#circuitpython-dev

1 messages · Page 407 of 1

minor plume
#

@tulip sleet I’m glad it’s not just me.

uneven mango
#

If this was directed toward my question about the sensor, it is i2c, so just a python library using busdevice and register.

tulip sleet
minor plume
#

Excellent! Now I have some things to read instead of blundering through the code by trial and error quite as much.

#

Thank you!

blissful pollen
uneven mango
blissful pollen
stuck elbow
#

it's very meta

mental nexus
#

In the core, what’s the preferred method for printing a “Warning” to the user and keep running (but not causing an Exception and stopping execution)?

tulip sleet
#

what did you have in mind?

orchid basinBOT
minor plume
#

So, 2 hours later, I got the results of the CI run on my first pull request. All of the test runs succeeded save for one, which failed with this error:

Error: Build sparkfun_redboard_turbo for ru took 34.35s and failed
make: Entering directory '/home/runner/work/circuitpython/circuitpython/ports/atmel-samd'
Use make V=1, make V=2 or set BUILD_VERBOSE similarly in your environment to increase build verbosity.

253708 bytes used, -12 bytes free in flash firmware space out of 253696 bytes (247.75kB).
11900 bytes used, 20868 bytes free in ram for stack and heap out of 32768 bytes (32.0kB).

Too little flash!!!

make: *** [Makefile:391: build-sparkfun_redboard_turbo/firmware.elf] Error 255

#

I’m not sure what to do next. Can someone help aim me in the right direction? This doesn’t seem like an actual failure with my code, but I don’t know what to do to fix it.

blissful pollen
#

Did you see this part:
253708 bytes used, **-12 bytes free **in flash firmware space out of 253696 bytes (247.75kB).

#

So your build on that particular board is short 12 bytes with your new code in. Not all that unusual on smaller boards

minor plume
#

Yeah. So what do I do about that?

blissful pollen
#

What is your PR doing? sorry not familiar with it

minor plume
#

It’s adding a check to the neopixel code so if you try to call show() on a neopixel after you’ve called deinit() on it, an exception will be raised instead of the show() silently failing with no message (or in some cases with a crash).

viscid pine
#

it means that board was probably very close to using all flash before, and your change just pushed it over the edge

blissful pollen
#

I'm assuming neopixel is important for that board so just seeing if maybe something else could be turned off. But while I can suggest I wouldn't want to decide

minor plume
#

@viscid pine That was my thought too. I just haven’t a clue what to do about that.

viscid pine
#

probably someone will help you disable some modules on that redboard to make some space

blissful pollen
#

To see what is included with the board builds you can look in the ports/atmel-samd/boards/boardname/mpconfigboard.mk file

minor plume
#

That would be great. And on the good side, this is my first PR and it passed 277/278 of the CI workloads on the first try, so I guess that’s a victory. 🙂

blissful pollen
#

if you look at say the feather_m0_rfm9x board vs the sparkfun_readboard_turbo you can see where the feather turns off the full build (CIRCUITPY_FULL_BUILD = 0) and then enables certain modules only.

#

so in theory you could try turning off something like USB_MIDI and that may save the space. You could try doing a build on your own computer against the sparkfun board to see what could work. Then make a comment in the PR

#

Hopefully that makes some sense

minor plume
#

It does. Let me see what I can come up with. Thanks!

blissful pollen
#

Feel free to ping me if you have a question I should be around for a bit

minor plume
#

Thanks!

tulip sleet
# minor plume Thanks!

also it is for a particular language (you can see that in the log). so do make ... TRANSLATION=ja or whichever language is overflowing.

#

The English translation is smaller than others

minor plume
#

It looks like it was the Russian translation that failed.

viscid pine
blissful pollen
viscid pine
#

audioio and audiobusio are ones i'm seeing disable in a few other boards

#

but i'm thinking, if this one samd21 board went over then others are likely to soon too, maybe some things should be disabled for all samd21 board with small flash

#

sorry if im overloading you with info, im thinking out loud a bit

minor plume
#

Seems likely. OK, trying again without audioio, audiobusio and USB MIDI - we’ll see what happens. And no problem - I’m happy to help solve a larger problem and not just my one bite-sized “first contribution” PR. 🙂

viscid pine
#

random trivia for anyone here: if you ever need more ram on a board without much ram, disabling usb midi helps a lot due to statically allocated buffers

#

(simply disabling it in boot.py doesn't do it)

minor plume
#

Interesting, and good to know. Change pushed, waiting for the 73 million CI runs again. 😂

viscid pine
#

were you able to get it building locally?

minor plume
#

The “Contribute to CircuitPython with Git” Learn guide probably needs a note somewhere that for contributions to CircuitPython itself, the CI process takes a lot longer than “a few seconds to several minutes”. LOL

#

@viscid pine Yup. I built it for that board and translation, and I can’t imagine it didn’t save the 12 bytes I needed. But we’ll find out soon enough.

#

And this is a good learning experience for me when I try to tackle larger issues.

viscid pine
#

another thing about the huge ci is that it amplifies random intermittent issues

minor plume
#

Yeah, for sure.

viscid pine
#

A bunch of boards on my recent PR failed with Error: Internal Server Error occurred while resolving "actions/cache@v2". Internal Server Error occurred while resolving "actions/checkout@v2.2.0". Internal Server Error occurred while resolving "actions/setup-python@v2". Internal Server Error occurred while resolving "actions/upload-artifact@v2"

#

some internal part of github actions just falling over

minor plume
#

Oops.

tulip sleet
#

I am going to turn off something else instead. This board is a lot like Metro M0, and there is something easier to turn off to gain small space

minor plume
#

@tulip sleet Sure thing. Just let me know what you need me to do and I’m happy to do it. I don’t even know the range of things I can turn off, so I’m sort of improvising as I go along right now.

#

Does the CI process build every combination of board and language that exists in the ecosystem?

tulip sleet
minor plume
#

Wow. No wonder that initial CI run takes a long time!

tulip sleet
#

it is like 2500-3000 builds

#

Each language doesn't have to recompile every file, only the files with messages

minor plume
#

That’s still a lot of stuff.

viscid pine
#

could you make a "dummy" language with the biggest translation for each string and build that?

#

or does building every language do more than just testing for size

tulip sleet
#

the strings needed vary because it depends on which features are turned on or not

#

not per language, but per board

minor plume
#

This process is making me wish we had a CI process for the work I do for my day job. (I’m a Data Engineer maintaining a date warehouse solution for a healthcare company, and our current platform’s tech stack is not really amenable to implementing a CI solution, unfortunately. Which hurts my heart, but I can’t change it.)

mental nexus
# tulip sleet We don't really have any examples of this, because the user might not have a ser...

I’ve got a build running that allows access to the REPL displayio group, so you can view the REPL screen on your display at the same time as the result of your other displayio code. I know its dangerous to toy with the REPL since you could mess it up and you won’t get feedback as to why it’s gone. So if your code ends or you call display.show(None) after you’ve placed the REPL group into another group, I want to rip it back out of any groups and put it back where it belongs. Anyway, I thought it would be useful to raise a warning to the user that something dangerous happened to get back access to the REPL. Probably not a very important use case to add something new.

manic glacierBOT
#
dglaude
[adafruit/circuitpython] Issue opened: #5947 Neopixel Broadcom on PiZero1W misbehave or crash

CircuitPython version

Adafruit CircuitPython 7.2.0-alpha.1-243-gf4872d346 on 2022-01-29; Raspberry Pi Zero W with bcm2835
Board ID:raspberrypi_zero_w

Code/REPL

"""CircuitPython Essentials NeoPixel example"""
import time
import board
import neopixel

print("time.sleep(20)")
time.sleep(20)

pixels = neopixel.NeoPixel(board.D18, 20, brightness=0.3, auto_write=False) # Raspberry Pi wiring!

RED = (255, 0, 0)
YELLOW = (255, 150, 0)
GREEN = (0, 255, ...
minor plume
#

Ok, next problem. My new code failed the pre-commit formatting test in a bunch of files under tests/circuitpython-manual but I didn’t change any of those files and “pre-commit run --all-files” passes locally. Waiting to see if the rest of the CI run succeeds.

tulip sleet
#

you can pull my changes

#

as a maintainer I can add commits to PR;s

minor plume
#

Ok. Thank you!

tulip sleet
#

I turned off atexit and getpass for these SAMD21 "full builds" (which cannot be so "full")

minor plume
#

Ah. Nice!

tulip sleet
#

there are other builds that are very close to full (a few dozen bytes away) so it was inevitable

minor plume
#

So pull the adafruit main branch into my local repo?

tulip sleet
#

just pull from your own branch

#

but I am seeing the same precommit issues. It's because the black formatter just published a new version today.

#

like a few minutes ago, I think

minor plume
#

If I just do a “git pull” in my local repo, git says my working branch has no tracking information.

manic glacierBOT
tulip sleet
#

you might google that

#

you should be able to recover

minor plume
#

Ah. I figured it out. For some reason when I pushed the local branch to my repo, The local branch and remote branch were not associated with one another. Got it fixed now. Weird.

tulip sleet
#

black decided to put spaces around the ** operator in Python now

minor plume
#

Lovely.

manic glacierBOT
#
dhalbert
[adafruit/circuitpython] New comment on pull request #5946: Fixes neopixel show\(\) after deinit\(\) not raising an exception

Extra changes:

  • ru translation failed on one SAMD21 board with external flash. Other builds were very close to overflowing. Disabled atexit and getpass for these SAMD21 CIRCUITPY_FULL_BUILD builds. This should give us some breathing room.
  • black went to version 22.1.0 in the past few minutes, so pre-commit Pytyhon black checks were failing. Only issue was it wanted spaces around the ** operator.
viscid pine
#

thanks for being awesome dan

minor plume
#

Seconded. Thank you!

manic glacierBOT
minor plume
#

Yay! My first PR merged! 🎉 Thanks everyone for the help!

blissful pollen
minor plume
# blissful pollen Congrats!

Thank you! Been wanting to contribute to CircuitPython for a long time, and got thrown for a loop by pandemic stuff and also a major medical crisis last year. Glad I’m able to start doing it now.

viscid pine
#

just received the new Seeed_XIAO_nRF52840_Sense board and circuitpython doesn't work, nothing on usb or bt, just a few led flashes

#

how do i go about debugging this?

tulip sleet
viscid pine
viscid pine
#

still trying to understand why he changed it to 7.0.1

#

uf2 doesn't contain sd, had to create my own uf2 from the combined hex file that has the sd

west birch
#

hey! I'm having trouble finding the usb_hid driver or helper for circuit python. I know it comes with the uf2 file but I installed blinka on my rasberry pi pico running micropython and I wanted to send keyboard inputs. I can't find the library I need to put in my folder. So far I put in the adafruit_hid folder in my lib folder but importing usb_hid gives me an error

viscid pine
#

usb_hid is part of circuitpython, you can't use it with blinka

west birch
#

ah

#

thank you

#

is there a way to send keyboard inputs in micropython?

viscid pine
#

i'm not sure

west birch
#

Alrighty, thanks for the help though!

viscid pine
#

i think that might be only on stm32 boards though

west birch
#

The article imports a "pyb" library which apparently is only supported on PyBoards

orchid basinBOT
orchid basinBOT
manic glacierBOT
orchid basinBOT
manic glacierBOT
orchid basinBOT
solar whale
#

hmm -- latest build of CP won't boot on my PyPortal --- 7.2.0-alpha.1 does boot...trying to narrow it down

manic glacierBOT
mental nexus
tulip sleet
#

was doing I2C testing on Metro M4

#

based on a bisect

manic glacierBOT
solar whale
#

Agreed. That was what I found as well

minor plume
#

I have a Metro M4 Express handy if there’s anything I can do to help with testing or whatever.

minor plume
#

And an ItsyBitsy M4. Somehow those seem to be the only SAMD51 boards I can find at the moment.

manic glacierBOT
#
dhalbert
[adafruit/circuitpython] Pull request opened: #5951 Track creation of board\.I2C\(\), etc\.
  • Fixes #5950.

Before #5422, the singletons for board.I2C() were stored as i2c_singleton, etc. and were NULL if not instantiated. #5422 allowed multiple board, objects, but removed the separate singleton storage slots, and relied on checking whether the objects were deinited() or not to see if they were valid. But all-zero objects were not recognized as deinited() in some ports (which caused crashes). For instance, in atmel-samd, deinited() checked for a pin being NO_PIN ...

manic glacierBOT
manic glacierBOT
manic glacierBOT
digital shoreBOT
#
adafruit
Owner

adafruit#3230

Category Channels

8

Text Channels

61

Voice Channels

6

Members

32941

Roles

36

ID: 327254708534116352 | Server Created ·
tulip sleet
#

<@&356864093652516868> Reminder that there is a CircuitPython Weekly meeting today. It is at the normal time of 11am Pacific / 2pm Eastern here on Discord. Everyone is encouraged to attend! Please add your hug reports and status updates even if you’ll be attending the meeting - it’s super helpful! If you are unable to attend but would still like to include updates, feel free to include them in the notes and we’ll read them off during the meeting. Hope to see you there! https://docs.google.com/document/d/1YU2kvEr3cFDF6GpmVm_l500xPcxhSu1ox3iXYrL_ZUo/edit?usp=sharing

Google Docs
CircuitPython Weekly for January 31st, 2022

CircuitPython Weekly for January 31, 2022 Welcome to the CircuitPython Weekly meeting notes! Feel free to add your Hug Reports and Status Updates early. During the meeting, we go through them as a round robin sorted by username. If you can’t make the meeting and would still like to participate...

proven garnet
#

Won't be able to make it, but early group hug to everyone! Love this community! love_parrot

orchid basinBOT
idle owl
#

@tulip sleet Is there a preferred board for asyncio? i.e. I'm going to be templating it and wanted to know if there's a specific board I should be using.

#

I planned to do it this week and did not think through bringing the right hardware. So I'm trying to sort out what I need. Might end up running home instead, but if I can fake it with stuff I can get from the shop, that would be ideal.

tulip sleet
idle owl
tulip sleet
#

I tested on a Metro M4 a lot

idle owl
#

Fair enough

tulip sleet
#

but an RP2040 should be fine too

idle owl
#

I'll give it a try on RP2040 so it's more tested that way along with all of this.

proven garnet
#

My auto-lighting menorah project runs on the RP2040 and I'm loving it! I think my use case does do some networking which I believe is in the pipeline to be more integrated, but otherwise I'm pleased and plan to stick with the Pico!

manic glacierBOT
#
alecharp
[adafruit/circuitpython] New comment on issue #4502: Hardfault when picking wrong UART pin on iMX RT1011

@ladyada sorry to ping you here but I'm trying to understand this in the context of using a KB2040 on a split keyboard. Both half should be communicating in UART and are linked on the Rx pin. With https://github.com/KMKfw/kmk_firmware/blob/e70ce5f4315db02beb25302f4ce027f739fba99b/kmk/modules/split.py#L111-L118, it means that the (default) left half uses the Rx pin as a receiver end, and the right half uses the Rx pin as the transmitter pin.

Is that code (KMK split module) broken with #451...

#
alecharp
[adafruit/circuitpython] Issue opened: #5953 UART connection between two KB2040 boards fails

CircuitPython version

Adafruit CircuitPython 7.1.1 on 2022-01-14; Adafruit KB2040 with rp2040

Code/REPL

import board, busio
busio.UART(tx=board.RX, rx=None)

Behavior

Traceback (most recent call last):
  File "", line 1, in 
ValueError: Invalid pins

Description

On a Sofle keyboard, with KB2040, the two D1 pins of the boards are connected through a TRRS connector (and cable). However, the connection cannot be made.
Trying ...

idle owl
thorny jay
#

I am super late for filling the meeting notes... There is a little risk if I write while someone read. Skip me if it is too draft.

tulip sleet
#

starting in a few minutes; we just finished our internal meeting

thorny jay
#

Love that keyboard, clicky.

onyx hinge
#

whoops I didn't add my notes yet.

tulip sleet
#

kinesis free style edge split kbd

idle owl
#

brb!

turbid radish
#

Anyone have any last minute news/projects/coming soon for the newsletter?

tulip sleet
errant grail
# turbid radish Anyone have any last minute news/projects/coming soon for the newsletter?

https://twitter.com/cedargrovemakr/status/1487653386048598018?s=21

CedarGrove (@CedarGroveMakr)

Wrapped up the dual load cell espresso scale project. Coded in @CircuitPython on @adafruit RP2040 Feather/TFT Wing and all three PyPortals. Touchscreen control of tares and alarms. Settings are stored in non-volatile memory thanks to @foamyguy's library. Custom @oshpark ADC Wing.

Twitter ·
turbid radish
#

No spam, easy to leave

#

Please join

lone axle
turbid radish
lone axle
#

https://xkcd.com/1987/

xkcd: Python Environment
The Python environmental protection agency wants to seal it in a cement chamber, with pictorial messages to future civilizations warning them about the danger of using sudo to install random Python packages.
turbid radish
#

The email can be very helpful in getting info to me

pallid echo
manic glacierBOT
raven canopy
#

is a lurky mclurkyface (hi everyone!!!! hope you're all well)

onyx hinge
#

Maybe our standard verbiage should say "and resellers"... also relevant when it comes to international people where it can make a difference to tarrifs.

idle owl
#

Always! 😄

pallid echo
minor plume
#

Nothing there yet because I'm still setting everything up, but my Twitch stream will be:
https://www.twitch.tv/tammymakesthings

Twitch
tammymakesthings - Twitch

Computer geek, electronics/space nerd, singer/musician, writer/novelist, photographer. Casual gamer (Nintendo Switch). LGBTQIA+ rights are human rights. Black lives matter. Down with racism and fascism. Feminism must be intersectional. Just say no to hatred and bigotry.

▶ Play video
lone axle
minor plume
#

Thanks! I appreciate that!

gilded cradle
#

thanks

mental nexus
#

Here’s a demo of a split screen showing a resizeable REPL/console along with some displayio elements. Perhaps useful as an interactive debug console for graphics.

crimson ferry
#

@solar whale do you have a public GitHub?

minor plume
blissful pollen
lone axle
ember iris
#

Thanks all!

solar whale
crimson ferry
#

@solar whale thanks! you work on a wide variety of interesting projects, and it would be interesting to see some behind-the-scenes 🙂

solar whale
idle owl
#

@lone axle Hello! I have a Thing ™️ I'd like to add to your list, but not for today. I want to try to schedule it for next week or the week after since you're hosting the meeting next week - don't want to overload you all on one day.

#

@lone axle I want to have Eva teach you how to do Adabot patches. Eva's the only one doing them right now, and it's a great task for her when she has the time, but I feel like having others able to do it is a smart move.

#

So I'm thinking two weeks from today.

lone axle
idle owl
#

Excellent. Thanks!

#

I think I'll sit in to learn it myself.

idle owl
#

It's not really a rush thing so that's ok.

lone axle
lone axle
solar whale
idle owl
mental nexus
orchid basinBOT
tulip sleet
#

Here is the notes document for the upcoming Monday, February 7, 2022 CircuitPython Weekly meeting. It is at the normal time of 11am Pacific / 2pm Eastern here on Discord. Everyone is encouraged to attend! Please add your hug reports and status updates even if you’ll be attending the meeting - it’s super helpful! If you are unable to attend but would still like to include updates, feel free to include them in the notes and we’ll read them off during the meeting. Hope to see you there! <@&356864093652516868> https://docs.google.com/document/d/1dlQ-p6WYbsgfwqtEoDw7NfvMKG8CntRsvUSQ-n7ZEb0/edit?usp=sharing

Google Docs
CircuitPython Weekly for February 7, 2022

CircuitPython Weekly for February 7, 2022 Here is the notes document for Monday’s CircuitPython Weekly meeting. It is at the normal time of 11am Pacific / 2pm Eastern here on Discord. Everyone is encouraged to attend! Please add your hug reports and status updates even if you’ll be attending th...

tulip sleet
manic glacierBOT
manic glacierBOT
#
kmatch98
[adafruit/circuitpython] Pull request opened: #5954 Expose splash group, add function to resize REPL terminal

This exposes the REPL's display group elements as supervisor.splash. Also, this adds a function to resize the dimensions of the REPL terminal window. This allows the user to view the REPL and console print() output as well as view any display elements at the same time.

Warning: This is somewhat dangerous since exposing the REPL display elements allows the the REPL to be relocated where it is no longer visible.

Here is an example creating a split screen with the REPL at the b...

sterile crag
#

Problem with countio: I have several Wemos/Lolin S2Mini board. According to the latest CircuitPython page, countio is available on the S2mini. I can import it ok. Unfortunately "Edge" does not seem to be defined. The documentation says there should be a class countio.Edge. When I run

>>> import countio
>>> dir(countio)
['__class__', '__name__', 'Counter']

I poked around in other modules without success.
When I run the example I get

stuck elbow
#

check the docs for the version you flashed

#

readthedocs shows the latest

#

there was a change in the api

sterile crag
#

I am running
Adafruit CircuitPython 7.1.1 on 2022-01-14; S2Mini with ESP32S2-S2FN4R2

stuck elbow
#

you can select the version in the lower left corner

sterile crag
#

Ok. 7.1.X does not have Edge.
Ahhh. How do I find which version is "latest"?
From the download page for the LOLIN S2 Mini I see a 7.2.0-alpha.1. Is that what the latest documentation is for?

viscid pine
#

it's either that, or the "absolute newset" below that

stuck elbow
#

latest is the repository main branch

#

you get that if you compile it yourself, or download the builds from ci

crimson ferry
#

I recall a discussion about making it so that (some) Safe Mode occurrences could be intercepted / detected by User Code so it could automatically reset, but I couldn't find it in Discord or Issues. Does anyone know where that discussion was? (it would be a really useful feature)

stuck elbow
#

iirc it was the battery brownout detection

#

in the issue tracker

manic glacierBOT
#
tannewt
[adafruit/circuitpython] New comment on issue #4574: Teensy 4\.0/4\.1 hardware RTC does not keep time unpowered while having vbat supply

I haven't worked on this code in quite a while. We've never officially started supporting it with Adafruit-funded folks so it's unlikely we'll get to it soon.

It does look like the rtc module talks to the internal RTC. The place to look is here: https://github.com/adafruit/circuitpython/blob/main/ports/mimxrt10xx/common-hal/rtc/RTC.c

That's the code that configures the RTC of the device. supervisor/port.c may have some clock initialization in it as well. Losing the value may be a po...

idle owl
#

@slender iron My 2022 post is done, but it's not anywhere accessible yet. I'm not feeling great so I think I'm done for the day. I'll get it up on the blog first thing tomorrow. Wanted to give you a heads up since it seemed like you might be waiting on me.

slender iron
#

@idle owl no worries. feel better

idle owl
#

Thanks, working on it

manic glacierBOT
#
tannewt
[adafruit/circuitpython] Pull request review submitted: #5948 Spi flash fix

I don't think you want to do this. Removing this configures will break cases where the flash is on a shared SPI bus with a device that has different settings. This is rare but possible.

Instead, the port should guard against repeated calls and exit early. https://github.com/adafruit/circuitpython/blob/main/ports/atmel-samd/common-hal/busio/SPI.c#L216-L222 This protects the SPI from similar issues when used by SPIDevice which also calls configure before each transaction.

#
tannewt
[adafruit/circuitpython] New comment on issue #5953: UART connection between two KB2040 boards fails

Is D1 only for reception and cannot be used as the transmitting pin? Do this means that the crossing of the Rx/Tx connections have to be done on the hardware and cannot be made on the software?

With busio.UART() yes because it uses the UART peripherals of the RP2040. However, you can use pio to do either UART TX or RX.

For TX, ensure you have the absolute newest version of CircuitPython. I recently added support for a PIO feature used by UART TX. There classes that present the same ...

#
prplz
[adafruit/circuitpython] Pull request opened: #5955 Only enable is31fl3741 for led glasses board

From the description of #5584, the is31fl3741 is supposed to only be enabled for the led glasses board. I found that it was mistakenly enabled by default on all nrf boards.

Also fixed the an uninitialized variable warning that prevents the module from building in debug mode. I don't actually know anything about this device and I just assumed the variable was supposed to start at zero. I don't own the led glasses board, so I was unable to test.

I confirmed the led glasses build still i...

#
tannewt
[adafruit/circuitpython] New comment on issue #5881: BLE API: returning the address for an existing connection

Ok thanks. I've got an iPhone that I tested this with.

Example code (will PR to the BLE repo too):

# SPDX-FileCopyrightText: 2022 Scott Shawcroft for Adafruit Industries
# SPDX-License-Identifier: MIT

"""
This example does a generic connectable advertisement and prints out the
manufacturer and model number of the device(s) that connect to it.
"""

import time
import adafruit_ble
from adafruit_ble.advertising.standard import Advertisement
from adafruit_ble.services...
#
anecdata
[adafruit/circuitpython] Issue opened: #5956 Safe Mode: mechanism for user code to recover without manual intervention

Feature request based on some discussion in this issue:
https://github.com/adafruit/circuitpython/issues/2694#issuecomment-740624686

It would be very useful to have a mechanism for user code to intercept or detect Safe Mode, and trigger its own reload or reset. Or, have a user-settable indicator not to go into Safe Mode for at least some causes. Ideally we'd want to identify and fix every last intermittent software cause of Safe Mode, but sometimes they are rare and very difficult to d...

#
tannewt
[adafruit/circuitpython] New comment on issue #5881: BLE API: returning the address for an existing connection

Here is a slightly improved version that sets values the other device can read:

# SPDX-FileCopyrightText: 2022 Scott Shawcroft for Adafruit Industries
# SPDX-License-Identifier: MIT

"""
This example does a generic connectable advertisement and prints out the
manufacturer and model number of the device(s) that connect to it.
"""

import time
import adafruit_ble
from adafruit_ble.advertising.standard import Advertisement
from adafruit_ble.services.standard.device_info im...
manic glacierBOT
manic glacierBOT
#
kmatch98
[adafruit/circuitpython] New comment on pull request #5954: Expose splash group, add function to resize REPL terminal

Thanks for the constructive comments.

  • being able to access the current Group would be more general. Yet if you’re somewhere already deep in your display code, it would require you to do a display.show(None) to access a handle to the splash group. Probably not a huge deal and you could turn off auto_refresh so it would be hidden to the user. Let me know your decision and I’ll make edits.
  • I assume that when code.py stops that it essentially does a display.show(None). I’ll hunt for...
#
ThomasAtBBTF
[adafruit/circuitpython] New comment on issue #5881: BLE API: returning the address for an existing connection

Yes, maybe.
But the other device is an IPhone / IPad and IOS connects to a keyboard if the user selects this in the Bluetooth Settings.
The use case is:
The (Blind) user has one Braille-Keyboard and an IPhone and one IPad he want's to pair his Braille keyboard with his IPhone and his IPad.
The user is using Voiceover as the screen reader on both devices to get Speech feedback and maybe a Braille Display to get braille dots from Voiceover. All this works if he also has two Braille Keyboa...

manic glacierBOT
main furnace
#

When Scott fixes it, it stays fixed.

manic glacierBOT
#
dhalbert
[adafruit/circuitpython] Pull request opened: #5958 I2c common hal write read

I believe this fixes all these ESP32-S2 I2C issues:

  • Fixes #5934.
  • Fixes #5906.
  • Fixes #5871.
  • Fixes #5680.

As suggested by @tannewt in https://github.com/adafruit/circuitpython/issues/5680#issuecomment-1024511731, this PR provides a new common_hal_busio_i2c_write_read() operation. shared-bindings/busio/I2C.c no longer calls write and read separately itself, but calls the combined routine.

  • On ESP32-S2, the common-hal I2C routines use the new ESP-IDF `i2c_master_device....
viscid pine
#

where are the vid/pid of circuitpython boards used?

#

or more directly, why is it important that they are unique?

manic glacierBOT
#
gamblor21
[adafruit/circuitpython] New comment on pull request #5955: Only enable is31fl3741 for led glasses board

The missing initialization is fixed in #5726 which has a lot of changes to the whole structure of the is31fl3741 module. The PR is still waiting to be merged.

As to which boards it is enabled on I cannot remember the initial discussion but an product like https://www.adafruit.com/product/5201 can take advantage of this module as well.

At the time I submitted the first PR I had only been working on NRF boards so had not had the time to test it on other types of boards like ESP/RP2040...

blissful pollen
# viscid pine or more directly, why is it important that they are unique?

They are used by the host system to identify the board. Similar to how your computer can know it is a Logitech mouse plugged in and process it correctly.
I'm not sure if anything currently checks it or not but if they shared VID/PIDs then the USB host would have no way to differentiate different types of boards

viscid pine
#

"I don't think it's much of an issue. Mu no longer depends on VID/PIDs to identify boards. "

#

so mu used to want it for some reason?

blissful pollen
#

I believe to set the mode. I haven't upgraded in forever and it detects a device to change the mode. Someone else probably has a much better answer then I do!

jaunty juniper
#

Mu used to depend on the vid/pid to find a board, and its serial port when opening the serial panel to connect to the REPL. Now it uses the Adafruit_Board_Toolkit module

#

and prompting to switch mode to Circuitpython when detecting a board

orchid basinBOT
manic glacierBOT
mental nexus
onyx hinge
#

.. no, boards.txt doesn't associate pid 0x8023 with the Adafruit Feather M0 Express, so Arduino environment support may not be the reason.

orchid basinBOT
manic glacierBOT
#
FoamyGuy
[adafruit/circuitpython] New comment on pull request #5954: Expose splash group, add function to resize REPL terminal

This is awesome! I have not tried it out on a device yet, but I'm looking forward to it.

Looking over the code and example presented above my only question ATM is around naming.

Is the "splash" term previously used somewhere for this concept (or something similar)?

I have seen it used as the name for a Group object in some example code scripts. But I've always thought that when originally written the object may have been named that way because it was used for a "splash screen" whic...

idle owl
#

@slender iron I have a link for you. It's not live yet because of the tricksy scheduling we have to do so Twitter doesn't freak out, but it should be live in 15 minutes or so. Ping me when you're around, since I imagine it'll be live by then and I'll drop you a link. (That way it's not a 404 here for the next 15 minutes.)

slender iron
#

@idle owl thanks! I just got to my desk so feel free to ping me whenever it's live. I'm not in a rush

idle owl
slender iron
#

thank you

manic glacierBOT
#
kmatch98
[adafruit/circuitpython] New comment on pull request #5954: Expose splash group, add function to resize REPL terminal

Thanks @FoamyGuy for the good feedback. Per @tannewt comments above I modified the functionality and naming. To clarify, here is the current status of what is added and the names.

  • display.root_group: the root displayio Group that is currently displayed. To access the “splash” screen group (with Blinka and the REPL terminal output) you first have to do display.show(None) and then procure the display.root_group. If you don’t want that to disturb the display you can turn auto_refres...
slender iron
manic glacierBOT
#
EmergReanimator
[adafruit/circuitpython] New comment on pull request #5948: Spi flash fix

Removing this configures will break cases where the flash is on a shared SPI bus with a device that has different settings. This is rare but possible.

I see. I also was not sure about this change. But I was not aware that shared SPI bus use cases are possible with circuitpython.

Instead, the port should guard against repeated calls and exit early. https://github.com/adafruit/circuitpython/blob/main/ports/atmel-samd/common-hal/busio/SPI.c#L216-L222 This protects the SPI from similar ...

lone axle
manic glacierBOT
slender iron
manic glacierBOT
slender iron
#

the unique parameter is the current time

lone axle
#

Perfect! I was just thinking: it would be ideal if NASA provided the data directly as JSON.

slender iron
#

🙂

lone axle
#

I looked a few weeks back when I built the MagTag project but couldn't figure out where the data came from.

idle owl
#

Have we done the adding a board.I2C1() thing yet to CP?

slender iron
#

it should be in absolute latest

idle owl
#

Oh!!

#

Excellent.

#

I will find out.

slender iron
#

I'm not sure all boards that could use it do though

idle owl
#

Fair enough.

#

I'll keep that in mind.

#

Nice!

#

Nope, not all boards do.

#

You are correct.

#

QT Py RP2040 does though.

jaunty juniper
#

yup there's board.STEMMA_I2C on the QT PY 2040

#

it should be added to the ESP ones in the same way

idle owl
#

Looks like ESP32... yes that.

jaunty juniper
#

and add it to to other boards as a simple alias to board.I2C if that's what we want to do

idle owl
#

Is there an issue open for this?

#

I'd like to start using it, but probably shouldn't until it's applied to all relevant boards.

jaunty juniper
idle owl
#

Right on.

#

Looks a bit more complicated than I'm used to in the core.

#

I'm sure I could figure it out, but not right now. Need to finish this guide.

manic glacierBOT
jaunty juniper
#

I did start the alias thing, like a year ago, it's basically just adding this line to all the boards that have the stemma connector on the same pins as the I2C pins, but I must be missing a lot now.

    { MP_ROM_QSTR(MP_QSTR_STEMMA_I2C), MP_ROM_PTR(&board_i2c_obj) },
idle owl
#

I like that idea. Consistency is good.

manic glacierBOT
#
tannewt
[adafruit/circuitpython] New comment on pull request #5954: Expose display’s root\_group, add function to resize REPL terminal
  • One other idea I toyed with is whether resizing the REPL should also forward over any text data from the previous incarnation of the Terminal tilegrid. Currently the tilegrid is obliterated and a fresh one is created. However that would require allocating the new tilegrid memory, copying the tilegrid data that will fit and then deallocating the previous tilegrid. Is it worth this effort/code space to keep the REPL text even when the size changes?

I think this would be way more work tha...

blissful pollen
#

So just making sure for a call to common_hal_busio_i2c_write() the stop bit flag is now gone? (just have to drop it in my PR merge conflict if so)

slender iron
#

ya, it's always true now

#

if you have a case where it's false, then you should combine it and the following read into a write_read

blissful pollen
#

thanks they're just writes so easy fix

manic glacierBOT
#
tannewt
[adafruit/circuitpython] Pull request review submitted: #5954 Expose display’s root\_group, add function to resize REPL terminal

Thank you for the iteration. It looks good!

The builds are failing because supervisor_stop_terminal is only linked in for boards that support displayio. I've added a suggestion that should fix it.

I agree the term "splash" isn't ideal but that's what I used when I added it. Maybe "serial_terminal" is clearer? I don't care if you change it or not in this PR. The term is only used internally so we're not committing to it.

#
anecdata
[adafruit/circuitpython] New comment on issue #5956: Safe Mode: mechanism for user code to recover without manual intervention

safemode.py would work fine. Anything that the user can control.

watchdog module doesn't catch these since the code stops feeding it, nor do Exceptions. I'm not sure if there would be any interactions / complications from having both. Nothing comes to mind.

Most (all?) of my Safe Mode occurrences are intermittent but fixed with a restart. 95% sure it's not power or hardware. The idea is that the device can self-recover with a software restart.

idle owl
#

@tidal kiln I need some help with hex addresses again. The base address on the Arcade QT doesn't follow the convention we made the table for, and so I'm confused because it's an ATTiny817, which I thought was a fixed concept in this area, but apparently that is wrong.

#

They're all tied to GND to begin with from what I can tell. So I assume it starts at the base address and goes up from there. But that means a new table. And me figuring out what the higher-end addresses are in hex.

crimson ferry
#

something like this?```py

starting_address = 0x80
soldered_address_pins = 0b0111 # {first bit not used) A2 A1 A0
final_address = starting_address + soldered_address_pins
final_address
135
hex(final_address)
'0x87'```

idle owl
#

Maybe?

#

Seems like that's not enough addresses

#

Let me post the current table

#

It's like this except that it starts at 0x3A instead of 30.

crimson ferry
#

ah, so in my example, the first bit would be A3

idle owl
#

So I would need to know what comes after 0x3F and how far beyond it I need to go

crimson ferry
#

the table isn't intuitive to me, because it has the least significant bits on the left, rather than like we do base-10 numbers with the tens then ones moving to the right

stuck elbow
#

after 3F comes 40 of course

idle owl
#

Hmm

stuck elbow
#

just like in base-ten

idle owl
#

I did it for beginners, I guess. Carter asked for it because of support issues

#

and told me how to do it

tidal kiln
#

@idle owl ok. back. looking...

idle owl
#

I'm confused why this one starts in the middle though.

#

To begin with.

#

Same seesaw chip and firmware afaik, so...??

crimson ferry
#

what's the starting address of this new thing?

tidal kiln
crimson ferry
#

"0x3V"? that's a new one

idle owl
#

Yeah

#

0x3A....

#

Did I typo it?

crimson ferry
#

in the guide... "0x3V -- cut traces A0 and A1"

idle owl
#

There's no guide yet, I'm writing it...

#

Which guide?

crimson ferry
#

the one cater linked

idle owl
#

oh in the product copy?

#

I'm confused

crimson ferry
idle owl
#

Ah.

#

That might be a typo.

crimson ferry
#

feedback left

idle owl
#

Because in the code it's listed as 0x3A through 0x3D I think.

tidal kiln
#

there all tied to GND

#

what base address is it showing up as?

idle owl
#

0x3A

#

At least that worked in code.

#

I'm not sure I scanned.

#

I can.

#

I2C addresses found: ['0x3a']

#

Traces are intact.

idle owl
crimson ferry
#

just an aside, due to routing constraints I assume, some boards have the addresses in AN ... A1 A0 order, other boards have them in reverse A0 A1 ... AN order, so yeah definitely a chart or aid is super useful in guides

idle owl
#

Hmm oki

tidal kiln
idle owl
#

You beat me to linking it.

#

Yeah.

#

So, back to part of the original question

#

Do the extra addresses follow on up the line of hex values?

tidal kiln
#

yes

idle owl
#

So the table won't work, but it should be consecutive?

#

Ok

tidal kiln
#

yep

idle owl
#

Thanks, @tidal kiln . And thanks @crimson ferry for popping in as well!

#

@tidal kiln Worth making a second table?

tidal kiln
#

yes please

idle owl
#

I'd need you to verify it.

tidal kiln
#

i've tried explaining hex notation and the address bits, etc.

#

it never works

idle owl
#

Fair enough

#

I wouldn't get it

tidal kiln
#

so need a table that just maps jumpers to "weird number that starts with 0x thing"

idle owl
#

so I see where you're coming from.

#

Should I flip the table per anecdata's suggestion?

#

At least I think that's what they were getting at

crimson ferry
#

ooh tough call... natural place order, or match the order on the silk

idle owl
#

I think it's as simple as dragging columns around, FWIW.

#

Silk is probably best

#

but Carter would know better

tidal kiln
#

i'd say keep as is. agree it's confusing for those that can read hex/bits naturally. but target audience is opposite.

idle owl
#

Yeah that.

#

@tidal kiln

Screen_Shot_2022-02-01_at_15.27.31.png
tidal kiln
#

yep

idle owl
#

keen. Thanks.

#

@tidal kiln Might ping you with the guide copy as well, since I'll need to update it to reflect the current starting address. Or wait. Should I even bother with that whole explanation since the table is there? Are folks learning from it? Or is it simply confusing?

tidal kiln
#

you should mention base address, since that'll be the default that needs to be used

#

the table would be used for any multi-board configs

#

do you have a code snippet near table?

#

that shows passing in non-default address?

#

"cut jumpers as you need/wish. use table to figure out resulting ADDR. use that value in call to begin()"

manic glacierBOT
idle owl
#

The examples are on the CP and Arduino pages.

#

Or will be anyway.

#

You have to pass in an address regardless or it doesn't work

tidal kiln
#

oh yah. would need CP code snippet also.

idle owl
#

so the examples will cover that

#

I don't think it makes sense on the Pinouts page

#

But mebbe am wrong on that.

#

Could link to the examples on the respective code pages

#

CP will have a two-board example

#

JP's Arcade Synth guide will stand as the multi-board Arduino example.

#

Still have to fix the Arduino example to match the silk.

#

@tidal kiln I went with the same thing as the other seesaw QT breakouts: ```In the center of the back of the board are four address jumpers. These jumpers allow you to chain up to 16 of these boards on the same pair of I2C clock and data pins. To do so, you cut the traces between the the two pads.

The default I2C address is 0x3A. The other address options can be calculated by “adding” the A0/A1/A2/A3 to the base of 0x3A.

A0 sets the lowest bit with a value of 1, A1 sets the next bit with a value of 2, A2 sets the next bit with a value of 4, and A3 sets the high bit with a value of 8. The final address is 0x3A + A3 + A2 + A1 + A0 which would be 0x49.

If only A0 is cut, the address is 0x3A + 1 = 0x3B

If only A1 is cut, the address is 0x3A + 2 = 0x3C

If only A2 is cut, the address is 0x3A + 4 = 0x3E
If only A3 is cut, the address is 0x3A + 8 = 0x42

So for example if A2 is cut and A0 is cut, the address is 0x3A + 4 + 1 = 0x3F.

The table below shows all possible addresses, and whether the pin should be high (cut) or low (closed).```

#

Check me if you have a moment.

tidal kiln
#

@idle owl lgtm

idle owl
#

Excellent, thanks

stuck elbow
#

wait, is the bit set when the jumper is cut, or when it's not cut?

idle owl
#

Cut?

#

They start connected...

#

If that answers your question

#

I'm fuzzy on a lot of this, which is why I've been asking for so much help. Trying to answer, not trying to be snarky if it's coming across that way.

stuck elbow
#

is there a schematic?

tidal kiln
idle owl
#

Yes. Not sure it's posted yet.

#

I can link it here though, give me a minute.

stuck elbow
#

the pins are pulled up

#

so the connection must be to gnd

idle owl
#

See if this works. Big file.

AQT_sch.png
#

Worked.

tidal kiln
stuck elbow
tidal kiln
#

that enables it. see other link above for actual addr incrementing code.

stuck elbow
#

right, but how is that variable set in this particular firmware?

#

without it we don't know

tidal kiln
#

via the "examples"

#

which are like board configs

stuck elbow
#

and by variable I mean constant

tidal kiln
stuck elbow
#

ok then, sorry, just checking

tidal kiln
#

the code setup is sort of weird for the firmware. sort of a hack. code (implementation) is in header files which are "configured" via the arduino sketches (.ino) in the "examples" folder

jaunty juniper
#

a C question, in the CIRCUITPY_BOARD_BUS implementation, could we remove the need for an additional define with the number of ports and instead use sizeof ?

#ifdef CIRCUITPY_BOARD_I2C_PIN
#define CIRCUITPY_BOARD_I2C (sizeof(CIRCUITPY_BOARD_I2C_PIN)/sizeof(board_i2c_pin_t))
#endif
idle owl
#

@tulip sleet The longer leg of the LED is + right? Because more lines needed to make a + than a -? (I'm hoping I remember this properly.)

tulip sleet
#

right, you disassemble the + sign and put the pieces end to end (that's how I think of it)

blissful pollen
#

I had never heard that but now I hopefully will never forget it!

idle owl
#

Ok that's what I thought you explained.

#

Thank you!

tulip sleet
blissful pollen
#

That's a great way

idle owl
#

I apparently haven't forgotten it 😄

#

Had to google it literally every time before. Or Dan- or JP-it.

blissful pollen
#

I'd also use the hook it up to my multimeter to see which way lit up

idle owl
#

I'm sad with multimeters.

#

Don't use them enough to have it stick. Always have to ask for help. Or google it.

#

It's pretty bad. Though I'm missing a lot of fundamentals due to the path I took to get here 😄

blissful pollen
#

If anyone is ever looking for an API for a demo project I just found the NHL has one with a ton of information. Was looking at how those automated goal light works and found they have crazy amounts of data. No API key even required

idle owl
blissful pollen
jaunty juniper
#

I'm trying to figure out what visual image I have, I believe I'm thinking of a faucet, the "plus" side is the longest so it it's stuck in the ground, and the current flows to the other side

idle owl
jaunty juniper
#

is that weird ?

led-faucet.jpg
idle owl
#

Haha not if it works for you.

viscid pine
#

i like it

blissful pollen
#

Good way to remember it as well.

manic glacierBOT
#
digixx
[adafruit/circuitpython] Issue opened: #5961 M4 Express: playing data array through audioio\.AudioOut does loop infinite when dac\.play\(sine\_wave, loop=False\)

CircuitPython version

Adafruit CircuitPython 7.1.1

Code/REPL

import audiocore
import audioio
import board
import array
import time
import math

# Generate one period of sine wav.
length = 8000 // 440
sine_wave = array.array("h", [0] * length)
for i in range(length):
    sine_wave[i] = int(math.sin(math.pi * 2 * i / 18) * (2 ** 15))
    print("Sin", sine_wave[i])

dac = audioio.AudioOut(board.A0)
sine_wave = audiocore.RawSample(sine_wave)
dac...
manic glacierBOT
proven garnet
crimson ferry
#

electrolytic capacitor ± markings get me... often it's "-" written sideways down the capacitor, but many of mine use an outline font and it looks like a "0" (zero), so of course I think 0 (negative) and 1 (positive), and then I end up looking at the lead length anyway to be sure

manic glacierBOT
proven garnet
#

Beats my strategy of buying two of everything, hoping for the best but watching for smoke haha

lone axle
manic glacierBOT
manic glacierBOT
manic glacierBOT
manic glacierBOT
#
walshbp
[adafruit/circuitpython] Issue opened: #5964 GPIO40 missing from boards module in ESP32S3 Builds

CircuitPython version

Adafruit CircuitPython 7.2.0-alpha.1-274-g5435a40c5-dirty on 2022-02-01; ESP32-S3-DevKitC-1-N8 with ESP32S3
Board ID:espressif_esp32s3_devkitc_1_n8

Code/REPL

import board
import pwmio

rl = pwmio.PWMOut(board.IO40)

while True:
    for cycle in range(0, 65535):  # Cycles through the full PWM range from 0 to 65535
        rl.duty_cycle = cycle  # Cycles the LED pin duty cycle through the range of values
    for cycle in range(6...
#
FoamyGuy
[adafruit/circuitpython] Pull request review submitted: #5954 Expose display’s root\_group, add function to resize REPL terminal

I attempted to build from this branch and run on a PyPortal but it seems that I'm getting a hard crash.

Immediately after the UF2 finishes pasting, the device reboots as normal. And for just one second I can see the serial output on the display as normal:

pyportal_troubleshoot

But immediately after this the display goes dark and I never see CIRCUITPY or the serial port connect ...

manic glacierBOT
manic glacierBOT
manic glacierBOT
smoky zephyr
#

3741

manic glacierBOT
manic glacierBOT
manic glacierBOT
#
larsks
[adafruit/circuitpython] Issue opened: #5965 Broken "edit on github" links in published content

CircuitPython version

N/A

Code/REPL

N/A

Behavior

N/A

Description

The "Edit on Github" links for core module documentation at https://circuitpython.readthedocs.io/en/latest/ appear to be broken. For example, on any of:

manic glacierBOT
#
Neradoc
[adafruit/circuitpython] Pull request opened: #5966 Add alias board\.STEMMA\_I2C for all boards with qwiic/stemma/QT

This aliases board.I2C and board.STEMMA_I2C on all boards where they are on the same pins. Meaning they both return the same singleton object. I went through the whole list on circuitpython.org, hopefully I didn't make too many mistakes.
I'm a little worried it might be confusing though. Documentation and guides will have to clarify when I2Cand STEMMA_I2C are the same and when they are not.
Note that one sparkfun board already also had a board.QWIIC alias.
Final step for #4638 ?

#
jerryneedell
[adafruit/circuitpython] Issue opened: #5967 RP2040 audiobusio,PDMIN not working with latest build

CircuitPython version

built from tip of main

Adafruit CircuitPython 7.2.0-alpha.1-274-g5435a40c5 on 2022-02-02; Adafruit QT Py RP2040 with rp2040

Code/REPL

Ran demo code from https://learn.adafruit.com/neopixel-mini-vu-meter/coding-the-neopixel-mini-vu-meter

uncommneted line to print PDM Mic readings.


# SPDX-FileCopyrightText: 2022 Liz Clark for Adafruit Industries
# SPDX-License-Identifier: MIT

import time
import array
import math
import ...
manic glacierBOT
#
sgauche
[adafruit/circuitpython] Issue opened: #5968 STM32F412Cx USART3 contradiction between CMSIS and HAL

CircuitPython version

Adafruit CircuitPython commit 9b32d549, Custom board with STM32F412CG

Code/REPL

None

Behavior

I am adding a board to CircuitPython with a STM32F412CG micro on the board. I ran into an error while building CircuitPython about USART3 functions:

common-hal/busio/UART.c:484:9: error: implicit declaration of function '__HAL_RCC_USART3_FORCE_RESET' [-Werror=implicit-fu...
digital shoreBOT
#
adafruit
Owner

adafruit#3230

Category Channels

8

Text Channels

61

Voice Channels

6

Members

32964

Roles

35

ID: 327254708534116352 | Server Created ·
manic glacierBOT
#
ADAMMNBF
[adafruit/circuitpython] Issue opened: #5970 Circuitpython crash

CircuitPython version

Adafruit CircuitPython 7.1.1 on 2022-01-14; Adafruit Matrix Portal M4 with samd51j19
Board ID:matrixportal_m4

Code/REPL

"""
MOON PHASE CLOCK for Adafruit Matrix Portal: displays current time, lunar
phase and time of next moonrise or moonset. Requires WiFi internet access.

Written by Phil 'PaintYourDragon' Burgess for Adafruit Industries.
MIT license, all text above must be included in any redistribution.

BDF fonts from the X....
#
dhalbert
[adafruit/circuitpython] New comment on pull request #5966: Add alias board\.STEMMA\_I2C for all boards with qwiic/stemma/QT

@dhalbert I want to see this happen. Consistency is important, and if we have it on the boards that use SCL1/SDA1, I think having it on all boards would work. We can document that it's an alias if that is desired, but I would love to see this across all boards.

OK, sounds good! I was not sure if there were hidden considerations. It will certainly make Learn Guide code easier. @ladyada -- I will not wait for your reply.

idle owl
#

@tulip sleet Good call on the potential hidden consideration. I discussed it yesterday with Neradoc, so I was already on board with it. 🙂

tulip sleet
#

excellent - I was just not sure if there was some template consideration, etc.

idle owl
#

Even if there is, I'll make it work.

#

I seriously love the idea of STEMMA_I2C()

#

So much more descriptive.

tulip sleet
#

it does sound good, because the code is interchangeable, and unless there are address conflicts, I2C and STEMMA_I2C can be the same bus

idle owl
#

We'll have to deal with the potential for folks trying to use both when they are the same and getting pin in use issues, but we get that anyway for who knows reasons.

tulip sleet
#

and the advantage for boards with separate buses is great, as you say, much more descriptive ("affordant") than SCL1/SDA1

idle owl
#

Exactly.

manic glacierBOT
inland tusk
#

when is the hackchat with the Guys on?

crimson ferry
inland tusk
#

Thanks @crimson ferry for the info. Do you know if they will be on any other channel

crimson ferry
#

not sure

inland tusk
#

I thought I remembered them going to be on a picast also

idle owl
#

Yeah that's a separate thing

inland tusk
#

When is it?

idle owl
#

The Pi Cast Celebrates 10 Years of Raspberry Pi: New Episodes With Ladyada, Eben Upton, and others. Adafruit’s Limor Fried will be on a livecast on February 15, 2022

inland tusk
#

Ok thanks @idle owl

idle owl
#

You're welcome, Charles.

#

@tidal kiln Do you have any of the ATtiny817 STEMMA QT breakouts and a Pi? Have you ever tried to use one on a Pi?

#

@gilded cradle ^^ Same question for you.

#

I assume if one works, they all work, but I haven't tested any of them and am adding it to the Arcade QT guide.

tidal kiln
#

@idle owl yes. no, or at least i don't think so.

idle owl
#

Fair enough. Thanks

tidal kiln
#

should just work. it's just i2c.

#

are you seeing something odd?

idle owl
#

No, I simply realised I've never tested it

#

It should work.

#

But things happen 😄

#

I try not to guide stuff I haven't at least sort of tested

tidal kiln
#

i could test breakout if you want

#

but don't have arcade

#

but also - pretty easy to connect/test if you have pi?

idle owl
#

I don't have a Pi with me or I'd have tested it 🙂

#

I think any of them is a good test

#

since they're all the same base hardware/firmware ish.

tidal kiln
idle owl
#

Ohhh. You only have the base breakout.

#

I was thinking of one of the other things that uses the chip.

#

Would be a good test.

#

I can't even remember if I tested that.

#

There was so much going on in that guide.

#

It's probably fine.

gilded cradle
tidal kiln
idle owl
#

On Pi? Don't think so.

tidal kiln
#

i have one of those. i'll test real quick.

idle owl
#

Ooh ok

#

Thank you!

tidal kiln
# 
(blinka) pi@raspberrypi:~ $ i2cdetect -y 1
     0  1  2  3  4  5  6  7  8  9  a  b  c  d  e  f
00:          -- -- -- -- -- -- -- -- -- -- -- -- -- 
10: -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- 
20: -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- 
30: 30 -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- 
40: -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- 
50: -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- 
60: -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- 
70: -- -- -- -- -- -- -- --                         
(blinka) pi@raspberrypi:~ $ python3 slider_test.py 
Traceback (most recent call last):
  File "slider_test.py", line 13, in <module>
    neoslider = Seesaw(board.I2C(), 0x30)
  File "/home/pi/blinka/lib/python3.7/site-packages/adafruit_seesaw/seesaw.py", line 150, in __init__
    self.chip_id, _SAMD09_HW_ID_CODE, _ATTINY8X7_HW_ID_CODE
RuntimeError: Seesaw hardware ID returned (0xc3) is not correct! Expected 0x55 or 0x87. Please check your wiring.
#

hmm

idle owl
#

Someone ran into this the other day............. And I have no idea what we did to fix it.

tidal kiln
#

going to sanity check on a qt py

# 
Adafruit CircuitPython 7.0.0 on 2021-09-20; Adafruit QT Py M0 with samd21e18
>>> import slider_test
576
289
132
221
#

hmmmm

manic glacierBOT
idle owl
#

@onyx hinge Do you understand enumerate in Python?

#

Pylint is telling me to use it and I don't have time to go through trying to figure it out on my own. Hoping someone else gets it and can help.

tidal kiln
#

it provides an index as well as the iterator

idle owl
#

Um..

#

It wants me to replace this for n in range(len(button_pins)):

#

which is used for button and LED numbers.

tidal kiln
#

for foo in some_list_of_foos:

#

vs.

idle owl
#

so would it be for count in enumerate(button_pins) or something?

tidal kiln
#

for i, foo in enumerate(some_list_of_foos):

idle owl
#

what is foo there though

#

the first one

#

I mean, why use it if I'm only using i in that scenario?

tidal kiln
#

each foo in list_of_foos

#

which can be anything

idle owl
#

seems extraneous.

tidal kiln
#

basic iterator usage

idle owl
#

Fair enough

tidal kiln
#

second provides an integer that can be used for indexing

#

i would be 0, 1, 2, 3 etc.

#

inside the loop

idle owl
#

Right.

#

Which is what I would be using

tidal kiln
#

what is button_pins?

idle owl
#

List of initialised button objects

#

wait

#

no

#

supposed to be buttons

#

which is a list of init button pins

tidal kiln
#

what are you wanting to do inside the loop?

idle owl
# 
    for n in range(len(buttons)):
        leds[n].value = not buttons[n].value```
jaunty juniper
# 
while True:
    for n, button in enumerate(buttons):
        leds[n].value = not button.value
#

*I meant enumerate

tidal kiln
#

^^ yah, that, was just typing same

idle owl
#

Doesn't the buttons.value need the iterator?.... oh

jaunty juniper
#

many typos

idle owl
#

weird. But ok 🙂

#

Thanks!

tidal kiln
#

nope. you get both.

jaunty juniper
#

I would say it's nitpicky in that context, but that's pylint for you

tidal kiln
#

first, think of the non-enumerate version

# 
for button in button_pins:
    button.some_button_func()
idle owl
jaunty juniper
#

yeah typically I use enumerate when I go from for X in Y and realize "oh but I need the index"

tidal kiln
#

enumerate does same thing, but also provides the integer counter

#

so two things are provided in loop. the samebutton as before, and then the i or n or whatever you setup to be the incrementer

#

^^ double yep. you typically get there by a "oh i need index" realization

idle owl
#

Fair enough

#

Thanks @tidal kiln and @jaunty juniper

tidal kiln
#

now back to this fun...

# 
pi@raspberrypi:~ $ source blinka/bin/activate
(blinka) pi@raspberrypi:~ $ python3
Python 3.7.3 (default, Jan 22 2021, 20:04:44) 
[GCC 8.3.0] on linux
Type "help", "copyright", "credits" or "license" for more information.
>>> import board
>>> from adafruit_bus_device.i2c_device import I2CDevice
>>> i2c_dev = I2CDevice(board.I2C(), 0x30)
>>> out_buf = bytes([0x00,0x01])
>>> in_buf = bytearray(1)
>>> with i2c_dev as i2c:
...     i2c.write_then_readinto(out_buf, in_buf)
... 
>>> in_buf
bytearray(b'\xc3')
>>>
idle owl
#

oof

tidal kiln
#

same thing on qt py:

# 
Adafruit CircuitPython 7.0.0 on 2021-09-20; Adafruit QT Py M0 with samd21e18
>>> import board
>>> from adafruit_bus_device.i2c_device import I2CDevice
>>> i2c_dev = I2CDevice(board.I2C(), 0x30)
>>> out_buf = bytes([0x00,0x01])
>>> in_buf = bytearray(1)
>>> with i2c_dev as i2c:
...     i2c.write_then_readinto(out_buf, in_buf)
... 
>>> in_buf
bytearray(b'\x87')
>>>
#

@idle owl you saw this before? was it here on discord or somewhere else?

idle owl
#

bee-are-bee lunch time.

digital shoreBOT
#
adafruit
Owner

adafruit#3230

Category Channels

8

Text Channels

61

Voice Channels

6

Members

32970

Roles

35

ID: 327254708534116352 | Server Created ·
idle owl
#

@tidal kiln Back if you're still fighting with the NeoSlider. I mean, I'm back regardless of what you're doing. But pinging you to let you know.

tidal kiln
#

@idle owl ok, yah, stil looking. seems odd.

#

@gilded cradle are you available to test also?

idle owl
#

@tidal kiln So.. Arduino: Adafruit PID 5296 I2C QT 4x LED Arcade Buttons test! seesaw not found! But I'm on a QT Py RP2040, so maybe that's my issue?

#

I'm not sure how to fix this and therefore test the code.

#

And... wait no, I have a TFT Feather ESP32-S2 with me too. Was going to say don't have any other MCUs with me, but I do. So I could try on that I guess.

digital shoreBOT
#
adafruit
Owner

adafruit#3230

Category Channels

8

Text Channels

61

Voice Channels

6

Members

32971

Roles

35

ID: 327254708534116352 | Server Created ·
tidal kiln
#

don't have any SAMDs?

idle owl
#

No. 😕

#

Wasn't thinking about code not working. Made the obviously false assumption that it would be fine.

tidal kiln
#

go ahead and try the ESP32

#

good to at least try a different mcu

idle owl
#

OK. I'll report back in a few.

tidal kiln
idle owl
#

Yeah, the pins are wrong for the LEDs and switches, but the rest should work.

#

Wrong as in backwards from the silk.

#

I'm going to fix it once GitHub is back up and I can verify it works.

#

Hmm same fail on ESP32-S2.

#

Seesaw not found.

#

Feather TFT STEMMA connector is same SCL/SDA pins.

#

So that's not the issue here.

#

I mean, obviously it worked at some point

#

The forum feedback we got wasn't "it doesn't work" it was "it's not matching the silk."

tidal kiln
#

try an i2c scan

idle owl
#

Ok, where is that code?

tidal kiln
#

it's failing the basic init

idle owl
#

I know where it is in CP, I wrote it. heh.

#

Don't know so much in Arduino.

tidal kiln
idle owl
#

Thank you

#

Nothing.

#

Going back to QT Py simply to try

tidal kiln
#

@idle owl oh yah, need to deal with i2c power on tft feather

idle owl
#

Nothing... I noticed the power LED wasn't on for the breakout when connected to the....

#

oh right.

#

That.

tidal kiln
#

for qt py, could be wire used

#

which qt py?

#

oh. 2040.

idle owl
#

Wire1

#

I think I got it uploaded properly, but I can't tell since it's all the same code. I changed a print to verify.

tidal kiln
#

you changed #define WIRE Wire

#

?

idle owl
#

Yes

#

to Wire1

#

is that not right?

#

(Nothing found either way.)

tidal kiln
#

stemma connector is Wire1

idle owl
#

Ok that's what I did

#

still not finding it.

tidal kiln
#

go back to tft feather

#

it has that lipo sensor on board, so should show something even if breakout is being odd

idle owl
#

It finds another sensor. As a side note.

tidal kiln
#

on qt py?

idle owl
#

ya

#

MSA311. Had it here so I plugged it in

#

Found it on next scan

tidal kiln
#

oh. another sensor tested. ok.

idle owl
#

Yeah

tidal kiln
#

good sanity check

idle owl
#

I thought so

#

So go back to Feather?

#

What magic is it to turn on the port?

tidal kiln
#

i think so. just to verify another way.

#

one sec. it's just a digital pin set.

idle owl
#

Also I thought Arduino did it by default.

tidal kiln
# 
pinMode(TFT_I2C_POWER, OUTPUT);
digitalWrite(TFT_I2C_POWER, HIGH);
#

add that to setup()

idle owl
#

testing now

#

LED is on

#

No devices found.

#

(╯°□°）╯︵ ┻━┻

#

JP got it to work ok on whatever he used for the synth project.

#

So Arduino does work for it...

#

But these are weird boards.

#

I don't have any not weird boards to try.

#

😕 sigh.

tidal kiln
#

remove the arcade

idle owl
#

Ok

tidal kiln
#

scan with nothing attached

#

see if it can at least see the lipo monitor

idle owl
#

No devices found...

#

ohh!

#

It's not seeing that either.

tidal kiln
#

yah, my tft feather is on the way. so i'm guessing at things.

idle owl
#

Oof.

#

Should I try CP as another sanity check?

#

This feels like going in circles.

tidal kiln
#

sure

#

oh..there's a BME also...

idle owl
#

Not on this one

#

There's a spot for it

#

but not populated.

#

CP is seeing something without the arcade attached

#

(on the display)

#

And the arcade when I plug it in

# 
I2C addresses found: ['0xb']
I2C addresses found: ['0xb']
I2C addresses found: ['0xb']
I2C addresses found: ['0xb', '0x3a']
I2C addresses found: ['0xb', '0x3a']
I2C addresses found: ['0xb', '0x3a']
I2C addresses found: ['0xb', '0x3a']```
tidal kiln
#

0x0B is the LC

idle owl
#

Right

tidal kiln
#

hmm...so working in CP

idle owl
#

Yeah

#

I am confuse.

#

I don't have Arduino chops to rely on here to troubleshoot this. 😕

#

That's what I get for not bringing a Metro or something.

tidal kiln
#

probably just missing something simple

#

and i'm lacking the feather to test

idle owl
#

QT Py RP2040?

#

Except that doesn't help because nothing built in

#

nm.

tidal kiln
#

got that. but not arcade.

idle owl
#

Right

#

@gilded cradle Did you say you have an Arcade QT board? Or do you have something else...

manic glacierBOT
#
tannewt
[adafruit/circuitpython] New comment on pull request #5948: Spi flash fix

The thing is that common_hal_busio_spi_construct sets default SPI frequency (baudrate) that differs from the SPI flash settings.

We're careful with the starting baudrate due to picky SD cards. That's where the 250k (IIRC) comes from.

I have added common_hal_busio_spi_configure back to spi_flash_read_data and spi_flash_write_data. But they are called before chip select signal is activated now. Moreover I have added common_hal_busio_spi_configure to spi_flash_init_device...

idle owl
#

I don't know who got one. JP did but I'm pretty sure he's slammed at the moment.

manic glacierBOT
idle owl
#

It's a bit of an unusual breakout, so it's probably not ubiquitous.

gilded cradle
idle owl
#

OK. I guess the Arduino page waits then. Probably until tomorrow when Limor might have more cycles.

gilded cradle
#

Sounds good

idle owl
#

@tidal kiln This is currently happening with CircuitPython: I2C addresses found: ['0x3a'] I2C addresses found: ['0xb', '0x3a'] I2C addresses found: ['0x3a'] I2C addresses found: ['0xb', '0x3a'] I2C addresses found: ['0x3a'] I2C addresses found: ['0xb', '0x3a'] I2C addresses found: ['0xb', '0x3a'] I2C addresses found: ['0x3a'] I2C addresses found: ['0x3a'] I2C addresses found: ['0x3a'] I2C addresses found: ['0xb', '0x3a'] I2C addresses found: ['0x3a'] I2C addresses found: ['0x3a'] I2C addresses found: ['0x3a'] I2C addresses found: ['0x3a'] I2C addresses found: ['0x3a']

#

The lipo monitor comes and goes.

#

The charge LED is inconsistent too. Must be something to do with the monitor itself.

tidal kiln
#

it's powered by VBAT...so that may be part of it

idle owl
#

Ah got it.

#

I'm less concerned about that, heh.

#

But it could explain Arduino not seeing it I guess. But it popped immediately with CP. So... eh.

manic glacierBOT
idle owl
#

Ok. @tidal kiln Thank you for slogging through this with me. I guess I'll move onto something else for now. I'm pretty sure the CircuitPython code for this was more important right now anyway because there were no examples for it previously.

tidal kiln
#

np. sounds good. i'm not helping much without hardware. just sleuthing from the schematics, etc.

idle owl
manic glacierBOT
blissful pollen
manic glacierBOT
slender iron
#

just doing my daily pass @blissful pollen 🙂

manic glacierBOT
#
sgauche
[adafruit/circuitpython] New comment on issue #5968: STM32F412Cx USART3 contradiction between CMSIS and HAL

@tannewt
I was able to build successfully by only updating the stm32f4xx_hal_rcc_ex.h file to the latest from ST's repo.

I noticed that circuitpython has an archived repo as a submodule I think for the ST drivers. Are archived repo's able to be modified? This is the archived repo: https://github.com/hathach/st_driver

I'm not familiar enough with circuitpython yet to say whether it would be feasible to update to ST's latest release... There must have been a reason to point to an archiv...

slender iron
#

@tulip sleet you around? I could use a BLE chat

tulip sleet
#

yes, but Carole just came home; I need to speak with her for a bit

slender iron
#

k, I'll go have lunch

tulip sleet
#

we were planning to go for a walk before it got dark. I can ping you again in about an hour

slender iron
#

sounds good. no rush from me

mental nexus
lone axle
#

I'm guessing I must be close but still need to tweak something else. Will definitely be tinkering around some more with it tonight.

manic glacierBOT
#
atctwo
[adafruit/circuitpython] Issue opened: #5973 ESP32S3 Sleep Behaviour using PinAlarm

CircuitPython version

Adafruit CircuitPython 7.2.0-alpha.1-278-ga04cd6444 on 2022-02-02; ESP32-S3-DevKitC-1-N8R2 with ESP32S3

Code/REPL

import alarm
import board
import digitalio
import neopixel
import time

np = neopixel.NeoPixel(board.NEOPIXEL, 1)

while True:
    np[0] = (50, 50, 50)
    time.sleep(1)
    np[0] = (0, 0, 0)

    # Create a an alarm that will trigger 5 seconds from now.
    time_alarm = alarm.time.TimeAlarm(monotonic_time=tim...
digital shoreBOT
#
adafruit
Owner

adafruit#3230

Category Channels

8

Text Channels

61

Voice Channels

6

Members

32971

Roles

35

ID: 327254708534116352 | Server Created ·
manic glacierBOT
#
kmatch98
[adafruit/circuitpython] New comment on pull request #5954: Expose display’s root\_group, add function to resize REPL terminal

Here's the latest demo code for moving the REPL position and adding some displayio objects:

import board
import displayio
import supervisor
import time


display=board.DISPLAY

# Create a bitmap with two colors
bitmap1 = displayio.Bitmap(display.width//4, display.height//4 - 10, 2)

# Create a two color palette
palette1 = displayio.Palette(2)
palette1[0] = 0xff00ff
palette1[1] = 0xffffff

# Create a TileGrid using the Bitmap and Palette
tile_grid1 = displayio.T...
mental nexus
#

I ran it on a pyportal.

lone axle
manic glacierBOT
manic glacierBOT
lone axle
#

@mental nexus this is such a neat way to interact with displayio! I've taken part of your code and split it off into a helper file that can be used from REPL https://gist.github.com/FoamyGuy/f1e6da12ccb07caac980e5093cfb74eb
so in the REPL you can do:

>>> from screen_splitter import split_screen
>>> import board
>>> displayio_group = split_screen(board.DISPLAY)

and then create and add whatever you want to displayio_group

mental nexus
manic glacierBOT
#
sgauche
[adafruit/circuitpython] Pull request opened: #5976 Change ST drivers to use ST's driver repos as submodules

This will update the ST drivers to the latest from ST, as well as fix the issue https://github.com/adafruit/circuitpython/issues/5968 .

I had to use CMSIS_5 tag 5.4.0, as I was getting the error in the below issue when trying to build CircuitPython.

https://github.com/ARM-software/CMSIS_5/issues/1397

Successful builds:

  • espruino_pico
  • espruino_wifi
  • feather_stm32f405_express
  • meowbit_v121
  • nucleo_f746zg
  • nucleo_f767zi
  • nucleo_h743zi_2
  • openmv_h7
  • pyb_nano_v2
  • py...
manic glacierBOT
manic glacierBOT
#
prplz
[adafruit/circuitpython] Pull request opened: #5977 Bitbang i2c: switch scl to input for reading

Some ports can't read a digital io while it is in output mode, see in the sda_read function it is switched to input, read, then switched to output again.

I used PULL_UP because the sda_read function also uses PULL_UP, but I'm not sure if this is the right thing to do. Should it depend on CIRCUITPY_I2C_ALLOW_INTERNAL_PULL_UP? And if so, should check for pullups during construction like the busio version does?

manic glacierBOT
#
dhalbert
[adafruit/circuitpython] New comment on pull request #5977: Bitbang i2c: switch scl to input for reading

Espressif, nrf, maybe stm and samd.

We have used bitbangio successfully on STM and SAMD, including with devices that need clock-stretching. So does it not work on Espressif? I mean, have you tested it?

Should it depend on CIRCUITPY_I2C_ALLOW_INTERNAL_PULL_UP?

This was added in #5237, due to a particular Espressif dev board that didn't include external pullups despite having on-board sensors. In general, the internal pullups are often marginal for I2C, or in cases like RP2040 (i...

#
prplz
[adafruit/circuitpython] New comment on pull request #5977: Bitbang i2c: switch scl to input for reading

We have used bitbangio successfully on STM and SAMD, including with devices that need clock-stretching. So does it not work on Espressif? I mean, have you tested it?

I don't have any devices that I know for sure need clock stretching, but I can get it to do a lot of weird stuff. One example is if I were to ground SCL and do a scan, I get a list of 100+ addresses instead of the expected TimeoutError: Clock stretch too long that I get with this PR.

manic glacierBOT
manic glacierBOT
digital shoreBOT
#
adafruit
Owner

adafruit#3230

Category Channels

8

Text Channels

61

Voice Channels

6

Members

32994

Roles

35

ID: 327254708534116352 | Server Created ·
manic glacierBOT
#
ThomasAtBBTF
[adafruit/circuitpython] New comment on issue #5881: BLE API: returning the address for an existing connection

Ok,
I checked out the code and tested it in the CP-application for our keyboards and I can use it to decide which one connection to keep and to disconnect the others.

I had hoped, that the .serial_number member of the DeviceInfoservice object would hold the serial_number of the IOS devices, but it didn't.
A quick (and dirty) fix for my problem would be to either put the mac-address string into the serial_number member if the serial number is not available for the device.
Or not so ...

manic glacierBOT
manic glacierBOT
#
dhalbert
[adafruit/circuitpython] New comment on pull request #5908: Add a timeout to busio\_i2c\_scan

This is not working, it looks like i2c_master_cmd_begin caps the delay at 1 second minimum: https://github.com/adafruit/esp-idf/blob/circuitpython/components/driver/i2c.c#L1172

This was a reported as a bug:
https://github.com/espressif/arduino-esp32/issues/5934
https://github.com/espressif/esp-idf/issues/4999

went back in time, and when I2C_CMD_ALIVE_INTERVAL_TICK was first introduced, it was a ceiling, not a floor:
https://github.com/espressif/esp-idf/blame/9075b507b54a99ce73...

manic glacierBOT
manic glacierBOT
#
jepler
[adafruit/circuitpython] Pull request opened: #5979 Enable mdns hostname setting

Having done this, I can

wifi.radio.connect(...)
wifi.radio.mdns_hostname("qtpy")

and then on my desktop computer I can ping qtpy.local. This will also be useful when using wifi boards as http servers, since they could be reached by name rather than by IP address on systems that support mdns.

A second step may be to add support for mdns services, which for instance allow a device to publish that it is a webserver, file server, printer, etc.

digital shoreBOT
#
adafruit
Owner

adafruit#3230

Category Channels

8

Text Channels

61

Voice Channels

6

Members

32996

Roles

35

ID: 327254708534116352 | Server Created ·
manic glacierBOT
manic glacierBOT
#
anecdata
[adafruit/circuitpython] Issue opened: #5980 ping too frequently results in Safe Mode

CircuitPython version

Adafruit CircuitPython 7.2.0-alpha.1-224-gac7a80753 on 2022-01-26; Adafruit QT Py ESP32S2 with ESP32S2
Adafruit CircuitPython 7.2.0-alpha.1-224-gac7a80753 on 2022-01-26; Saola 1 w/Wrover with ESP32S2

Code/REPL

import time
import wifi
from secrets import secrets

DELAY = 0.5  # this is fine, but 0-0.4 or so lead to Safe Mode after a couple of pings

wifi.radio.connect(secrets['ssid'], secrets['password'])
while True:
    print(...
blissful pollen
crimson ferry
#

yes, strong preference for varieties with extra PSRAM (R2, R8)

blissful pollen
#

Thanks, was ordering some other stuff and thought about it

crimson ferry
#

Although the -S3 has more internal SRAM... 512KB, increase from 320KB in the -S2. That extra 192KB is really nice.

blissful pollen
#

Never enough memory 🙂

crimson ferry
#

internal good for wifi and other esp stuff; PSRAM for CircuitPython

#

digikey doesn't actually call out the PSRAM unfortunately, you're left to interpret the model numbers

#

Adafruit has N8R2 in the shop, the others with PSRAM (N32R8 (!) and N8R8) are sold out atm

blissful pollen
#

Thanks! This was at least partly motivated by getting to free shipping on a digikey order I was already making

crimson ferry
#

I have so much inventory for that exact reason

#

"hey, free shipping if I also get this"

blissful pollen
#

Yup, i mean maybe I'll use two of these even if I just wanted one. Maybe even three!

tulip sleet
slender iron
#

thanks for looking @tulip sleet. I think nimble is the right direction. ESP support is helpful

tulip sleet
#

yes, the others didn't look very promising, though it's good to know now, esp the packetcraft ones

manic glacierBOT
#
muendlein
[adafruit/circuitpython] New comment on issue #5924: High Deep Sleep current with QT Py ESP32\-S2

It turns out this is a hardware problem specific to the QT Py ESP32-S2. We know how to fix it, but it's not an easy trace-cutting or something like that. For now, you may want to switch to a different board. We could get somewhat less sleep current by using ESP-IDF "light sleep" (which is not the same as CircuitPython's light sleep), but it would still be several mA.

I seriously have to question the QA procedures because something like should have never happened in the first place.

slender iron
#

@runic crest just replied to you on the forum. I don't have a good answer for you

manic glacierBOT
lone sandalBOT
fiery fossil
#

@slender iron Looks like my PR might be good to go other than the seeeduino_xiao issue I see there is a PR for now

viscid pine
idle owl
#

@lone axle The CircuitPython Org Bundle (library bundle)... Is that worth mentioning when mentioning the library bundles? Or what's the deal there. The name is kind of confusing and requires explanation I think, which this page doesn't have room for. But I don't want to leave it out if it's as important as the others.

#

And I guess I'm not even sure how to describe it in a few words. "It is made up of..." and I don't know how to end that sentence.

#

@slender iron Same question to you, I suppose. ^^ Since you kind of spearheaded the CircuitPython Org in the first place.

slender iron
#

(will get back to you after we're done chatting with jimmo)

idle owl
#

Ok thanks.

lone axle
# idle owl <@!382939733107408897> The CircuitPython Org Bundle (library bundle)... Is that ...

It's made up of libraries maintained collectively by the community (rather than individually by single members like the community bundle).

I'm am in favor of listing it because I think the higher visibility means higher chance people will use the libraries within, and higher chance we find more folks that want to get involved with it.

However in it's current state I would say it's fair to say it's behind the other two bundles both in terms of libraries contained and active contribution. So I could understand leaving it off for the time being. My hope is that it continues to expand with DisplayIO widgets and other libraries that the community is interested in supporting rather than Adafruit directly.

idle owl
lone axle
#

IIRC It has some infrastructure things that need doing as well still like the automated bundle generation via adabot or similar tool.

idle owl
#

Sounds about right.

#

Hmm. Maybe shouldn't include it yet since it's not easy to find.

#

The others are on /libraries

#

I think I'll leave it out for now since I'm linking to the bundle download page for this section, and the Org bundle isn't there yet.

#

I will specify the Community Bundle is "individual community members" in preparation for adding the Org bundle being "maintained collectively by the community."

lone axle
#

Sounds good. The things in it ATM are probably a bit deeper than would be needed for primary audience of a /start page if I am imagining it correctly anyhow.

idle owl
#

Added it in a comment so it's there somewhere, but not public facing

digital shoreBOT
#
adafruit
Owner

adafruit#3230

Category Channels

8

Text Channels

61

Voice Channels

6

Members

33000

Roles

35

ID: 327254708534116352 | Server Created ·
slender iron
#

@idle owl thanks. I agree with foamyguy

idle owl
manic glacierBOT
manic glacierBOT
zealous current
#

hi everybody, I'm new to discord, so please bear with me 🙂 @slender iron kindly suggested that it would be easy to add os.utime() because FATFS has a corresponding function.

idle owl
zealous current
#

thanks @idle owl

#

so is this the right place to ask for (a lot of) help with the above task?

idle owl
#

Absolutely!

#

Ask away, but be patient, as folks aren't always around. Someone will help as soon as they can though!

stuck elbow
zealous current
#

@stuck elbow - yes, I found that but I'm a Python guy ignorant of c and its type definitions, pointers, etc. First of all, I guess for circuitpython it should be sufficient to implement os.utime(path, times=None), right?

slender iron
#

welcome @zealous current! the best place to start is to find a similar function that we do have implemented

zealous current
#

@slender iron that's what I'm trying to do. So I would start with STATIC mp_obj_t os_utime(mp_obj_t old_path_in...

#

now what do I do for an optional tuple of ints? (none of the other functions have something like that, afaics)

slender iron
#

kwargs are easiest to do with the MP_KEYWORD stuff

#

taking a step back. python objects are C structs under the hood

#

shared-bindings is all about validating those structs to make sure they are the right type and then converting them to the corresponding C types

manic glacierBOT
#
t6b
[adafruit/circuitpython] New comment on issue #5538: Sending Requests Failed on second network call in CircuitPython 7\.0\.0

@makermelissa This issue is just an issue with the example. I have traced through the code and discovered:

  1. There can only be one socket_pool per wifi.radio
  2. Initializing the radio twice isn't helping
  3. The fix can be done entirely in code.py using these libraries

My only minor gripe is that I have to load secrets inside my code. It would be great if there was a public accessor or a way to pass the secrets through MagTag into PortalBase and NetworkBase...

zealous current
#

@slender iron so does STATIC mp_obj_t os_utime(mp_obj_t path, mp_obj_tuple_t times) {... look right?

lone axle
manic glacierBOT
slender iron
zealous current
#

@slender iron thanks, I'll have to muse over that a bit. Close to midnight here - gotta go. I'll be back.

slender iron
#

thank you! this is the best place to get help

manic glacierBOT
#
rp10007
[adafruit/circuitpython] New comment on issue #5930: M4 Express enters safe mode

I don't know if this is part of the crash behavior, but on the countio side I have discovered an odd thing about the M4 port that isn't true elsewhere afaik: if I define a counter in a file that I then import (e.g. encoder = countio.Counter(board.MISO)), that counter doesn't work. It uniformly reports 0. If I deinit that counter in the repl and define a new counter on the same pin, it works fine. If I define the counter in code.py directly instead of defining it in in an imported file, it wor...

mental nexus
# lone axle <@!366059101203202052> I noticed in your 2022 thread one of the widgets you ment...

That definitely caught my eye last night in show and tell. Thanks for sharing. I’m also thinking about how to make a big scrollable text box, like with a long paragraph. With limited memory, main issue is you don’t want to store a huge bitmap. Last year I made something that just rendered what you see and maybe a line above and below. Then as you scrolled it rendered what needed to come into frame. It got way complicated way fast. Seems like there should be an easier way than what I tried.

lone axle
#

Ah, paragraph scrolling would be nice as well. The PlaylistDisplay widget in the same project could be molded a bit more into a paragraph text box I think. It's set up for specifically 3 lines atm though.

mental nexus
#

Also, I see you could use a text line selector widget to select tracks. Maybe that’s the same as a vertically scrollable text box + touch selection. I remember you made a hamburger selector. Maybe there’s a core building block you can make into several offshoot text widgets?

manic glacierBOT
#
rp10007
[adafruit/circuitpython] Issue opened: #5982 Countio on M4 Express doesn't work when counter is imported

CircuitPython version

Adafruit CircuitPython 7.1.0 on 2021-12-28; Adafruit ItsyBitsy M4 Express with samd51g19

Code/REPL

code.py (finally working version):
from BrocItsyHelper import *

#encoder.deinit()
encoder = countio.Counter(board.MISO)

motor1.throttle=1; time.sleep(3); motor1.throttle = None
print(encoder.count)

BrocItsyHelper.py (works with above code.py, failed code commented out)
# BrocItsy helpers

import time
import board
import p...
mental nexus
manic glacierBOT
mental nexus
blissful pollen
manic glacierBOT
mental nexus
lone axle
mental nexus
blissful pollen
shell bolt
#

Hello! To what extent, if any, does MicroPython's Zephyr support extend to CP?

manic glacierBOT
idle wharf
#

One of the reasons for that, CP has a common hardware layer across all ports, where in MP it is port specific and not common\the same for every port.

shell bolt
proven garnet
#

Rereading things from today, I didn't know there was another bundle besides the Adafruit and Community, how do I access the Org one?

#

And is it possible to contribute to it? Also the intended difference between them? Take for example the CSV module I wrote, circuitpython_csv, I would actually prefer that to widely examined by others and contributed to, but realistically I'm the only one doing so right now, since it's not like a collaboration. Is there a preferred bundle?

jaunty juniper
jaunty juniper
tulip sleet
# shell bolt Got it. I'm still learning but what confused me is that fact that each port incl...

See https://circuitpython.readthedocs.io/en/latest/docs/porting.html#architecture for how we separate the Python API from the port-specific implementations. There is a "common HAL" API called from the Python API side, and it's reimplemented for each port (or shared, if possible). ports/name_of_port/common-hal/.... On the port-specific side, we call that port's HAL, or sometimes use raw register manipulation, depending on what we need.

manic glacierBOT
#
jepler
[adafruit/circuitpython] New comment on pull request #5979: Enable mdns hostname setting

While it's possible we might use a different device with a TCP/IP stack that doesn't support mdns, it would be possible to have these properties always return None and error if a non-None value is ever assigned.

I could remove the "instance name" for now, it is more useful when your device has an "mdns service". In this case, the "friendly name" (instance name) can be shown in the service browser. For example, in the avahi-browse -a output line like

+ enp8s0 IPv4 OctoPrint instance...
proven garnet
manic glacierBOT
jaunty juniper
#

well the idea is to be a place for libraries that people collaborate on, but I tend to think that most libraries start with one person and others come later so I don't know how relevant that is

#

but the idea of the Circuitpython Organisation is to put more stuff in there rather than under the Adafruit name, so more people (that don't work for Adafruit) would be part of it and maybe create their libraries there, and then put them in that bundle ?

weary hill
#

Hello all! I created an open source command line tool called piku to help my CircuitPython development. Its patterned after npm and poetry. You can use it to create a project, download and add modules from the bundle, deploy to the device, and connect to serial via the command line. I'll try to post about it again. I hope it is useful to someone else, and I would love your feedback: https://github.com/mraleson/piku

GitHub
GitHub - mraleson/piku: A small command line utility for managing C...

A small command line utility for managing CircuitPython projects - GitHub - mraleson/piku: A small command line utility for managing CircuitPython projects

frosty root
#

Is this the right channel in which to post a question about stepper motor control using circuitpython?

blissful pollen
frosty root
#

ok thanks

manic glacierBOT
#
jepler
[adafruit/circuitpython] New comment on pull request #5979: Enable mdns hostname setting

@tannewt I clarified the documentation to state that if the networking stack doesn't support mdns then

  • the value of the property is always None
  • setting it to a non-None value will raise a ValueError

does this allay your concerns about a future networking stack that doesn't support mdns, or do you still want this in a separate module? If a separate module, would you care to suggest a name? Here are the first two alternatives that came to mind for me:

import mdns
import...
#
jepler
[adafruit/circuitpython] New comment on pull request #5976: Change ST drivers to use ST's driver repos as submodules

It's also possible to silence diagnostics strictly within source files and headers from a third party, meaning it may not be necessary to fork a repo just to make the ST code compile cleanly with our preferred CFLAGS.

To silence a warning inside a header controlled by a third party, modify the CFLAGS so that it is the "-isystem" flag that introduces the location where the header resides, rather than "-I" or "-iquote". For example, in the espressif port we have

espressif/Makefile:    ...
manic glacierBOT
manic glacierBOT
blissful pollen
#

Woo! I got a native built mpy module loading

proven garnet
# weary hill Hello all! I created an open source command line tool called piku to help my Cir...

This is awesome! I really wanted something like this! I'll definitely check this out and let you know. Is there any kind of git integration? My big problem is sometimes my local git repo and what's on the board don't match up, because I'll edit the code directly on the board and forget to copy/paste the changes back to my local repo. Having it all be ONE place is super helpful, so if there were any sort of way to use this with an existing git folder (or create a git folder within/as the piku folder) that would be incredible!

slender iron
#

@weary hill please email cpnews@adafruit.com and we'll include it in the circuitpython newsletter