circuitpython-dev

[manic glacier]

tannewt

[adafruit/circuitpython] New comment on issue #1239: External Flash Device Winbond W25Q16JV

Yup! I think so.

tannewt

[adafruit/circuitpython] Issue closed: #1239 External Flash Device Winbond W25Q16JV

[cedar beacon]

thanks @meager fog and @slender iron I've made a start here https://github.com/robmarkcole/circuitpython-on-home-assistant I'm assuming I can use MQTT over wifi with the esp boards and circuitpython?

[solar whale]

@cedar beacon you can use umqtt from the Micropython libs https://github.com/micropython/micropython-lib?files=1

[cedar beacon]

👌

[manic glacier]

dhalbert

[adafruit/circuitpython] New comment on issue #1326: Trellis M4 latest build causing issue with NeoPixel code

Here is a safe-mode build. I built this last night but then thought you probably had a clock issue. Unzip before using.
trellis_m4-SAFE-MODE.uf2.zip

[manic glacier]

zhouF96

[adafruit/circuitpython] New star added

alongruss

[adafruit/circuitpython] New star added

[manic glacier]

dynanets

[adafruit/circuitpython] Fork created: dynanets/circuitpython

[manic glacier]

notro

[adafruit/circuitpython] New comment on issue #1120: Extra \\r in Traceback lines

Indeed it did, thanks.

notro

[adafruit/circuitpython] Issue closed: #1120 Extra \\r in Traceback lines

[idle owl]

@cedar beacon To clarify the USB point, the thing the ESP doesn't have is USB mass storage. When you plug in an ATSAMD board, a USB drive appears called CIRCUITPY and allows you to drag and drop files, and edit them live on the board. The ESP doesn't support that. When we refer to it not having USB, it's not a reference to it physically having a USB connector, it's a reference to the lack of mass storage capabilities.

@slender iron I wasn't suggesting giving it more power for all the pixels, I simply remembered from discussion that something was going to be added to handle that particular failure better. The turn off command on reset is what I was thinking of, but I forgot what it specifically was.

@tulip sleet Thank you for responding to that post, that is what I was referring to when I asked Scott what we'd done to deal with that.

[cedar beacon]

Thanks for the clarification @idle owl

[inland tusk]

@slender iron tan I will not be there this week. Hope to see you next werk.

[slender iron]

ok thanks @inland tusk. talk with you next week!

<@&356864093652516868> Meeting in two minutes!

[gusty kiln]

i can hear y'all, but i might be text input only.

cool.

[tidal kiln]

@slender iron lurking this week. group hug.

[cunning crypt]

Boo. I am busy with paperwork for new job.

[idle owl]

@cunning crypt Congrats!

[slender iron]

@cunning crypt +1 for a new job

[meager fog]

here in text but will be in and out in meatings

congrats andon!

[slender iron]

k feel free to post hug reports and status updates early @meager fog I'll slot them in as we go

[cunning crypt]

Thanks guys, but gah. Paperwork.

I'll give my customary hug report to everyone since you are all awesome. Status I think I covered.

[meager fog]

thanks - big hug report to tannewt and hathach for a huge usb rewrite

[errant grail]

... in the workshop so no microphone today.

[marble hornet]

I'll be late to the meeting b/c school ends at 200

[meager fog]

brennen & kattni for character LCD hackin'

[marble hornet]

@slender iron My hug report: HUGE hug to the community. I looked down at my work today and there is no way I could have done this at my current level without python. Your work and standardization has made it so easy and enjoyable to code for controllers. Thank you all so so much. Happy Thanksgiving to all.

[gusty kiln]

pasting, one sec

@tidal kiln for picking up chained TLC5947 support after I dropped the ball.
@idle owl for taking on CharLCD library updates.

[errant grail]

Group hug to the team and community. The help and support are amazing. @idle owl’s character display work will be a real time-saver. Wildly anticipating @slender iron’s USB implementation! Also, I appreciate the exposure to advanced topics even though I don’t understand many – yet.

[marble hornet]

Here

[gusty kiln]

@idle owl congrats on getting that stuff done. seems like a big step. :)

[idle owl]

@gusty kiln thanks! 😊

[gusty kiln]

Some assistance last week to @idle owl on CharLCD library updates / testing.
Hacky install script for libgpiod on the Pi / Raspbian.
Working on libgpiod_pulsein, a small C shim for polling GPIO values on Linux systems.
Testing TLC5947 chaining.
Improving Blinka platform detection slightly.
Out Thursday & Friday.

[errant grail]

Deployed the CircuitPython and Trinket M0 holiday window lights this weekend, although they won’t be activated until after Thanksgiving. Each controller box has its own piezo that plays a fanfare whenever the lights turn on or when they automatically turn off after 6 hours.
Also tested the front panel and audio output features of the lunchbox music synth (CircuitPython 3.1.1, Feather M0 Express, custom purple PCB). Next steps are to finish testing the analog inputs and outputs, implement the MIDI stack, and tune the default DSP-G1 synth voices. https://www.youtube.com/watch?v=U0NItzSlINU&feature=youtu.be

[solar whale]

Gemma M0 and lis3dh “ embedded”

[fierce girder]

fyi - https://github.com/adafruit/Adafruit_CircuitPython_BusDevice/pull/22#issuecomment-435209002

[neat folio]

"unused pins are wasted pins" 😉

[drowsy fox]

https://i.imgur.com/FNnf8V2.png

[marble hornet]

I have one

[gusty kiln]

i was gonna say, reloadexception does that, right?

[slender iron]

https://circuitpython.readthedocs.io/en/3.x/shared-bindings/supervisor/__init__.html#supervisor.reload

[stuck elbow]

that is also useful to me, thanks!

nothing to report

[marble hornet]

bye

[neat folio]

now I want to make things on a lathe 😦

[errant grail]

Thanks!

[gusty kiln]

thanks y'all.

[errant grail]

Better to show and share!

[gusty kiln]

also off to get lunch. later all.

[onyx hinge]

late hug report to @C47D for fixing the docstring I wrote incorrectly!

I had started by intending to clone the linux-specific time.clock_gettime_ns which is where the "clock ID" came from

[slender iron]

<@&356864093652516868> Here is the recording from today's meeting: https://youtu.be/l1dRzurIXq0

[marble hornet]

thanks @slender iron

[slender iron]

And the diode.zone link: https://diode.zone/videos/watch/4e9cb207-1425-42a2-a154-6ba6627fb73f

[winged leaf]

Hi, I’m trying to figure out how to read from a file in the emulated storage.

Writing to a file is well documented https://learn.adafruit.com/data-logging-with-feather-and-circuitpython/circuitpython-code but I can’t seem to find anything about reading from a file.

Any ideas or help would be greatly appreciated, thanks a bunch.

[marble hornet]

@tulip sleet or @slender iron any pseudo code you have in mind that could be added to the issue?

[tulip sleet]

storage.persistent_ram[0] = 12

or something like that. Something like nvm but in ram: survives reloads but not hard resets.

[marble hornet]

tnx

i'm thinking this too:

storage.next_p_r_size(# of bytes)

[tulip sleet]

yeah - I forgot tht part

[marble hornet]

what about storing objects from python?

storage.p_r_obj(name_of_obj = 'obj') than storage.p_r() #returns list

[tulip sleet]

that's harder because the objects are not part of the Python heap, and potentially contain pointers to other objects. Instead, we provide a bytearray: you can use struct to serialize primitive objects (ints, floats, strings, etc.) or just store strings as bytes

[marble hornet]

rather a dict

[tulip sleet]

that's why nvm is just a bytearray also

[marble hornet]

??

wym

[manic glacier]

dopplet

[adafruit/circuitpython] New star added

[marble hornet]

ohhh 👌

[manic glacier]

codydg

[adafruit/circuitpython] Fork created: codydg/circuitpython

tannewt

[adafruit/circuitpython] New comment on issue #1340: nRF boards slow to enter REPL

@jerryneedell what terminal program are you using? Are you typing anything once connected?

@dhalbert Your trace doesn't have the linecoding packets until 12.5 seconds in and CircuitPython responds promptly. Can you try with something besides picocom? Does an atmel board do the same?

[idle owl]

.mpy strips out docstrings and comments, right?

[marble hornet]

from my understanding it leaves only executable lines of code

but! a string will stay, so multi line comments are executed, just like typing: "jgjhdjksa this is a string" into the repl

[raven canopy]

argh. totally forgot to put my meeting notes in last night. 😦

[idle owl]

(╯°□°）╯︵ ┻━┻

[manic glacier]

jerryneedell

[adafruit/circuitpython] New comment on issue #1340: nRF boards slow to enter REPL

I use screen. I connect then enter “enter” until it responds.

[marble hornet]

@idle owl 😂 😂 😂 😂 😂

did I answer your question well enough? (i'm trying to hone my explanation skills)

all feedback welcome 😊

[idle owl]

@marble hornet Presumably, yes. I appreciate your help. Removing all my docstrings etc did not resolve it. I had already found the fix but was trying to make sure there wasn't another option.

[marble hornet]

what's the difference between docstring and text?

[raven canopy]

@marble hornet the way i took your explanation was to mean that The result: wouldn't be "removed" in the following:

print('The result: %' % result)

which i think is correct. however, a docstring would be removed:

def Foo():
    """ Foo returns Bar
    """

[marble hornet]

ooooh 👌 thank you. does that work in classes ?

[raven canopy]

docstrings? yep.

# comments are also removed

[idle owl]

@raven canopy Thank you.

That was what I was referring to.

[manic glacier]

tannewt

[adafruit/circuitpython] New comment on issue #1340: nRF boards slow to enter REPL

Ok, I tried with my Arch box and it seemed fine. I'll try with my Ubuntu one. It could be auto-mount related.

[marble hornet]

@raven canopy would """" this is a multi line comment""" be removed if it was say the first line in main.py?

[raven canopy]

yep, should be.

[marble hornet]

could you give me some help w/ the mpy compiler? it's not working for me

please

wait

just figured it out

forgot about ./ 🤦

[raven canopy]

hehe. sorry, was elsewhere. and, i do that all the time. 😄

[marble hornet]

i have another question

okay two...

[raven canopy]

ghead

[marble hornet]

do you use macos?

/

[raven canopy]

nope. Win10 and Ubuntu.

[marble hornet]

do you know or have a script that mpys everything in a folder ?

oh

[raven canopy]

i don't. but, you could use circuitpython-build-tools to make a bundle just like Adabot does. also, run ./mpy-cross -h...there may be a parameter that will do a folder/batch.

actually, just looked at the source. none such option. 😦

[marble hornet]

no ah well

i tried -h

does this mean anything to you? apply bytecode optimizations of level N

[raven canopy]

not really. i would guess its similar to -O<n> for GCC, but i don't know the details on how it would optimize mpy...

[raven canopy]

😆

[marble hornet]

trying to understand things the first time

[solar whale]

@winged leaf what are you reading from -- is ithe Fassh system on a CircuitPython Board or an SDCard attached to a CP board. For a file on the FLash drive: ```>>> f = open("/boot_out.txt","r")

print(f.readline())
Adafruit CircuitPython 4.0.0-alpha.3 on 2018-11-15; Adafruit Feather M0 Express with samd21g18

[tidal kiln]

ah! those ani gifs scared me. was in another tab.

[marble hornet]

ani?

[tidal kiln]

animated

[slender iron]

@gentle bronze are you around?

when you are can you merge https://github.com/hathach/tinyusb/pull/10 so that I can update circuitpython to it?

I think you fixed a couple bugs I had

[red dock]

@slender iron So I've been playing with displays (I now have one of 3 LCD displays i'm playing with. How hard would it be to port the FT6206 Driver for Capacitive Touch to Circuit Python?

[short canyon]

Does anyone know whether the mu-editor works with the Feather Huzzah ESP8266?
(I asked at the mu gitter discussion, but they didn't know there.)

[red dock]

@short canyon I'm pretty sure it does. I don't have a ESP8266, but it's one of the Feathers I was looking at.

[manic glacier]

dhalbert

[adafruit/circuitpython] New comment on issue #1340: nRF boards slow to enter REPL

@tannewt: I started Beagle datacenter, then plugged in the board, counted to 5 or 6 seconds to make sure it was done enumerating, then connected via picocom, then waited for the REPL prompts to appear. I didn't type anything; the prompts just appear spontaneously.

[gentle bronze]

@slender iron AH ok, just merged

but I haven't tested with samd yet. still adding bsp support for it https://github.com/hathach/tinyusb/blob/develop/hw/bsp/metro_m4_express/board_metro_m4_express.c#L52

[ruby lake]

hm, a lot of crosstalk in the trellis m3 audio output

~m4

[gentle bronze]

@slender iron I started to get some usb transfer with samd51, I am doing more tests, maybe you should wait a bit longer , I am doing more testing with it

if you have segger embedded studio installed (with same51 package), you could run and debug the minimal usb example here (nrf52 also) https://github.com/hathach/tinyusb/tree/devlocal/examples/device/device_composite/ses

[cedar beacon]

Giving a short presentation on circuitpython at work python meetup today https://gitpitch.com/robmarkcole/circuitpython-on-home-assistant/master?grs=github&t=night

[winged leaf]

Thanks @solar whale it's the internal file system I want to read from. Cheers!

[ebon horizon]

@cedar beacon I think it looks really nice. But, visually, you never show a link related to CircuitPython. The only "visual" links are micropython and mu. (the others require clicking on something, which doesn't always translate to a presentation.)

[cedar beacon]

thanks for the feedback @ebon horizon

[winged leaf]

I was hoping there's a way to turn off the 'CIRCUITPY' mass storage? Is this something that's doable, mounting/unmounting the drive from within the code?

[marble hornet]

@tulip sleet or @slender iron could the size of the 'scratchpad' change between soft reboots or after hard ones ? if cp is soft rebooting won't the cp code stay in the same spot? how could the pad get bigger? (writing issue now)

[tulip sleet]

@winged leaf we are working on run-time choices of what to expose via USB. what is your use case?

@marble hornet the reload clears the heap, but I think the persistent ram size may not be able to change without a hard reset, because there are non-heap allocations used for usb, file system, etc.

[winged leaf]

@tulip sleet I want to create a keyboard, using the HID libraries, the mass storage isn't really needed so I'd like to be able to disable it, but it would be cool if it could be enabled/disabled with certain key combinations.

[tulip sleet]

The way usb works, we’d have to reset the usb connection to change what’s presented. We plan to make specifying which devices show up on usb be done in boot.py, which runs before usb is set up.

[winged leaf]

That would be awesome, @tulip sleet appreciate your work! Do you reckon that is a feature that is very far away?

[upbeat plover]

on the feather M4, NVM address 0 gets erased when you change CP version but at address like NVM[200] it stays there between builds... guessing the size of the flash is smaller then the total ROM

[upbeat plover]

Can you use FRAM with CP?

would i just use "busio" "I2C"?

[tidal kiln]

thusly?
https://github.com/adafruit/Adafruit_CircuitPython_FRAM
theres a SPI and I2C flavor, currently the lib only has I2C support, SPI support is PR'd and in review

[bleak tiger]

Hi, I can't find away to release the coils on a stepper in motor_featherwing. Could this be added to the library as a function? Similar to release() in Adafruit_MotorShield

[tidal kiln]

@bleak tiger there are a couple of recent issues added to the underlying library that may be related. take a look:
https://github.com/adafruit/Adafruit_CircuitPython_Motor/issues
and feel free to add another one if those don't cover it

[bleak tiger]

Will do, thanks!

[marble hornet]

Thanks for all your help @tulip sleet ☺️

[slender iron]

@winged leaf you should chat with @sudden coral and @bronze shadow they are working on keyboard firmware and we had a similar conversation. Some discussion is here: https://github.com/KMKfw/kmk_firmware/issues/52

[prime flower]

just when I thought I pushed thru the encryption parts of tinylora, the MIC has its own key generation routine 🙃

This library is starting to get really long..I feel I should start breaking the encryption routines out into separate files/classes

[manic glacier]

0m3rta13

[adafruit/circuitpython] New star added

[prime flower]

@slender iron would an implementation where I have a tinylora_encryption.py file with classes for aes and mic work? or should I break the two routines into separate files?

[slender iron]

@prime flower will you always use both?

[idle owl]

@tidal kiln I finally tested LSM9DS0. I requested two small changes - I think it's copypasta. Otherwise, it runs great with those changes!

[prime flower]

sending a packet will always use both. The MIC uses the aes routines.

[slender iron]

then its up to you

if they were separately used then I

[tidal kiln]

@idle owl probably. that whole driver was copy-pasta.

[slender iron]

'd say they need to be split

[idle owl]

@tidal kiln Yah I figured as much.

[prime flower]

Okay, thank you

[idle owl]

@tidal kiln Do you know off the top of your head of an example of a library that has I2C or SPI split into a different file in the driver package? I know we have some, but I don't know off the top of my head.

[manic glacier]

TG-Techie

[adafruit/circuitpython] Issue opened: #1346 Scratch Ram or Something

This topic was discussed in CP Weekly November 19th, 2018.
At about time 35:50 here:
https://youtu.be/l1dRzurIXq0?t=2150

The discussion stemmed from a desire to prevent memory errors when switching between many modules. The idea is to be able to pass information between soft reloads and then have some user-defined main.py act accordingly from the saved data. The group came to an idea of a 'ram scratchpad', this would be like nvm except made of ram.

the difference form nvm is that...

[tidal kiln]

@idle owl https://github.com/adafruit/Adafruit_CircuitPython_PN532/tree/master/adafruit_pn532

[idle owl]

Thank you

[tidal kiln]

@idle owl you tested LSM9DS0 on SPI?

[idle owl]

@tidal kiln Yah...

[marble hornet]

how do you add tags to a git issue?

[idle owl]

The variable names don't match. PyCharm caught it before I even tested it.

[tidal kiln]

and it worked?

@idle owl oh. i see. you actually changed code from the PR version to get it to work?

[idle owl]

Yes.

[tidal kiln]

yah. that totes copy pasta.

[idle owl]

Figured 😃

[tidal kiln]

for some reason names changed between DS0 and DS1

[idle owl]

¯_(ツ)_/¯

[tidal kiln]

it should'nt have even run with that PR code

[idle owl]

it didn't without my changes

[tidal kiln]

should've got a "no such member" or some such error

[idle owl]

Yep

I already disconnected it, so I can't paste the error. But yeah it failed initially. I changed the things to match and it ran.

[tidal kiln]

ok. makes sense. good enough. no need to recreate.

yay for HW testing. good catch.

[idle owl]

Thanks 😃

[tidal kiln]

@idle owl update pushed

[marble hornet]

issue created

https://github.com/adafruit/circuitpython/issues/1346

[idle owl]

@tidal kiln hmm... now it failed. Until I swapped the csag and csm wires from what the example has. I noticed when I pasted the new version of the driver into the file, gcs and xmcs swapped. self._gyro_device = spi_device.SPIDevice(spi, gcs, baudrate=200000, phase=1, polarity=1) self._xm_device = spi_device.SPIDevice(spi, xmcs, baudrate=200000, phase=1, polarity=1)

that's what it is now, but before they were opposite.

so if that's correct, the example is now backwards?

[tidal kiln]

i swapped the two lines to match the order they are set up in the I2C class

[idle owl]

hmm. And the example code you got from the other version?

[tidal kiln]

ah...i see what you're looking at. maybe more copy pasta. let me look at that also.

[idle owl]

It works if I have csag connected to D6 and csm connected to D5.

wait... it's not working right. Looking at the output, all of the tuples match themselves, as in (x, y, z) for each of them are the same.

and temp is off...... I'm not freezing to death here.

Magnetometer (gauss): (-0.185,-0.185,-0.185)
Gyroscope (degrees/sec): (-20.247,-20.247,-20.247)
Temperature: -39.250C```

[tidal kiln]

blah! the CS order is different between the two also.

LSM9DS0:
def __init__(self, spi, xmcs, gcs):
xmcs = mag
gcs = gyro

LSM9DS1:
def __init__(self, spi, xgcs, mcs):
xgcs = gyro
mcs = mag

[idle owl]

oi.

[tidal kiln]

so maybe it's not as simple as originally thought

[idle owl]

I've learned nothing is ever simple.

at least not for me.

[tidal kiln]

esp.. when you make an assumption that it is

[idle owl]

yep.... rug == ripped out.

[tidal kiln]

would you be up for taking over the fix?

[idle owl]

I guess? I don't know if I can fix it though. All of the readings being the same is presumably something else going on in the code and I don't understand bitwise operations or registers or bit shifting or whatever any of that involves.

And it's going to be a bit. I'm still mired in charllcd.

[tidal kiln]

i don't think this is a high priority driver, so taking the time isn't an issue

could be a good way to level up, you could use the LSM9DS1 as a ref

[idle owl]

Except apparently I can't use it as a ref, it's not the same.

[tidal kiln]

yep. but the general bit-foo stuff that's there should help as a notional guide.

based on what you are seeing - temperature way off, other values not updating - something seems to be up and it's best resolved by having the HW available for testing

[idle owl]

Yah I get it.

[tidal kiln]

i could try and scan the code and see if it's something obvious, but having the HW really helps

[manic glacier]

bwshockley

[adafruit/circuitpython] New comment on pull request #1344: Add Mini SAM M4 board\.

I am not excellent with git or github, but I think I removed the changes I'd made to devices.h

[idle owl]

Nah, it's fine. I'm commenting on your PR.

[tidal kiln]

i also hooked up the saleae to verify some stuff on the DS1

[sly wagon]

Finally got both sides connected for my #blackpanthercosplay haha, but now I need to work on couple small new problems. I've got a question re: power source and splitting power between a usb battery pack between 2 circuit express playgrounds ... what's the correct room to ask questions about that, friends?

Idk why the preview won’t show ha. I’ll upload to Twitter and post here

[idle owl]

@sly wagon You might get more help in #help-with-projects regarding your power issue.

[sly wagon]

Maybe projecthelp is where I ask that stuff ?

Kk thank you ! @idle owl

[manic glacier]

tannewt

[adafruit/circuitpython] Pull request opened: #1347 Update for Grand Central Rev B

[manic glacier]

deanm1278

[circuitpython:dm\-mixer] 1 new commit

1bf6c58 DM: mixer voice stuff - deanm1278

deanm1278

[adafruit/circuitpython] New comment on issue #1306: request: volume for mixing and/or playback

ok @tannewt have a look at the new interface in this commit:
https://github.com/adafruit/circuitpython/commit/1bf6c588e7e1ea11c865c33f660ec126f752b1dc

It is not finished but I would like to get approval on the general structure before I continue

import array
import math
import time
import board
import busio
import audioio
import struct

VOICES = ["voice01.wav", "voice02.wav", "voice03.wav", "voice04.wav"]

def parse_wav(filename, *, scale = 1.0):
    print("Reading file...

[tulip sleet]

@upbeat plover the location of the nvm block may have changed somehow between versions (but I thought not). If [0] is changing but [200] is not, can you file an issue? We might have an off-by-one (or four). Maybe also look at [1], [2], [3], and [4].

[cedar beacon]

Hi is anyone aware of a Flask/Django web app for managing micropython/circuitpython boards? Something that allows flashing firmware, uploading files etc via a web browser. If there were a Docker image available that would be the icing on the cake. Thanks

[stuck elbow]

no

[pastel panther]

nope

[marble hornet]

could blinka be considered a shim?

[tulip sleet]

Sure, I would say that.

it adapts one interface to another (CPy hw to gpiod hw)

[marble hornet]

gpiod? what is the d?

[runic hornet]

"daemon?"

[pastel panther]

probably

[cedar beacon]

I've flashed circuitpython 3.1.1. onto a raw esp8266 and attempt to connect with screen /dev/ttyUSB0 115200 (works for my huzzah) but I just see first a blank screen, then on resetting the board I see jibberish {ll��|�$�|��l�#|����r�b�c�p~�o�$on��#cp��c$r$p�n��lbo�|�$$�#��on�l��$�on�{lor���N��p�n�#o�|#��on��l�onl or���N�� �n$���b�N��No��{lp�o�bn�|b��no�l�$�o$nr�����ܜ$r��n��� {��Nd�|���r��o�#�obp

Any advice?

perhaps I should flash board with lower baudrate?

[manic glacier]

gbafana25

[adafruit/circuitpython] New star added

[prime flower]

@cedar beacon are you using AMPY?

[cedar beacon]

not for flashing firmware

I followed https://learn.adafruit.com/welcome-to-circuitpython/circuitpython-for-esp8266#get-esp8266-ready-for-bootloading-13-6

[prime flower]

doing an esptool flash at 115200 baud?

(gibberish/junk terminal output is usually an incorrect baudrate, I use 115200 for my ESP8266 running )

[cedar beacon]

yes

also tried 460800

no errors

esptool.py v2.5.1
Serial port /dev/ttyUSB0
Connecting....
Detecting chip type... ESP8266
Chip is ESP8266EX
Features: WiFi
MAC: 5c:cf:7f:fd:80:7a
Uploading stub...
Running stub...
Stub running...
Changing baud rate to 460800
Changed.
Configuring flash size...
Auto-detected Flash size: 4MB
Flash params set to 0x0040
Compressed 606212 bytes to 395009...
Wrote 606212 bytes (395009 compressed) at 0x00000000 in 9.9 seconds (effective 490.3 kbit/s)...
Hash of data verified.

Leaving...
Hard resetting via RTS pin... ```

[manic glacier]

tannewt

[adafruit/circuitpython] New comment on issue #1342: samd: REPL disconnect problems after tinyusb PR

FYI, I'm waiting for this PR: https://github.com/hathach/tinyusb/pull/11

Once its in, I'll update CircuitPython to it.

[solar whale]

@cedar beacon a blast of gibberish is normal at start, but it should settle to normal REPL. Try control C

Flash is at 460800 REPL at 115200

[cedar beacon]

@solar whale still get gibberish

[solar whale]

Sorry, can’t get to my HW to check anything. The gibberish is a normal output of the bootloader but for me it recovers.

Did you do a flash erase before loading. Needed if going from Arduino

esptool erase_flash

@cedar beacon if I recall correctly I had similar issues if I did not erase first

[tulip sleet]

@cedar beacon When I upload I do this (via some aliases):

esptool.py erase_flash
esptool.py --baud 460800 write_flash 0 something.bin

[pastel panther]

@slender iron I started working on a build that supports my st7735 display via displayio didn't have a saleae to test/debug it at the con but I'll let you know how it goes.

[slender iron]

nice!

[pastel panther]

As as starting point I aped the setup for the hallowing in its board.c and updated pins.c to have a DISPLAY; anything else I should keep in mind?

[tawny creek]

Hi All! Trying to get this working on CP4 Alpha 2 (maybe ill try alpha 3)
https://raw.githubusercontent.com/adafruit/Adafruit_CircuitPython_IS31FL3731/master/examples/text.py

and getting this error

  File "code.py", line 39, in <module>
  File "code.py", line 14, in <module>
AttributeError: 'module' object has no attribute 'MONO_VLSB'

[pastel panther]

I don't think framebuf is working in 4.x

[slender iron]

I turned it off

[pastel panther]

that would keep it from working 😉

[tawny creek]

ah I have included framebuf.mpy so i was able to get some of the examples working with it

would this matter if it's 'turned off' ?

[tulip sleet]

the "turned off" version is a native (C) implementation. framebuf.mpy is a Python version that is slower and has less functionality

like, no text support

[slender iron]

yup, sorry for not being clear

I turned off the C version

[tawny creek]

ahh gotcha, makes sense for the incoming display.io stuff

@tulip sleet what's the quickest way for implementing text support?

[tulip sleet]

"quick" unfortunately is not the operative word here (nor "easy"). Here's an old guide about doing text, with some obsolete info: https://learn.adafruit.com/micropython-displays-drawing-text. Doing it right is part of displayio.

[tawny creek]

@tulip sleet actually this looks very familiar to me! I think I got this to work with SSD1306, thanks!

[tulip sleet]

There is some work described here that is going on as well: https://forums.adafruit.com/viewtopic.php?f=60&t=142003&hilit=fonts

[tawny creek]

@tulip sleet got it working :D!

thank you!

[tulip sleet]

great!!

[tidal kiln]

@tulip sleet oh guru of all things window oddness. i beseech thee:
https://forums.adafruit.com/viewtopic.php?f=8&t=143708#p710466

[tulip sleet]

@tidal kiln assuming windows 7 works, maybe they should update the bootloader on Windows 7, and then try again. https://learn.adafruit.com/adafruit-neotrellis-m4/update-bootloader. (The bootloader should be updated in any case.) Then try on Windows 10 again. I would also have them disable Norton Security Suite temporarily to see if that changes the Win 10 situation. And is Norton running only on Win 10?

i sort of doubt the bootloader, unprotected or not, is actually affecting Windows 10. More likely it's Norton or something like that.

[tidal kiln]

norton will be selective like that? is it M0 vs. M4 or something else?

also - thanks. didn't know some of the trellis's got out without bootprot....hmmm....may explain other things....

i'll have them do that regardless as first step

[idle owl]

@tidal kiln contrast is the new show().

[tulip sleet]

I don't really know why it would be different M0 vs M4. Right, that's odd. Could be some USB timing thing. Maybe also get port info: USB2, USB3, hub, USB-C??

[tidal kiln]

@idle owl oh no. but my life is based on the current API. now what am i going to do?

[idle owl]

@tidal kiln you'll call show() and all you'll get is boxes in place of your text.

[tidal kiln]

████ ██ ██████

[idle owl]

It's begun!

[pastel panther]

https://www.youtube.com/watch?v=2U_OkCsexjw

also https://www.youtube.com/watch?v=3qYbVQu7YAQ

[tawny creek]

Hi guys! :3 trying to build CP4 alpha 3, and getting this:
No rule to make target `lib/tinyusb/src/portable/microchip/samd21/dcd.c', needed by `build-espress0/genhdr/qstr.i.last'. Stop.

other than
git submodule update --init --recursive
is there other things I need to do?

[pastel panther]

that should cover it but let us know if it complains about something else after doing so

[tawny creek]

@pastel panther the complaint that shows up is ```No rule to make target lib/tinyusb/src/portable/microchip/samd21/dcd.c', needed by build-espress0/genhdr/qstr.i.last'. Stop.

[pastel panther]

try doing a make <BOARD STUFF> clean if you haven't

then try again

[tawny creek]

same issue 😦 .. i've cloned a fresh one and that one seems to work

[pastel panther]

I think it might be that the submodule update doesn't by default clean up submodules that are no longer needed

tinyusb isn't in 3.x so it would make sense that it was complaining about it if there were leftovers

[tawny creek]

Hm, it was from 4 alpha 2

[pastel panther]

that might also be pre tinyusb

it was a pretty recent change

[tawny creek]

thanks @pastel panther , Imma just work on the fresh clones -- easy enough to port changes over

[pastel panther]

try deleting the lib/tinyusb directory

or not

[slender iron]

git submodule sync

that updates things when the submodules remote changes

[tawny creek]

on alpha 2 I used a W25Q16JV flash and that worked fine, but now getting this:

                 from ../../supervisor/shared/external_flash/external_flash.c:31:
../../supervisor/shared/external_flash/devices.h:226:6: error: 'external_flash_device {aka const struct <anonymous>}' has no member named 'has_quad_enable'
     .has_quad_enable = true, \
      ^
<command-line>:0:57: note: in expansion of macro 'W25Q16JV_IM'
../../supervisor/shared/external_flash/external_flash.c:48:78: note: in expansion of macro 'EXTERNAL_FLASH_DEVICES'
 const external_flash_device possible_devices[EXTERNAL_FLASH_DEVICE_COUNT] = {EXTERNAL_FLASH_DEVICES};```

trying to get the spi flash recognised now

ah ok @slender iron , will try that too

[slender iron]

I changed the quad enable a bit because I added support for different bit positions of it

[tawny creek]

I'm not sure which flavour of the chip I have, as Circuitpy isnt mounting

what does a steady solid blue status mean?

[slender iron]

blue is boot.py

iirc

have a jlink?

[tawny creek]

yup

the definition that worked for 4.2 and 3.x for the flash was

// Datasheet: https://www.winbond.com/resource-files/w25q16jv%20spi%20revf%2005092017.pdf
#define W25Q16JV {\
    .total_size = (1 << 21), /* 2 MiB */ \
    .start_up_time_us = 5000, \
    .manufacturer_id = 0xef, \
    .memory_type = 0x40, \
    .capacity = 0x15, \
    .max_clock_speed_mhz = 133, \
    .has_sector_protection = false, \
    .supports_fast_read = true, \
    .supports_qspi = true, \
    .has_quad_enable = true, \
    .supports_qspi_writes = true, \
}

[pastel panther]

there isn't W25Q16JV anymore; it's been split into two types

[tawny creek]

yeah, so the first kind (of W25Q16JV boots fine.. but drive doesnt mount

second kind boots but blinks white for a lil, then solid blue after

doesn't mount a drive on the second one either

actually, had to comment out .has_quad_enable to build

[cedar beacon]

@tulip sleet @solar whale @prime flower ok solved the flashing of the raw esp, had to add -fm dio and use --baud 460800, should this be updated on the docs?

[timber mango]

Is there a list with all the Keycodes from the Keycode library? I want to type "/", "." and ":" but I don't know the keycode

[solar whale]

@timber mango is this what you are looking for? https://github.com/adafruit/Adafruit_CircuitPython_HID/blob/master/adafruit_hid/keycode.py

[timber mango]

seems like it, thanks a lot!

[timber mango]

hi all,
iam currently trying to write a library for an advanced TI LED-Dvier (TLC5957 48ch 16bit)

i want to use the busio.SPI to allow for fast data output.
but i need to have a latch pulse synced with the last spi clock cycles.

so my idea was to write out most of the data per busio.SPI and the remaining data with the synced latch as bitbanging DigitalInOut

now i discovered that the pins are locked against multiple uses..
so it seems i need to deinit SPI to get access to the DigitalInOut and also the other way round..

but this means creating new Objects every time - and that sounds not fast...

are there any other ways to achieve this?

[timber mango]

ok after a quick test it seems that the bitbangio.SPI class is not locking the pins...
so will try with this one..

[marble hornet]

if one's using blinka would they have all the ram not used by the os and pyhton ?

[tulip sleet]

@marble hornet you mean on an RPi? You're just using all the memory Python can get, which is a lot, and it's a real Linux system, so you're not limited by the physical RAM (it's virtual memory).

@cedar beacon Great! Could you post the exact command lines you used, and which Learn Guide seems to be incomplete? Thanks.

[marble hornet]

i've just been looking for somehting that can run cp and has 1MB of more of ram. even if it means i have to try(big uncertainty) to port it. if rpi zero or compute modules runs cp 🤷

[tulip sleet]

They don't run the CircuitPython firmware, BUT, the blinka library adapts the CircuitPython devices modules (busio.SPI, etc.) to the native device support on Raspbian (you may need to install gpiod - I can't remmber - it's in the Learn Guide). So you are running "regular" Python (the regular CPython implementation of Python), and using CircuitPython-based libraries.

[marble hornet]

ah, okay I think I understand your distinction. should I have said i've been looking for something with >1MB that support cp modules?

[tulip sleet]

yeah, that's more precise. I think the Learn Guide should be clearer about what "CircuitPython" means. So yes, an RPi something or (soon) a BeagleBone Black works because adafruit-blinka works on those boards. So you'll have 512MB instead of just >1MB 😃

[marble hornet]

🤤

is the beagle bone info up on the learn guides? rather do you know what color they will use?

[timber mango]

regarding beagle-bone - not really- i read through the guides some days ago...

[tulip sleet]

the library is being developed, not done yet, stuff going on here: https://github.com/adafruit/Adafruit_CircuitPython_BusDevice/pull/22 to get it to work. See PR's in adafruit blinka: https://github.com/adafruit/Adafruit_Blinka/pulls?utf8=✓&q=is%3Apr

[timber mango]

i would love if adafruit-blinka could work as 'hardware emulator' so i can test the code on my desktop computer...
or if the qemu port would work in some way like a emulator with hw...
but i think this are dreams fare away 😉 (as i also don't have the time to try to get it running...)

[tulip sleet]

@timber mango since the main idea is to control external devices in various ways, I'd think you'd also need emulators for the external devices (e.g. some I2C device). Or are you thinking of some other way of doing this?

there's some of this idea in the simulators that are in MakeCode, but the set of devices is limited to, say, the on-board devices on the CPX.

[marble hornet]

thank you Mr. H,

[tulip sleet]

sure, yw

[timber mango]

@tulip sleet the emulators for the external devices would be great - this way you could really test your libraries in automated ways...
also it can not guarantee that you catch all corner-cases of the chips..
but for the start some 'we let the code run and do not crash' would be cool.. and for all output things an simple 'show data' would be helpful..

ok - i see as i get into this it gets complicated very fast 😉

[tulip sleet]

it's a great aspiration! I mean ideally mfrs would provide simulators for their sensor chips, etc. But too often they are flaky in weird timing ways, etc.

[timber mango]

i imagined a more basic way - so just the *io libraries would have some way to show what they normally would send out - and some way to simulate the input data...
and this scriptable - with that you could write tests for the libraries...

did you notice my question from earlier regarding HW SPI?

[tulip sleet]

no, i'll go back and look

can't find it, can you provide channel/time/timezone?

[timber mango]

ill copy it here:

s-light 13:35 Uhr
hi all,
iam currently trying to write a library for an advanced TI LED-Dvier (TLC5957 48ch 16bit)
i want to use the busio.SPI to allow for fast data output.
but i need to have a latch pulse synced with the last spi clock cycles.
so my idea was to write out most of the data per busio.SPI and the remaining data with the synced latch as bitbanging DigitalInOut
now i discovered that the pins are locked against multiple uses..
so it seems i need to deinit SPI to get access to the DigitalInOut and also the other way round..
but this means creating new Objects every time - and that sounds not fast...
are there any other ways to achieve this?
s-light 14:08 Uhr
ok after a quick test it seems that the bitbangio.SPI class is not locking the pins... so will try with this one..

(it was on this channel 😃 )

[tulip sleet]

hmm, the pins are claimed, so I'm not sure why it works for you. That chip sounds like a pain 😃

[timber mango]

😉 its a display driver - originally targetet at big mutliplexed led-walls

i like to play with these chips...

[tulip sleet]

but how did they expect the data to be sent in the first place: bit banging?

did you see this? https://github.com/adafruit/Adafruit_CircuitPython_TLC5947

[timber mango]

good question - i think they target is more of a fpga or something like this...
in the userguide (www.ti.com/lit/ug/slvuaf0/slvuaf0.pdf#page=4&zoom=160,-94,702a)
the controller generates all this fast things - so its more of a bare metal or high speed thing you normally need...

[marble hornet]

Dan, can I pick your brain?

or do you prefer Mr.H?

[tulip sleet]

Dan is fine!

[timber mango]

i did not spot the TLC5947 lib. thanks for the Link! i will have a look!
(i only used the TLC59711 lib as a reference...)

[tulip sleet]

also we have an arduino lib: https://github.com/adafruit/Adafruit_TLC5947

[marble hornet]

I'm thinking about having two cp-devices on my board, any thoughts of how to communicate between them?

[tulip sleet]

UART is simplest, I'd say.

There's an I2CSlave impl, but it's not compiled into most of the builds

[marble hornet]

the tx and rx pins?

[tulip sleet]

yeah, TX->RX and RX->TX between the two boards. People have done this all the time with Arduino in the past: there are probably a lot of examples of that.

[timber mango]

@marble hornet https://circuitpython.readthedocs.io/en/latest/shared-bindings/busio/UART.html

[marble hornet]

thanks

[tulip sleet]

https://learn.adafruit.com/circuit-playground-express-serial-communications

[timber mango]

@tulip sleet i had a look at the 5947. you can write all bytes and than latch them 😉 that is easy
my chip has an 48bit (6byte) shift register
and the latch is coded:
the chip counts the clock cycles while the latch line is high -
this is used then for different functions... (writing the data as Grayscale or as function commands and others...)
i will experiment a little more with the bitbangio..

[tulip sleet]

does the latch have to be synchronized with the end of the SPI, or could it be brought low after the SPI is finished?

[marble hornet]

is buf a list?

[tulip sleet]

you can use repr(your_list) on one side and eval() on the other side. Or if it's a list of numbers, you could use struct.pack and .unpack to format it into a byte stream

[timber mango]

the latch has to be synced with the spi clock cycles at the end of the data package

[tulip sleet]

ugh

[marble hornet]

repr... googling 1mo please

[tulip sleet]

you could use SPI for most, and bitbang the last byte

[marble hornet]

okay...

[tulip sleet]

maybe PulseOut would be helpful??

TG - what is it a list of?

[timber mango]

repr is similar to print -
but repr = representation - should print a python representation of a object that is 'python code' (if i remember correctly)

the idea to bitbang the last byte is what i have tried to implement..

[marble hornet]

i'm saving string

[tulip sleet]

a list of strings?

[marble hornet]

nope

[timber mango]

but the problem is - for this to work i would need to deinit the SPI object and create the digitalIO objects for all pins -
and after bitbanging the thing out deinit the pins and create a new SPI object...
and i think that this deinit and newly creating of objects could get very slow...
(i have to do this 16 times to write all data out..)

[marble hornet]

i'll check the site

[tulip sleet]

@timber mango use OR gates to drive the data/clock lines from two pins 😃

seriously maybe open drain outputs and a pull up or down?

wired-or

[timber mango]

hm could be an option - i will keep that in mind.

currently i have it running with the bitbangio.SPI class
(https://circuitpython.readthedocs.io/en/latest/shared-bindings/bitbangio/SPI.html)
this one does not lock the pins - and i can reuse them as digitalio 😃

[tulip sleet]

what chip is this that requires a synchronous latch?

not actually TLC5947?

[timber mango]

no its TLC5957 (the second last digit is 5)

http://www.ti.com/lit/ug/slvuaf0/slvuaf0.pdf#page=15&zoom=160,0,702

its a 48ch 16bit chip

the bitbangio clock iam generating is ~8.4kHz - so not bad for bitbanging 😉

[tulip sleet]

i'm looking at the datasheet, don't see that the latch has to be sync'd with the last clock pulse, unless it's this sentence: "5. During the same period of step4, GS data for next line should be written into GS data latch. Using LATGS
command for the last group of 48bit GS data loading."

but you've tried a delayed latch and it doesn't work?

[timber mango]

the problem is i have to generate 1 up to 15 clock pulses during the latch high time..

and with every clock the data gets shifted...

[tulip sleet]

You've seen this, I assume? http://www.ti.com/lit/ug/slvuaf0/slvuaf0.pdf

i see the multiple latching in like figure 11

[marble hornet]

thank you

[timber mango]

yes - its the main 'datasheet' as in the real sheet there are no commands given..

[tulip sleet]

and "traditional" mode is no better?

[timber mango]

no traditional mode is
send 48bit per chip to all chips -
every 48bits are representing one RGB output pixel

[tulip sleet]

ugh, ok, good luck, sri you have to deal with this. I have to eat bkfst, will be afk for a while

[timber mango]

the 'pocker' mode is really wired:
send 1bit for every output channel... so you can send only 9bit...

thanks for your help

Enjoy your meal

[timber mango]

is wemos d1 mini a good board to get started with micropython?

[prime flower]

maybe micropython? I've seen them mainly used for ESP development

[timber mango]

yeah i just discovered micropython and i have these wemos boards lying around that i could experiment on

so ill have to lear at least the basics. i'm wonding though if i should get something else to start with.

if you have it already just try it?!

i have started my CircuitPython journey with the ItsyBitsy M4 Express some days ago..

ok 😃

lol i bought a couple of those d1 mini to make the deauthor thing, didn't even realize it wasn't using the arduino language

[manic glacier]

sommersoft

[adafruit/circuitpython] Pull request opened: #1348 Fix Driver RTD Links

This fixes the following occurrence for Adabot checks:

CircuitPython drivers page missing driver - 9
  * https://github.com/adafruit/Adafruit_CircuitPython_CPython
  * https://github.com/adafruit/Adafruit_CircuitPython_HTU21D
  * https://github.com/adafruit/Adafruit_CircuitPython_MPRLS
  * https://github.com/adafruit/Adafruit_CircuitPython_L3GD20
  * https://github.com/adafruit/Adafruit_CircuitPython_NeoTrellis
  * https://github.com/adafruit/Adafruit_CircuitPython_LIDARLite
  * ...

[tawny creek]

I'm trying to build the latest cp for a custom board, here are my board definitions from 3 to 4.2 that worked followed by what i'm trying to use for 4.3

https://gist.github.com/eighthree/2181f94f80a0aaea46244b540e73725e

the 4.3 ones end up in a solid blue status light, how do I use a jtag to debug this or are the board definitions incorrect for 4.3?

[exotic pumice]

is line 56 of the first file right? I heard they split up that flash chip

[tawny creek]

Yeah, I've tried using both

[exotic pumice]

ok

that's all I got

[cedar beacon]

@tulip sleet the working commands for flashing the raw ESP are at https://github.com/robmarkcole/circuitpython-on-home-assistant/blob/master/assets/cheatsheet.md and the docs to update are https://learn.adafruit.com/welcome-to-circuitpython/circuitpython-for-esp8266#program-esp8266-13-11

[idle owl]

@slender iron Any idea why RTD would successfully build for the updated version number, but latest is not updating?

[cedar beacon]

I got the correct commands (for raw esp) from http://docs.micropython.org/en/latest/esp8266/tutorial/intro.html#intro evidently my esp has the particular FlashROM configuration described there

[tawny creek]

When using W25Q16JV_IQ as the definition the status light blinks white then stays solid blue
When using W25Q16JV_IM as the definition this error comes up:

In file included from ../../supervisor/spi_flash_api.h:32:0,
                 from ../../supervisor/shared/external_flash/external_flash.c:31:
../../supervisor/shared/external_flash/devices.h:226:6: error: 'external_flash_device {aka const struct <anonymous>}' has no member named 'has_quad_enable'
     .has_quad_enable = true, \
      ^
<command-line>:0:44: note: in expansion of macro 'W25Q16JV_IM'
../../supervisor/shared/external_flash/external_flash.c:48:78: note: in expansion of macro 'EXTERNAL_FLASH_DEVICES'
 const external_flash_device possible_devices[EXTERNAL_FLASH_DEVICE_COUNT] = {EXTERNAL_FLASH_DEVICES};
                                                                              ^~~~~~~~~~~~~~~~~~~~~~
make: *** [build-espress0/supervisor/shared/external_flash/external_flash.o] Error 1

So at one point I tried removing that line .has_quad_enable = true, \ and it builds fine, but CIRCUITPY doesn't mount

@exotic pumice

[exotic pumice]

so does it get further when you use IM?

it gets through the bootloader but not to circuitpy?

[tawny creek]

@exotic pumice yup confirmed

CIRCUITPY doesn't mount

[manic glacier]

kattni

[adafruit/circuitpython] Pull request closed: #1348 Fix Driver RTD Links

kattni

[circuitpython:master] 4 new commits

d7aa790 add trailing '/' to links missing them - sommersoft
59430ea fix L3GD20 link - sommersoft
4254ae6 add 74HC595 - sommersoft
f0d02fc Merge pull request #1348 from sommersoft/fix_dr... - kattni

[tawny creek]

I had to remove .has_quad_enable = true, \ though

[exotic pumice]

yeah, I'm just looking at that file now

[tawny creek]

also can get into REPL

[idle owl]

Nevermind. It finally updated latest.

[exotic pumice]

I can't figure out why it can't see the has_quad_enable in the struct

is it in your local copy?

devices.h:52

[slender iron]

@idle owl glad it did because I had no idea

[idle owl]

Yeah I don't know what the deal was. I built it manually multiple times and it refused to update.

¯_(ツ)_/¯

[tawny creek]

@exotic pumice It isn't

[exotic pumice]

alright, you probably have to pull, submodule sync, et al

something like

git submodule sync
git pull --recurse-submodules
git submodule update --init --recursive
```

[tawny creek]

Its from a fresh clone too, but will run this

I don't see bool has_quad_enable : 1; on the latest release

[exotic pumice]

I see it on master

[tawny creek]

https://github.com/adafruit/circuitpython/blob/master/supervisor/shared/external_flash/devices.h

[exotic pumice]

oh whoops I'm not on master

I'm on siddacious fork

[tawny creek]

also just plopped that line into struct ( bool has_quad_enable : 1;) and circuitpy still doest mount

[exotic pumice]

did you add the backslash too?

@slender iron, it looks like you might've accidentally undone siddacious' changes? https://github.com/adafruit/circuitpython/commit/9d91111b1b6a1eee77a88b13faf695ee0c5caea3

or actually, it looks intentional

[slender iron]

I redid it

[tawny creek]

@exotic pumice backslash?

where

[exotic pumice]

@tawny creek try this .quad_enable_bit_mask = 0x02, \

[tawny creek]

That line is already in both

[exotic pumice]

ok, sorry

yeah so removing the .has_quad_enable = true, \ was probably the right thing to do

as for CIRCUITPY not mounting ¯_(ツ)_/¯

@slender iron it looks like you missed removing some of the has_quad_enable = trues

[slender iron]

maybe

I built all the board defs

[tawny creek]

ahh ok thanks @exotic pumice !

[exotic pumice]

want me to PR @slender iron?

[slender iron]

sure! want to add a board too?

[exotic pumice]

what board?

@cascade's board?

Maybe I'll let them take the PR

[tawny creek]

I can do the PR for the board but CIRCUITPY isn't mounting :S so have to make that work first no?

[exotic pumice]

Do you have a debugger @tawny creek ?

[tulip sleet]

@cedar beacon thanks - i have saved that post and will look at this, though maybe not until after thanksgiving

[tawny creek]

@exotic pumice jlink mini

[timber mango]

Mu is giving me the error MemoryError: memory allocation failed, allocating 1884 bytes does this mean I have written too much code? I am confused ecause the file is only 7kb big and the Trinket M0 I am using still has over 30kb left?

pasted the wrong thing, thats the error I'm getting

I am new to this btw

[tulip sleet]

you're running out of ram, not filesystem space

[timber mango]

oh okay

[tulip sleet]

yes, there's too much code, or similar. sometimes doing the biggest imports first helps a little

[timber mango]

Okay I actually had two unused imports and after that I moved my imports around to make the biggest imports first, now the error message changed, it no longer says "1884 bytes", now it says "%u bytes". Why is it doing that?

[tulip sleet]

it doesn't even have space to allocate the string to print the error message 😃

[timber mango]

oof but it should actually have more space now.. Is there any way to make it consume less space, or do I need to cut out unneccesary code?

[exotic pumice]

@tawny creek maybe try stepping through the circuitpy mounting code?

[tulip sleet]

so did you compile some of the .py's to .mpy?

[timber mango]

or try to remove some code somehow, because I actually need all of it right now

no, I didn't

does that help?

I was about to try that, then I figured It'd probably better to ask here first

[tulip sleet]

yes, because compiling the .py's on the board can fragment the heap, or even not work because the code is too big

just get the mpy-cross for your version and platform. It compiles one .py file at a time.

[exotic pumice]

@tawny creek or see if there's any error messages in your os logs

[tawny creek]

@exotic pumice 😅 I’ve never done this process before, is there a suitable learn guide for this

[exotic pumice]

https://learn.adafruit.com/debugging-the-samd21-with-gdb/overview

[timber mango]

Can I do the mpy-cross compiling on windows too? Because It says I need a UNIX based system and I don't want to boot up my raspberry lol

[tulip sleet]

you can use the mpy-cross .exe that you see in the release assets. What vesrion are you using?

e.g. https://github.com/adafruit/circuitpython/releases/download/3.1.1/mpy-cross-3.x-windows.exe

[timber mango]

okay nevermind I was completely lost, all I found was the soruce code with the Readme saying that I need a unix system

thanks and sorry for my stupid questions

[tulip sleet]

they're not stupid! this stuff is complicated!! That's why we're here.

[marble hornet]

if you try and except a block of code but there is a memory error in the code is the tried stuff wiped from memory?

[tulip sleet]

everything is in memory up to the time the exception happens

[tawny creek]

@exotic pumice I got a setup running, mind walking me through how to properly debug this?

[marble hornet]

so when i want to exit a program might i flush it from memory like that?

"flush" being the wrong term

[tulip sleet]

you have to do a soft reload (raise ReloadException) or a hard reset

[marble hornet]

ah well, just hopefully double checking, thank you

[tulip sleet]

the soft reload will clear everything from the heap, but maintain the USB connection

[exotic pumice]

@tawny creek find wherever the circuitpy is mounted in the code, place a breakpoint nearby, step through and see what happens

[tawny creek]

@exotic pumice okies, looking for it now

@exotic pumice https://gist.github.com/eighthree/42ae852e0ca83bc7565f3eb67bf8ab5e

It went from a white status light (normally flashes), and cant step any further as the status stays blue and stepping doesn't do anything

[exotic pumice]

That looks like res != FR_OK to me, right?

I haven't looked closely at this code

[tawny creek]

did some of this as well: https://gist.github.com/eighthree/42ae852e0ca83bc7565f3eb67bf8ab5e#file-debugger_2

[exotic pumice]

So I think we've identified the error a little more if we're triggering that res != FR_OK block, now you should investigate what that means and what is causing it

[slender iron]

I'm around to help for a bit too. @tawny creek is your current code pushed anywhere?

[tawny creek]

@slender iron nope, I only have a new board definition folder -- I can publish it

[slender iron]

ya, its easiest to push it all I think

[exotic pumice]

oh I was wrong

[tawny creek]

https://github.com/eighthree/circuitpython/tree/espress0-board-def/ports/atmel-samd/boards/espress0

@slender iron here's a branch with the board def

[slender iron]

looking now, thanks!

[exotic pumice]

I'll let @slender iron take over

[slender iron]

do you have gdb going? does the board work with an older version?

[tawny creek]

yep i have GDB running and yes, up to 4.2

[slender iron]

k, can you run it until it freezes, ctrl-c then get a backtrace?

[timber mango]

If I compile code into a .mpy, and the code I am compiling relys on other libraries, do I need to do something special?

[slender iron]

nope, it just makes the one file smaller

[timber mango]

Okay because I am running into some strange issues

[tawny creek]

@slender iron https://gist.github.com/eighthree/42ae852e0ca83bc7565f3eb67bf8ab5e

freezes around line 78 on that gist

[slender iron]

try it with continue and not next

next can be funny since its trying to break somewhere close

[timber mango]

I have a method called wString in which I am calling another method called writeLetter very often. It compiles fine, but in certain lines, I get an error saying the name writeLetter isn't defined

[slender iron]

it does sound like something funny with the flash

@timber mango have you posted your code somewhere yet?

[timber mango]

No Idea why and before I tried putting into its seperate class to compile it to make it take less ram (and before it was too big to run) everything worked,

no

[tawny creek]

https://gist.github.com/eighthree/42ae852e0ca83bc7565f3eb67bf8ab5e#file-backtrace-2

not sure if i should Ctrl+C or wait a certain time

I hit ctrl+C when it stays a steady blue

[slender iron]

I'd wait a little bit

that is when its running boot_py which is blue

does boot.py do anything?

actually, it hasn't started yet. its looking for the file

[tawny creek]

waiting a minute or two yields the same backtrace

[slender iron]

its totally possible I broke it

try printing never_reset_sercoms

[tawny creek]

so with the IQ variant, it has this behaviour. With the IM variant I need to remove .has_quad_enable = true, \

[slender iron]

https://github.com/eighthree/circuitpython/blob/espress0-board-def/ports/atmel-samd/common-hal/busio/SPI.c#L44

[tawny creek]

IM boots but spi flash doesnt load

[slender iron]

whats actually on the board? I'd put just the one on the board in the list

[tawny creek]

never_reset_sercoms
$1 = {false, true, false, false}

[slender iron]

k, that looks ok

[tawny creek]

W25Q16FVSIG

[slender iron]

ok, let me pull up the datasheet

[tidal kiln]

@timber mango sounds like a scope issue. can you post code?

[timber mango]

Directly here or to github?

(its not much at all)

[sacred edge]

Brand New to Circuit Python or most any programming really, Received my circuit playground express today.. now what?

[tawny creek]

@sacred edge have you done the tour? https://learn.adafruit.com/adafruit-circuit-playground-express/overview

[exotic pumice]

There's also https://learn.adafruit.com/welcome-to-circuitpython/overview

[sacred edge]

yes, I did the tour and ran the basic "Blinky" program, and have the library's loaded, and examples

Keep in mind, I am a 63 year old newbie to this programming stuff!!

[slender iron]

@tawny creek I can't find a datasheet for it

[tawny creek]

iirc its the same as https://www.winbond.com/resource-files/w25q16fw reve 10072015 sfdp.pdf

[slender iron]

is it a W not V?

[tidal kiln]

@timber mango you can gist or pastebin somewhere, or directly here if you want. you can attach as file or just the code if it's only several lines.

[idle owl]

@sacred edge Have you gone through this one? https://learn.adafruit.com/circuitpython-made-easy-on-circuit-playground-express/

[tawny creek]

It shows a V but when I look for the datasheet and what those letters mean that part only determines the size

[slender iron]

k, its a 1.8v part which is a bit weird but should be ok

[sacred edge]

kattni, Yes, I just did... now I need to learn this new language I suppose.. or find code I can run and learn form

[tawny creek]

ah sorry I meant this: https://www.winbond.com/resource-files/w25q16jv spi revd 08122016.pdf

[slender iron]

@tawny creek what happens when you just put that in the list?

[sacred edge]

@idle owl that is how I found the discord group... 😃

[idle owl]

@sacred edge There are a lot of projects available on the Learn system that use the Circuit Playground Express. I would see what you can find there for new ideas. The next step will be getting a few peripherals and attaching them to your Circuit Playground Express.

[tidal kiln]

@sacred edge and this one:
https://learn.adafruit.com/circuitpython-essentials/circuitpython-essentials

[sacred edge]

@tidal kiln , Thanks!! something new to follow...

[idle owl]

You've already gone through the intro stuff, the Essentials guide starts to introduce some more complicated code.

There are some other items needed to do everything in that guide though.

[tawny creek]

@slender iron let me try again, i copied the definition for W25Q16JV before the split earlier

[sacred edge]

Thanks @idle owl

[tawny creek]

ill also try the FW to double check, and only using 1 spi

vs. defining three different ones

[idle owl]

@sacred edge You're welcome! Good luck! We're here to answer any questions you have as you go along 😃

[slender iron]

@tawny creek kk, the definition should basically be the same

[tidal kiln]

yep. welcome/intro guides then example projects. that's a good progression.

[slender iron]

I did move spi flash to using the internal common_hal apis instead of dedicated code

[timber mango]

https://github.com/NTDuckler/DuckyPy I am trying to recreate a USB RubberDucky using a Trinket M0 btw @tidal kiln

and I am new to python so I probably have lots of errors lol

[tidal kiln]

where's wString?

nvm. i think it see what you're trying to do.

[timber mango]

Iirc i renamed it to just string

I am actually just trying to split a string and then type the character

but It seems like I am to stupid

[tidal kiln]

you need to add self. in front of your calls to the class member, ex:
https://github.com/NTDuckler/DuckyPy/blob/master/DuckyPy.py#L12

[timber mango]

everything worked fine until It got too big, then I tried putting most of it into a seperate class which I would then try to pre compile, but that is where I got a lot of strange errors

[tidal kiln]

needs to be:

 self.writeLetterU(Keycode.A)

[timber mango]

oh okay

will try that, thanks :D

[tidal kiln]

is that your first time writing a class?

[timber mango]

yes

[slender iron]

https://github.com/adafruit/Adafruit_CircuitPython_HID/blob/master/adafruit_hid/keyboard_layout_us.py#L197

[timber mango]

this is my first python project actually

[tidal kiln]

awesome. jumping right in!

[slender iron]

the us layout driver can type a string for you too

[timber mango]

Now I feel stupid for not figuring that out on my own rip

really..?

[tidal kiln]

have you done anything with classes before?

[tawny creek]

Added:

// Settings for the Winbond W25Q16JV 2MiB SPI flash.
// Datasheet: https://www.winbond.com/resource-files/w25q16jv%20spi%20revf%2005092017.pdf
#define W25Q16JV {\
    .total_size = (1 << 21), /* 2 MiB */ \
    .start_up_time_us = 5000, \
    .manufacturer_id = 0xef, \
    .memory_type = 0x40, \
    .capacity = 0x15, \
    .max_clock_speed_mhz = 133, \
    .has_sector_protection = false, \
    .supports_fast_read = true, \
    .supports_qspi = true, \
    .has_quad_enable = true, \
    .supports_qspi_writes = true, \
    .write_status_register_split = false, \
}

and the mconfigboard.mk is:

SPI_FLASH_FILESYSTEM = 1
EXTERNAL_FLASH_DEVICE_COUNT = 1
EXTERNAL_FLASH_DEVICES = "W25Q16JV"
LONGINT_IMPL = MPZ

Yields the same backtrace results, and (gdb) print never_reset_sercoms
$1 = {false, true, false, false}

[timber mango]

you mean in another coding language?

[slender iron]

don't feel bad, if you couldn't find it its a doc problem

[tidal kiln]

yep. in a general programming sense.

[timber mango]

yeah, in java

[tidal kiln]

looks like you figured out needing to have at least a self parameter for each method

[slender iron]

@tawny creek do you have an logic analyzer?

[timber mango]

I need to add that self in the other methods like closeApp too, right?

[tawny creek]

@slender iron I doo but nothing like a salae

[slender iron]

what do you have? can you hook it up to the flash?

[tawny creek]

I have this https://www.sparkfun.com/products/15033

[slender iron]

should be fast enough

I wonder if is actually starting up ok

[tidal kiln]

@timber mango in general, need to add self. in front of anything to access the object members. your kbd is defined at the class level though, so you could prepend with class name.

[tawny creek]

How shall I connect it?
Also, I tried with the FW variant and it boots but still no CIRCUITPY

I think this has to be with the 16J vs 16W though

[slender iron]

you can break inside this return to see if it finds the flash: https://github.com/eighthree/circuitpython/blob/espress0-board-def/supervisor/shared/external_flash/external_flash.c#L210

[timber mango]

Okay, thanks!

[slender iron]

(I usually add asm("nop"); to give it an instruction to break on)

if it ends up there then something is mismatched with the jedec_id_response

where did you buy the chip from? its weird that there is no datasheet for it specifically

[tawny creek]

I'm going to revert to using W25Q16JV_IQ

@slender iron this one's from a far-east supplier

[slender iron]

ah ok

that logic analyzer should help snoop on it but you'll have to figure out how to connect it. the saleae comes with tiny grabbers

[tidal kiln]

@timber mango more specifically - python passes in the instance as the first parameter to every function. that's why you need at least one parameter for every function. you can actually name it whatever you want, but convention is to use self

silly example:

>>> class Foo:
...     def __init__(turkey_dinner):
...         turkey_dinner.msg = "gobble gobble"
...     def speak(gravy):
...         print(gravy.msg)
... 
>>> f = Foo()
>>> f.speak()
gobble gobble
>>> 

[tawny creek]

@slender iron have tiny grabbers on this hehe

hooking it up now

[slender iron]

nice!

[tawny creek]

which lines shall i connect

[slender iron]

I'd do the whole spi bus

that way you can see the whole conversation 😃

[timber mango]

Oh my god it finally works

thanks for everything guys :D

[tawny creek]

I need finer probes I think, they are pushing each other off ^^;

[tidal kiln]

look at the probes just posted over in #help-with-projects 😀

[tawny creek]

@tidal kiln oooh! I can't get my hands on one right now... so i just soldered wires directly onto the flash

[tidal kiln]

no worries. more of a joke suggestion. they are $pendy items.

[exotic pumice]

members of the circuitpython team should do an interview on the Talk Python Podcast. Or have you already?

[tawny creek]

@slender iron oki got the logic analyzer connected

This is with 4.3

This is how things looks like with CP 4.1

[manic glacier]

dhalbert

[adafruit/circuitpython] Pull request opened: #1349 Latest Feather nRF52840 pin revisions

Pins revised based on latest rev C schematic. Some pin names revised.

Only Q is if the SWITCH should be called BUTTON, since it's momentary. Left it at SWITCH for now.

[sacred edge]

in makecode I built a program, and it generated a .uf2 file... now what?

[tawny creek]

@sacred edge double press the reset button, drag the .uf2 file on the CPLAYBOOT drive

[slender iron]

@tawny creek you are only capturing one of the four channels so its hard to debug

I think its stuck waiting for status at some point

[tawny creek]

@slender iron Ah sorry! First time using this...

[slender iron]

np 😃

[tawny creek]

where should I connect SPI enable to? Pin 1, 3, or 7?

[slender iron]

its also known as chip select

(the protocol analyzer doesn't work without it)

[tawny creek]

ok, got it (I think)

http://www.timothygarcia.ca/cp/8-MHz-480MSamples.logicdata

[teal thorn]

has anyone tried to add support for micro sd cards as the storage? this could be a way to add more storage space for boards without the flash chip.

[exotic pumice]

probably. especially with grand central on the way

that's just me speculating though

in the mean time there's this https://github.com/adafruit/Adafruit_CircuitPython_SD

[slender iron]

@tawny creek it looks like that is missing enable still

[tawny creek]

@slender iron http://www.timothygarcia.ca/cp/sample.logicdata

how long should I run it for?

[slender iron]

5 seconds is plenty

hrm, the decoder is still unhappy

[tawny creek]

the initial state of the CLK line does not match the settings

[slender iron]

ya, it might be running too fast for your probe

and having wires on it can slow things down too

you can drop the max_clock_speed_mhz to reduce it

always good to do when debugging

[tawny creek]

where can I change this setting? thanks for all this help btw & the patience ><

[slender iron]

in the devices.h file

[tawny creek]

If i capture as it continues, i dont see that error - is that because It doesn't see the initial state? also tried dropping max_clock_speed_mhz to 100 and still getting output with that on the decode panel

[slender iron]

100 is really fast because its in megahertz

try 1

(its 12 or 8 to start)

[tawny creek]

okay, I dont see that same error anymore:

http://www.timothygarcia.ca/cp/sample.logicdata

also set the speed to 1

[tawny creek]

swapped flash chips to ones that say W25Q16BVG and that works on the same board definition on Alpha 2

[manic glacier]

neeels

[adafruit/circuitpython] Issue opened: #1350 using factory delivered main\.py with 4\.0\.0 Alpha 3 fw leads to "OSError: \[Errno 5\] Input/output error" on feather\_m0\_basic

I recently purchased a feather M0 basic. Just now I installed the 4.0.0 alpha 3 firmware
and updated the .mpy files, from:

• adafruit-circuitpython-feather_m0_basic-en_US-4.0.0-alpha.3.uf2
• adafruit-circuitpython-bundle-4.x-mpy-20181121.zip

obtained from the releases pages on github, installed by copying onto FEATHERBOOT / CIRCUITPYTHON usb drives.

The feather boots fine with 4.0.0 alpha 3, I get a ttyACM0, but the main.py fails to run:

Press any key to enter the REPL. Use C...

[teal thorn]

i could use the sd card library, but i don't think it would be accessible from usb,. a native sd implementation would be better.

[manic glacier]

neeels

[adafruit/circuitpython] New comment on issue #1350: using factory delivered main\.py with 4\.0\.0 Alpha 3 fw leads to "OSError: \[Errno 5\] Input/output error" on feather\_m0\_basic

Here is the complete main.py as it is now. It does in fact contain minor changes from me playing with it, but only after the failing line:

# Feather Basic IO demo
# Welcome to CircuitPython! :)

import time
import simpleio
from adafruit_hid.keyboard import Keyboard
from adafruit_hid.keycode import Keycode
from digitalio import DigitalInOut, Direction, Pull
from analogio import AnalogOut, AnalogIn
import touchio
import neopixel
import board

# Built in red LED
led = Digit...

neeels

[adafruit/circuitpython] New comment on issue #1350: using factory delivered main\.py with 4\.0\.0 Alpha 3 fw leads to "OSError: \[Errno 5\] Input/output error" on feather\_m0\_basic

...and I notice the fw 3.1.1 works fine without having to unstall the 3.0.0 bundle .mpy files, so the error appears to be in the 4.0.0 alpha 3 fw, not in the .mpy.

[cedar beacon]

OK so there is a web-ide for micropython, and the author is interested in adding circuitpython and serial support http://www.1zlab.com/ide/#/

[slender iron]

@gentle bronze I'm going to work on the Makefile now

[gentle bronze]

@slender iron ah thanks, that will be very helpful. But you should sleep now instead :)

[slender iron]

@gentle bronze I'm in seattle so its only 10:21 pm. I got a good couple of hours in me 😃

[gentle bronze]

👍

Don't worry to support all mcu or ports, just get your samd example running. We can generalize it later :D

[slender iron]

sounds good

[gentle bronze]

Please pull first, I move the example around a bit, you should start with the cdc_msc_hid example

[slender iron]

yup, I did @gentle bronze

[onyx hinge]

hm a few speed bumps with this clone I hadn't updated in awhile, had to monkey with the fetch URLs for the submodules. no harm done.

[tawny creek]

Thanks for the assistance yesterday @slender iron and @exotic pumice ! Ordered myself other chips from LCSC to bundle with a recent order and maybe i can try and poke at this at a later time~

I learned a few new things I can do to debug ☺

[manic glacier]

jepler

[adafruit/circuitpython] Pull request opened: #1351 json: enable on nrf boards and "large flash" samd boards

.. as discussed in #1307.

Testing performed: Only that it builds -- none of my CP hardware is handy.

Keda87

[adafruit/circuitpython] New star added

[manic glacier]

turbinenreiter

[adafruit/circuitpython] New comment on issue #703: Ethernet FeatherWing Library

Hi,

I tried this with the latest alpha. When I run the script from the comment in the Merge Request, I get

• no serial port
• no CIRCUITPY drive

Just wanted to ask what the current status here is and if it's supposed to work, before I file a detailed bug report.

Using an M4 Express with the Ethernet Featherwing.

[timber mango]

@danh#1614 regarding SPI bitbanging and TLC5957 LED-Driver (my topic from yesterday...) :
it basically works ;-)
my write is a little slow - but that is expected..

[manic glacier]

tannewt

[adafruit/circuitpython] New comment on issue #703: Ethernet FeatherWing Library

Does USB only fail with the script present? We've had a lot of USB churn recently so it could be unrelated to the network code.

[meager fog]

happy 🦃 day for those who are celebratin'

[manic glacier]

turbinenreiter

[adafruit/circuitpython] New comment on issue #703: Ethernet FeatherWing Library

Will try to test that tomorrow.

ladyada

[adafruit/circuitpython] New comment on pull request #1349: Latest Feather nRF52840 pin revisions

checkin out by checking out

[manic glacier]

ololobus

[adafruit/circuitpython] New star added

[lone sandal]

MicroHex

[adafruit/awesome-circuitpython] New star added

[manic glacier]

ladyada

[adafruit/circuitpython] New comment on pull request #1349: Latest Feather nRF52840 pin revisions

ok there's enough weird stuff - QSPI not workin, neopixel malloc crashes, and then tinyusb, that will need to pause reviewing this until there's some other fixes. i did get the red led to blink but thats kinda it :)

micro-led

[adafruit/circuitpython] Fork created: micro-led/circuitpython

[manic glacier]

ladyada

[adafruit/circuitpython] New comment on pull request #1349: Latest Feather nRF52840 pin revisions

im ok preliminarily merging even tho QSPI isnt working - all the other GPIO are tested!

[manic glacier]

dhalbert

[adafruit/circuitpython] New comment on pull request #1349: Latest Feather nRF52840 pin revisions

I set the SPI flash chip to GD25Q16C, per the schematic. Is that correct? Previously was GD25Q64C, maybe from the Trellis?

[manic glacier]

ladyada

[adafruit/circuitpython] New comment on pull request #1349: Latest Feather nRF52840 pin revisions

yeha its correct - we got it working in SPI mode, but not QSPI mode, it busywaits for some reason. we're not sure why but i dont think its related to the pin names, maybe something in the QSPI peripheral?

[worldly kayak]

Hi. I have just written a driver for the SH1106 OLED display and wanted to create a pull request, but with it being a submodule of the Adafruit_CircuitPython_Bundle I am not sure how to submit the pull request. The new driver can be found at https://github.com/winneymj/Adafruit_Python_SH1106. I forked the Adafruit_CircuitPython_Bundle and created a submodule for my driver under the libraries/drivers/sh1106, but I am not sure what to do now to create a pull request for a submodule.

[tulip sleet]

@worldly kayak Great! FYI, the regular bundle is just a collection of existing libraries with their own repos., and is all official drivers from Adafruit. Since we don't sell an SH1106 display, it makes more sense to submit to the community bundle: https://github.com/adafruit/CircuitPython_Community_Bundle. The naming convention is <something>_CircuitPython_SH1106, or just CircuitPython_SH1106. A "Python" library would be for conventional Python. and we've mostly deprecated those in favor of the CircuitPython ones.

[manic glacier]

dhalbert

[adafruit/circuitpython] New comment on pull request #1349: Latest Feather nRF52840 pin revisions

OK by me if you want to merge it with QSPI broken, so we can test it against upcoming USB changes easily. If you approve I will merge. The QSPI thing is weird, but if those are different pins or a different chip than anything used with the nRF52840 before, I could believe it needs more debugging.

ladyada

[adafruit/circuitpython] New comment on pull request #1349: Latest Feather nRF52840 pin revisions

cool - yeah i think its not the definition here. approvin'!

ladyada

[adafruit/circuitpython] Pull request closed: #1349 Latest Feather nRF52840 pin revisions

ladyada

[circuitpython:master] 2 new commits

5fe746f Latest Feather nRF52840 pin revisions - dhalbert
3ca74d8 Merge pull request #1349 from dhalbert/feather5... - ladyada

[lone sandal]

occoder

[adafruit/awesome-circuitpython] New star added

[errant grail]

Wrapped up the front panel and CV/audio output tests tonight (CircuitPython 3.1.1 / Feather M4 EX w/custom Eurorack-compatible motherboard). Using a potentiometer as the voice selection "switch" to gain space behind the panel. Coding the selection helper to include noise reduction and hysteresis was kinda fun. Should be easy (?) to incorporate the MIDI stack and the DSP-1 default voices now. https://youtu.be/I1jVkATQ81E

[pastel panther]

@errant grail That's awesome! Eurorack in a lunchbox!

[manic glacier]

ptone

[adafruit/circuitpython] Fork created: ptone/circuitpython

[velvet badger]

Do f strings work in circuitpython, yet?

[pastel panther]

format strings?

[velvet badger]

yeah. e.g. print(f'value: {value}')

[tidal kiln]

Adafruit CircuitPython 3.1.1 on 2018-11-02; Adafruit Feather M0 Express with samd21g18
>>> foo="world"
>>> f'hello {foo}'
Traceback (most recent call last):
  File "<stdin>", line 1
SyntaxError: invalid syntax
>>>  

[velvet badger]

bummer. ah well.

thanks for the info

[tidal kiln]

yah. looks like no. probably one of those space things.

[velvet badger]

no doubt. and likely not a priority since there isn't typically a whole lot of print output except for error messages to the console.

[tidal kiln]

there may be support and it's just not compiled in for these targets, but i'm not sure

[velvet badger]

I don't think they were in regular python until 3.5 or 3.6. I think the old way was something like print('Hello %s', world)

Will dig around. Thanks again!

[manic glacier]

Gigahawk

[adafruit/circuitpython] Issue opened: #1352 Human readable OSErrors should include 30 and 28

#837 added messages for common errors, but the original issue that brought it up (#221) asked for code 30 (Read only filesystem) and this tutorial talks about 28 (out of storage).

I feel like these exceptions are pretty common and should warrant human readable descriptions to go with them.

YazzyYaz

[adafruit/circuitpython] New star added

[slender iron]

@velvet badger MicroPython (and therefore CircuitPython) was targeted at 3.4

[velvet badger]

Ah, ok thanks @slender iron

[manic glacier]

BillHeintz

[adafruit/circuitpython] New star added

BillHeintz

[adafruit/circuitpython] Fork created: BillHeintz/circuitpython

[half sedge]

Has this thing been released yet? Or at least more than just hinted on twitter: https://twitter.com/adafruit/status/1055549781873999874

adafruit industries (@adafruit)

#comingsoon open-source python powered badge and gaming platform that is @adafruit feather compatible - will have e-ink version too @CircuitPython #adafruit

Twitter

[meager fog]

no its not out yet

INOYDA

[main meteor]

The one I'm waiting for is Grand Central (and hoping it uses the SAMD51 high-speed microSD interface).

[pastel panther]

so much good stuff in such a little board

"little"

[meager fog]

@main meteor its plain SPI interface, we dont have a driver for the SD HS interface

[main meteor]

Ah well. I guess microSD isn't really supposed to be fast storage anyway (which is why there's a QSPI flash chip on there, I suppose).

[meager fog]

if someone contributes a driver, we'd be into changing the pins

the delays in SD cards are the internal FS erase/writes - that wont be any faster

same with QSPI

cameras have huge SRAM buffers

SPI for SD card is well supported, and understood 😃

[main meteor]

My 1200fps camera really beats up the SD card.

[meager fog]

yeah the engineering in cameras is sucking that data from the sensor, buffering it, and managing the SD card. its kinda the whole job of the camera 😄

we're not quite there yet

[slender iron]

@meager fog want a new nrf build? I think the usb is happier with thach's fixes

[meager fog]

ooooh yes

im not home where my nrf52's at

but...ill do it when i get there

[main meteor]

INOYDA? "I know, yeah, don't ask"?

[slender iron]

actually I don't have the new board def in my repo

[meager fog]

its not out yet dont ask

hmm can you cherrypik it

i merged

[main meteor]

I was close! 😃

[slender iron]

ah! I do have it. I didn't realize it was merged

has been ignoring github emails

[manic glacier]

tannewt

[adafruit/circuitpython] Pull request opened: #1353 Update tinyusb to include control fixes\.

I had to remove a few unhelpful warnings and tweak the build but easy overall. Works better than before in my testing with nrf and samd51.

tannewt

[adafruit/circuitpython] New comment on issue #1342: samd: REPL disconnect problems after tinyusb PR

Please try with https://github.com/adafruit/circuitpython/pull/1353. I believe it will fix it.

[meager fog]

@slender iron oh wait i have the boards here

ok can you give me a uf2

ill test it now

[manic glacier]

tannewt

[adafruit/circuitpython] New comment on issue #1340: nRF boards slow to enter REPL

Please try with https://github.com/adafruit/circuitpython/pull/1353 in my limited testing it fixed the slowness. (It was due to an incorrect/missing response to setting the CDC line coding on connection.)

[slender iron]

yup, lemme build

nrf840 feather

[meager fog]

oh shoot i forgot we didnt commit the SPI stuff or neopixel removal

[slender iron]

I can do it

@meager fog do you have the spi pin defines handy?

[meager fog]

hold on

lemme see

[slender iron]

I think I got it

MOSI is DATA0 iirc

[meager fog]

thats what i used to build - not perfect but worked

[slender iron]

yup thats what I have in the last firmware I posted

[meager fog]

ok thanks yeh that works

lemme see if i can cDC

yeah!

it works 😃

i can putty in

n1c3

[slender iron]

yay!

[meager fog]

ok we do have some pin deinit stuff htat isnt workin'

Auto-reload is on. Simply save files over USB to run them or enter REPL to disable.
code.py output:
Traceback (most recent call last):
  File "code.py", line 7, in <module>
ValueError: SWITCH in use

on reset

[slender iron]

ya, maybe

[meager fog]

ok hmm

it does seem a lot more stable

[slender iron]

lemme see if I can spot the issue

[meager fog]

npnp

i can save MSD files, it detects, resets etc

CDC is solid, can copy / paste

[slender iron]

good! thats @gentle bronze 's good work!

[meager fog]

my review is 💯

until i can do more testing

[slender iron]

https://github.com/adafruit/circuitpython/pull/1353

[meager fog]

yep yep

[manic glacier]

tannewt

[adafruit/circuitpython] Pull request closed: #1353 Update tinyusb to include control fixes\.

tannewt

[circuitpython:master] 2 new commits

324301e Update tinyusb to include control fixes. - tannewt
e77c06e Merge pull request #1353 from tannewt/usb_update - tannewt

[meager fog]

excitin!

[slender iron]

yup! hope it fixes both open usb bugs

[meager fog]

i think acually MSD was stable before but i didnt see the in-use errors

so i thought the code hung on save, it was saving fine, just not running

[slender iron]

ah

[meager fog]

🤷

[slender iron]

that makes sense because it was the control endpoint with issues I think

msc doesn't use it much

it does control over the bulk endpoints

[meager fog]

rite

[slender iron]

haha

I know why the pin stuff isn't being reset

[half sedge]

Is this on your radar: https://github.com/boochow/micropython-raspberrypi/releases

[half sedge]

Barbone micropython on Raspberry Pi Zero...

[slender iron]

I know someone had tried that before

[meager fog]

https://blog.adafruit.com/?s=boochow

yep we blogged it even!

https://blog.adafruit.com/2018/08/10/micropython-on-bare-metal-raspberry-pi-zero-zero-w-boochowp/

[slender iron]

@meager fog does that new build work better for pin reset?

[meager fog]

testin'

[half sedge]

Now there is a release. Just unzip that on a microSD, add two file from the foundation, and it boot.

[slender iron]

https://github.com/adafruit/circuitpython/blob/master/ports/nrf/common-hal/microcontroller/Pin.c#L51

debugging code that I left in

[meager fog]

lol

ok hold on

[pastel panther]

haha

[slender iron]

that gets a second one

[meager fog]

yep

thats good now, no errors

[pastel panther]

resetz? who needs dem?

[meager fog]

ok try #2

[slender iron]

hopes the code below works

[meager fog]

this is totally like what it used to be progammin

you'd submit your card stack

check the next day for the output

[main meteor]

Can confirm.

[pastel panther]

That's about how I feel every time I have to compile and flash some code.

[meager fog]

super solid @slender iron

thank you 😃

sm00th as gravy

[slender iron]

k, I just thought of a simple neopixel fix

standby

[meager fog]

standin byyyy

[slender iron]

may take a bit more time

[meager fog]

no biggie

[slender iron]

and stack space 😉

[meager fog]

you will get your package soon too

i am testing neopixel in userland just fine

so i dont need it for the runtime

[tropic shuttle]

I am trying to run the example code in the readme of the bluefruit SPI library (https://github.com/adafruit/Adafruit_CircuitPython_BluefruitSPI/blob/master/README.rst) on a Feather M0 Bluefruit LE, but I get an error saying that board doesn't have an attribute D8. I guess this makes sense since there isn't a physical D8 on the board, but I'm confused as I thought this was an example for the board I am using.

This guide (https://learn.adafruit.com/adafruit-feather-m0-bluefruit-le/configuration) also uses pin 8. How do I use pin 8 if board doesn't have D8?

I am on CircuitPython 3.1.1 btw

[meager fog]

use the adalogger circuitpy build

[tropic shuttle]

@meager fog Thanks! That did the trick

[slender iron]

@meager fog status neopixel

[manic glacier]

tannewt

[adafruit/circuitpython] Pull request opened: #1354 A few fixes for nRF52840 feather QSPI and neopixel

[meager fog]

@slender iron ok want me to try again

[slender iron]

sure

[meager fog]

@slender iron enumerated just fine with MSD

does that mean QSPI is fixed?

[slender iron]

and the status neopixel works?

[meager fog]

hold on i gotta remove my code

[slender iron]

no, its setup to be spi

[meager fog]

k np

[slender iron]

I'll look at qspi when I get the hardware

[meager fog]

ok got it, yeah neopixel is good now

[slender iron]

great!

[tough flax]

Just received 3x Argon and 3x Xenon feathers from Particle... CP on nRF52 just got way more interesting for me 😃

[slender iron]

nice! I still haven't gotten my shipping notification

[tough flax]

They included a rebranded tripler - nice of them but I don’t like the blue 😃

I can send you one until you get one if it helps w/development

[slender iron]

nah, it should be any day

[tough flax]

Ok

Crazy mad stupid stress through Sunday’s event

[slender iron]

well I'm sure it'll go well 😃

[meager fog]

koool

[tough flax]

Will share vids esp the two cp projects

[meager fog]

adding board support should not be too hard if you want to attempt it bill!

its kinda a cross between the nrf52840 feather and the dongle (no SPI flash)

just gotta look up the pins 😃

[tough flax]

After the event I’m happy to help. I think the xenon won’t be hard

[meager fog]

totally - that would be great 😃

[tough flax]

The argon has an esp32 as well

[meager fog]

we also have the nrf52840 support in arduino coming along

that's also not too hard!

just needs the pin mapping.

[tough flax]

Is it SPI slave?

[meager fog]

the esp32?

[tough flax]

Yep

[meager fog]

absolytely no idea

😄

[tough flax]

Hehe

[meager fog]

i don't have one, not even a proto

we'll be stocking after they ship to backers of course

[tough flax]

Want one? I’m still at the ups store?

[meager fog]

nooo

i gotta finish the turky on my plate

[tough flax]

Gotcha

[meager fog]

grand cental and nrf52840

[manic glacier]

ladyada

[adafruit/circuitpython] New comment on pull request #1354: A few fixes for nRF52840 feather QSPI and neopixel

neopixel works!

ladyada

[adafruit/circuitpython] New comment on pull request #1354: A few fixes for nRF52840 feather QSPI and neopixel

lol

[tough flax]

Can I ask you something? Did you intentionally make the esp8366 able to be disconnected from the Lopo circuit so it could act like a wing? Or is that and jumper there for another reason?

Kilo

Stupid phone

LIPO

[pastel panther]

(s/search/replace)

doesn't do /g though so it only catches the first instance

[tough flax]

(s/ESP8266/ESP8266)

[pastel panther]

no parens, sorry

and it only matches your most recent post, I think

[tough flax]

Yep

Ok. Good to know! Full regex with back refs?

[pastel panther]

probably not

haven't tried though

[tough flax]

Ok. Gotta get heads down on this event

[timber mango]

It's lame but it basically will s/foo/bar and give surprise endings

Still easier than not at all so I use it a lot.

[pastel panther]

it's good enough

one can always just use edit

[timber mango]

or one could design a proper command line interface to a primarily text-driven user interface ;)

[pastel panther]

if one wanted to spend their time doing something like that

[timber mango]

oops I just looked at what room this is.

[meager fog]

@tough flax nah that jumper is so people can use AAA's or somehin

but you could use it as a shield for sure 😃

[tough flax]

I was thinking it would be a great stopgap for cp

Run spi slave sketch on it

Have cp initialize it and send / receive web requests using the esp8366 to do all the connection mgt and parsing

Just pass request/response back and forth

And the hardware already (kinda) exists. I think you need that jumper cut and perhaps not connect all the power pins

I certainly think it would irk when connected to usb or 5v. LIPO might not work

You could even run an http server on the esp and have it trigger the m0/m4 when a request is is received via a pin

[meager fog]

@tough flax yep we have some ideas on how to do it, theres's some 'ESP8266 as coprocessor' examplse out there

and we've done similar for BLE

but no ETA

[tough flax]

Gotcha. If you’d like help (after the event) let me know

[meager fog]

its a bit of an undertakin 😃

i have ideas where to start

[stuck elbow]

there are so many possibilities, you just have to write the code :(

[meager fog]

lol yeppp

story o my life!

happy 🦃 day 😃

l8r f0lx

[stuck elbow]

even in hardware, it always comes down to writing the code

[pastel panther]

anyone know of a serial plotter with a time scale that you can record traces on?

hmm.. maybe I can futz with arduino's to get it to do what I want

[tidal kiln]

there's one in mu

[pastel panther]

mu doesn't like me