitpython] New review comment on pull request #9349: Tools for setup-free CircuitPython builds

Yeah this is one of the things that could probably be improved. The container doesn't have to be run with sudo but the processes inside the container run as root so the owner of anything created by the container ends up as root which makes git unhappy. The reason I'm having it run as sudo is so that I can patch this up after the fact and change the owner back to who it should be.

#
bill88t
[adafruit/circuitpython] Pull request review submitted: #9349 Tools for setup-free CircuitPython builds

I also do not think this belongs in the repo.

Also the current approach, other than completely breaking permissions (cough cough, add --user $UID:$(id -g)), cannot do specific languages, which sometimes we need to test flash size.

I feel like having it separate would be better since then it could work as is with a local clone of the repo, but it could also be told to just clone it (or a fork, specific commit, etc.) on the fly and discard everything other than the output folder.

#
timchinowsky
[adafruit/circuitpython] New comment on pull request #9349: Tools for setup-free CircuitPython builds

I also do not think this belongs in the repo.

Also the current approach, other than completely breaking permissions (cough cough, add --user $UID:$(id -g)), cannot do specific languages, which sometimes we need to test flash size.

I feel like having it separate would be better since then it could work as is with a local clone of the repo, but it could also be told to just clone it (or a fork, specific commit, etc.) on the fly and discard everything other than the output fold...

#
timchinowsky
[adafruit/circuitpython] New comment on pull request #9349: Tools for setup-free CircuitPython builds

Thanks for doing this. It's a one-stop shop for someone who doesn't want to set up a build enviornment.

I don't know how docker works. Does the Dockerfile run on every invocation, or are you just including it to show how you built timchinowsky/build_circuitpython?

Rather than put this into the circuitpython repo, I am thinking that it could be left as a repo that you control, since you control the docker image anyway. Otherwise we become ultimately responsible for maintain...

manic glacierBOT
#
timchinowsky
[adafruit/circuitpython] New comment on pull request #9343: Fix delays and rounding in samd PWM

Another place to put the tests is in tests/circuitpython-manual. I'm not sure we should put the very detailed testing screenshots and discussion in the repo. I'd suggest putting it in this PR and putting a link to it somewhere with the tests.

tests/circuitpython-manual is certainly on-point. I'd want to put a couple images there to illustrate what "good" and "bad" look like. I put the test data in the repo bc it's hard to wrangle a lot of images if you just "attach" them. Probably...

manic glacierBOT
manic glacierBOT
manic glacierBOT
manic glacierBOT
#
dhalbert
[adafruit/circuitpython] New review comment on pull request #9347: Revise Arduino Nano 33 BLE definition to include pins for the Nano 33 BLE Sense Rev2

The build failed because tools/ci_check_duplicate_usb_vid_pid.py noticed that this PID is the same as the original board. I looked at the latest Arduino Board Support Package and that has not included the Rev2 as a separate board. So I'm guessing that the USB PID is still the same, though having one in hand would verify that. For now, you can add this new board to the exceptions list in tools/ci_check_duplicate_usb_vid_pid.py.

It still makes sense for this to be a new board, because it...

manic glacierBOT
manic glacierBOT
manic glacierBOT
manic glacierBOT
#
MaxWright17
[adafruit/circuitpython] Issue opened: #9354 Adafruit Dark saber code help problem

CircuitPython version

9.0.5

Code/REPL

# SPDX-FileCopyrightText: Kattni Rembor for Adafruit Industries
# SPDX-FileCopyrightText: Limor Fried for Adafruit Industries
# SPDX-FileCopyrightText: 2019 Liz Clark for Adafruit Industries
#
# SPDX-License-Identifier: MIT

"""
Prop-Maker based Darksaber
Adapted from the Prop-Maker based Master Sword code
by Kattni Rembor & Limor Fried
Adafruit invests time and resources providing this open source code.
Plea...
manic glacierBOT
manic glacierBOT
manic glacierBOT
manic glacierBOT
#
mMerlin
[adafruit/circuitpython] Issue opened: #9355 collections.deque status not matching documentation

CircuitPython version

Adafruit CircuitPython 9.0.5 on 2024-05-22; Adafruit PyPortal with samd51j20

Code/REPL

from collections import deque
a = deque((), 5)
len(a)
a.append(0)
a.append(1)
a.append(2)
a.append(3)
a.append(4)
len(a)
a[0]
list(a)

Behavior

0
5
Traceback (most recent call last):
  File "", line 1, in 
TypeError: 'deque' object isn't subscriptable
Traceback (most recent call last):
  File "", line 1, in 
TypeEr...
orchid basinBOT
manic glacierBOT
lone sandalBOT
manic glacierBOT
manic glacierBOT
manic glacierBOT
manic glacierBOT
orchid basinBOT
manic glacierBOT
manic glacierBOT
#
aguaviva
[adafruit/circuitpython] Issue opened: #9356 esp32-cam failing to init camera

CircuitPython version

CircuitPython 9.1.0-beta.3 (https://circuitpython.org/board/ai-thinker-esp32-cam/)

Code/REPL

import espcamera
import board

cam = espcamera.Camera(
    data_pins=board.CAMERA_DATA,
    external_clock_pin=board.CAMERA_XCLK,
    pixel_clock_pin=board.CAMERA_PCLK,
    vsync_pin=board.CAMERA_VSYNC,
    href_pin=board.CAMERA_HREF,
    pixel_format=espcamera.PixelFormat.JPEG,
    frame_size=espcamera.FrameSize.SVGA,
    i2c=board....
tiny token
#

Hey all, I've encountered this issue (https://github.com/adafruit/circuitpython/issues/9235) with my QTPY ESP32-C3 and while the work around is effective, I don't like adding it every time I refresh my QTPY. I'm willing to help get a more permanent fix implemented into Circuit Python for this (and possibly other) C3 boards. However, I'm not sure where to start. I've looked at the board.c code and several of the radio.x files found in common_hal and shared_bindings folders, but I'm still not sure on how to implement that same (wifi.radio.tx_power=) call in the c code for the board. Any pointers to get me moving in the right direction?

tulip sleet
#

in ports/espressif/common-hal/wifi/Radio.c.

tulip sleet
#

Then I'd say do this in common_hal_wifi_init(), conditionalized for C3 only.

tiny token
#

so for testings sake, if I set it using esp_wifi_set_max_tx_power(34) in board_init(), i shouldn't need the boot.py code anymore?

brazen hatch
#

Also, I have one specific network that doesn't work on just 3D antennas.

#

I have yet to find time to find the value that would work for it

#

So yea, this needs to be done in board_init().

#

Alternatively, we could have a build flag for every antenna type.

#

To make it a bit more abstract and updatable.

manic glacierBOT
manic glacierBOT
#
dhalbert
[adafruit/circuitpython] New comment on issue #9235: Wemos C3 Mini should pre-set tx_power to 8.5

It appears from reading the documentation, though I am not sure of this, that wifi.radio.tx_power=8.5 may be lowering the max power from its default. Would you be willing to do some testing to fetch and print out the power before and after the setting and see how its changed? It might get printed before USB is ready, so you could ESP_LOGI() it, or else turn off wifi workflow and do import wifi by hand to see the print.

manic glacierBOT
#
bill88t
[adafruit/circuitpython] New comment on issue #9235: Wemos C3 Mini should pre-set tx_power to 8.5

@dhalbert The current default is 20.0, set by the "default wifi settings" from idf as far as I can tell.
So yes with 8.5 it's less than half power.
But that's what works on all the antenna types.
Compatibility vs range.

And besides, the tiny ceramic antenna did connect through closed doors to an AP 3 rooms away. It's honestly fine.

Also, as I already mentioned, for those who need the range, it's a settable value. They can crank it all the way.
However, the default should be what wor...

manic glacierBOT
#
karlfl
[adafruit/circuitpython] New comment on issue #9235: Wemos C3 Mini should pre-set tx_power to 8.5

So, I took some time to perform some tests using the station example from the ESP-IDF. I discovered that my QTPY C3 would effectively connect up through power values of 18dBm (72 in the ESP API) but would never connect using the highest (and default) power of 20dBm (80 in ESP API). Lower values would connect fairly well, but 18 seemed to connect within 1 or 2 tries every time.

When I was digging into the [ESP WIFI API](https://docs.espressif.com/projects/esp-idf/en/stable/esp32/api-refe...

manic glacierBOT
manic glacierBOT
manic glacierBOT
#
bablokb
[adafruit/circuitpython] New comment on issue #9359: Display updates extremly slow after PR#9344

No AnalogIn, no wifi. Just plain displayio with labels and images:

def show_colors():
  clear_display()
  N_SHADES = 4
  N_COLORS = N_SHADES*N_SHADES*N_SHADES
  COLOR_STEP = 256/N_SHADES
  stripe_width = display.width // N_COLORS
  i = -1
  for red in range(N_SHADES):
    for green in range(N_SHADES):
      for blue in range(N_SHADES):
        i += 1
        color = ((min(int(COLOR_STEP*red),255) << 16) +
                 (min(int(COLOR_STEP*green),255) << 8) +
        ...
manic glacierBOT
#
bablokb
[adafruit/circuitpython] New comment on issue #9235: Wemos C3 Mini should pre-set tx_power to 8.5

@karlfl

discovered that my QTPY C3 would effectively connect up through power values of 18dBm (72 in the ESP API) but would never connect using the highest (and default) power of 20dBm (80 in ESP API).

Nice insight. A few hours too late for me: I just replaced shortly before your post the onboard ceramic antenna on my QTPY C3 with a (very!) cheap external antenna and had a good and stable connection at once. Before it didn't even find all available networks when trying a scan.

#
gitcnd
[adafruit/circuitpython] New comment on issue #9336: is /cp/serial/ working over mDNS names? (web workflow)

@RetiredWizard @dhalbert

Yes! (kindof) Looks like it might be an extension issue? Chrome without extensions "half works" (and - weirdly - when working also fixes the other problem whereby any connection attempt from chrome locked up the serial console on the board making it unusable). Chrome does not work in that same --disable-extensions browser in incognito mode however (the IP address did load, but not the mDNS one).

The ESP32 required rebooting if connections failed (even the...

manic glacierBOT
manic glacierBOT
#
elipsion
[adafruit/circuitpython] Issue opened: #9360 USB audio source/sink

I've been tinkering around with the audio infrastructure and, while it's fun to do stuff like playing WAV files of an SD card, am kind of missing the functionality to get audio streams to and from my computer over USB.

Ideally I'd like to be able to expose a number of sources/sinks to the host and perform mixing on them together with the audiobusio module.

slender iron
#

@tulip sleet at my desk for an hour or two. need anything?

manic glacierBOT
tulip sleet
#

there are some PR's that could be reviewed if you are so inclined, but not rush

slender iron
#

I have some tabs open to do so. Not sure how far I'll get

manic glacierBOT
slender iron
#

@tulip sleet ok if I merge the pwm testing PR 👆 ?

tulip sleet
#

author proposed a pdf, but that's still large, and in the repo

slender iron
#

he removed most

tulip sleet
#

it's still 1060kB for two images. I dunno. Maybe I shouldn't worry.

slender iron
#

I worry more about some of our submodules

tulip sleet
#

yeah, but they are other repos. well maybe it's not worth worrying

#

right now a zip of what's in the repo is 14.7MB

slender iron
#

can you compress the images?

tulip sleet
#

I just tried in gimp. it's the same size

slender iron
#

they are helpful in understanding the test because the script generates those images

tulip sleet
#

let me try to compress them. I might be succeeding if it twiddle some settings.

slender iron
#

@tulip sleet I'll let you follow up on the PR

manic glacierBOT
#
bablokb
[adafruit/circuitpython] Issue opened: #9362 Safemode due to heap allocation failure (since PR#9325)

CircuitPython version

since PR#9325

Code/REPL

def send_file_to_host(src_filename, dst_file, filesize, buf_size):
  import sys
  import binascii
  try:
    with open(src_filename, 'rb') as src_file:
      bytes_remaining = filesize
      buf_size = buf_size // 2
      while bytes_remaining > 0:
        read_size = min(bytes_remaining, buf_size)
        buf = src_file.read(read_size)
        sys.stdout.write(binascii.hexlify(buf))
        bytes_rem...
#
elipsion
[adafruit/circuitpython] New comment on issue #9360: USB audio source/sink

TinyUSB already has audio support IIRC so this would involve hooking that into CP's audio subsystem.

I did check that before I submitted the request, to not put in an unreasonable FR.

Is there anything I can do to help with implementation? I realize it might be a bit of a tough nut to crack as a first contribution, but I would gladly help out with testing or documentation.

#
dhalbert
[adafruit/circuitpython] New comment on pull request #9357: Configure default espressif wifi power

@bill88t We started with choice of 8.5 dBm was from @askpatrickw on a particular board: #6540.

@askpatrickw It's been a while, but did you choose that value as the maximum practical for the board, or was it a quick choice that worked?

@karlfl's experimentation (https://github.com/adafruit/circuitpython/issues/9235#issuecomment-2181890602) where 18dBm seems like the max. I would be inclined to use a somewhat lower value, but I'm not sure about choosing 8.5 dBm, which is quite a bit less....

#
bill88t
[adafruit/circuitpython] New comment on issue #9235: Wemos C3 Mini should pre-set tx_power to 8.5

I also had another find.
The other boards with ceramic antennas, like the Waveshare ESP32-S2 Pico (and probably the same S3 board) do not ground the ceramic antenna. There is a ground plain keepout area.
On the supertiny however, the antenna is grounded to the plain, intentionally.

What I don't understand now is why the 3D antennas connect to that one network with the reduced power.
They aren't grounded.

#
dhalbert
[adafruit/circuitpython] New comment on issue #9235: Wemos C3 Mini should pre-set tx_power to 8.5

The SuperTiny schematic, such as it is, is here: https://www.aliexpress.us/item/3256806297878595.html?gatewayAdapt=glo2usa4itemAdapt.

Chip antennas are fed at one end and aren't meants to be grounded at the other end. But they need to be near a ground plane, at a particular orientation and distance: https://www.globalspec.com/learnmore/semiconductors/communication_ic/chip_antenna. Sometimes there isn't room for this or the designers are sloppy.

manic glacierBOT
#
FoamyGuy
[adafruit/circuitpython] Issue opened: #9363 circuitpython_setboard command python version support

the circuitpython_setboard command currently supports python 3.10+ only due to a change between py3.9 and py3.10.

Here is an example of the traceback that results from trying a version below 3.10:

$ circuitpython_setboard adafruit_feather_esp32s3_4mbflash_2mbpsram
setting board: adafruit_feather_esp32s3_4mbflash_2mbpsram
Traceback (most recent call last):
  File ".../.venv/bin/circuitpython_setboard", line 8, in 
    sys.exit(set_board())
  File ".../.venv/lib/python3.9/si...
manic glacierBOT
manic glacierBOT
#
bill88t
[adafruit/circuitpython] New comment on pull request #9357: Configure default espressif wifi power

I started testing (on the supertiny) by constantly switching between two networks, one really close (10cm) and one that is purposefully barely in range (about ~25m diagonally, vision obstructed).

I started from tx_power=18 and climbed down 1.0 at a time.
While some values did work once or twice, once the board heated up (due to the wifi transmission power), the reliability plummeted to 0.
The board was let to cool before every value swap.

I decided 11 is fully reliable at normal ope...

manic glacierBOT
manic glacierBOT
#
Timeline8
[adafruit/circuitpython] New comment on issue #9291: ESP32-S3 ADC use causes crashes when WiFi in use

CONFIG_FREERTOS_UNICORE=y, so that only one core is used: problem seems to go away. That would explain why ESP32-S2,which has only one core, doesn't have the problem.

Circling back to this comment, which was before you thought it was fixed only for us to find it may not be, is this statement still true? And if so, is this something that can be done through CircuitPython?

I ask because as I try to decide which single configuration to use going forward on my project using S3 modu...

manic glacierBOT
manic glacierBOT
manic glacierBOT
manic glacierBOT
#
aguaviva
[adafruit/circuitpython] New comment on issue #9356: esp32-cam failing to init camera

This works like a charm:

import socketpool
import wifi
from adafruit_httpserver import Server, Request, Response

import board
import espcamera

cam = espcamera.Camera(
    data_pins=board.CAMERA_DATA,
    external_clock_pin=board.CAMERA_XCLK,
    pixel_clock_pin=board.CAMERA_PCLK,
    vsync_pin=board.CAMERA_VSYNC,
    href_pin=board.CAMERA_HREF,
    powerdown_pin=board.CAMERA_PWDN,
    reset_pin=None,
    i2c=board.I2C(),
    external_clock_frequency=20_000_000,
    ...
#
Sola85
[adafruit/circuitpython] New comment on pull request #9342: esp: make cpu frequency settable

I like the idea of automatically "rounding down" to the next lower supported frequency. As far as I can tell, overclocking is not supported by esp_pm_configure anyway, so that would not be an issue.

Regarding the size of the build: Instead of disabling the feature on C3 builds, would it be acceptable to modify the partition sizes to make the changes fit?
I think it's only partitions-4MB-no-uf2.csv that would require a small reduction in user_fs size, in order to increase the size of...

manic glacierBOT
manic glacierBOT
#
dhalbert
[adafruit/circuitpython] Issue opened: #9365 usb_cdc.serial.read() doc says it reads one byte, but it will read as much as possible

Right now the documentation for usb_cdc.serial.read() gives the signature as usb_cdc.serial.read(size = 1). That is, it reads a at most a single byte. This does not match the implementation: it actually tries to read as much as possible, because it uses the standard stream operation read(), which has those semantics.

Pyserial's read() has the same signature and does read only one byte if no argument is given. I think our original intention was to match pyserial.

So rather than c...

manic glacierBOT
manic glacierBOT
#
DLBPointon
[adafruit/circuitpython] New comment on issue #8425: In KMK python firmware, Specifying "KC.MO(1)" in code.py crashes CircuitPython

@Bob837217
Apologies for commenting here, but it is the only place I have seen a similar issue to what I am having.
Hi, did you manage to solve this at all?

I've tested this on all major and latest beta releases of CircuitPython since 8.2.8.

Running the layer assignment before appending Layer(), ends up with the keyboard not working due to Invalid LT key, adding it after the assignment there is an Invalid HT key both of which stop the board working. removing the new layer result...

manic glacierBOT
#
Bob837217
[adafruit/circuitpython] New comment on issue #8425: In KMK python firmware, Specifying "KC.MO(1)" in code.py crashes CircuitPython

Hi @tannewt @DLBPointon
It has been a while since I last did any work on that keyboard project. I have not tested the latest beta etc and do not know if this is still an issue or not. I did things in a different way in KMK to avoid any issue for my project in the end. I also note that for my final version (last I looked at it), things were stable for a couple days uptime, then the rp2040/firmware would freeze and need to be power cycled to regain functionality. I have moved onto other pe...

manic glacierBOT
#
FoamyGuy
[adafruit/circuitpython] New comment on issue #9363: circuitpython_setboard command python version support

I don't necessarily understand the reason behind the issue, everything underneath our code in the stack trace is a bit cryptic to me. Something is resolving to None, but it's unclear to me what specific thing is actually None. I searched around for a bit with different parts of the stack trace and was able to find references to the same or similar issues, but nothing really stood out as a cause or way to change the code to make it work.

The crux of it is that `resources.files("board-stubs...

lone sandalBOT
manic glacierBOT
#
jepler
[adafruit/circuitpython] New comment on issue #9363: circuitpython_setboard command python version support

OK, here's the reason. Because it does not contain a init.py, board-stubs is a "namespace package". Support for namespace packages was added in 3.10: https://github.com/python/cpython/issues/86295

pip-installable "importlib-resources" supports namespace packages https://github.com/python/importlib_resources so it could be used as a substitute across any supported python version. I don't know if simply adding board-stubs/__init__.py could be a better solution.

With `/tmp/t1/__ini...

short tendon
#

@tulip sleet, sorry to bother, but is there anyone else I can tag to review some open library PRs?

manic glacierBOT
manic glacierBOT
#
FoamyGuy
[adafruit/circuitpython] New comment on issue #9363: circuitpython_setboard command python version support

Knowing the real cause now though, I do think adding init.py in the necessary directories is probably the best solution.
I will test that locally in a little bit and close that PR and make a new one.

I tested that by manually adding a __init__.py inside of both board-stubs/ and board_definitions/. I was worried that the one inside board-stubs/ would take precedent over the __init__.pyi file which is the only other thing in that dir, but inside the IDE it does still c...

spare jacinth
#

The check could be just sys.version_info < (3, 9)

Not commenting on github cuz im on phone 😅

manic glacierBOT
tulip sleet
short tendon
tulip sleet
manic glacierBOT
manic glacierBOT