circuitpython-dev

[raven canopy]

guess i'll just stick to reading the repo like i do with asf4. thanks @tulip sleet!

[tulip sleet]

@raven canopy you can clone the nrfx repo and build the documentation with doxygen. It's explained in the README. My impression is that nrfx changes much more slowly than the SDK.

@tawny creek do you have your board pushed to a repo?

i mean the circuitpython code

[raven canopy]

yeah, i was thinking about doing that just now. 😄

@tulip sleet i imagine you don't like reading nrfx. i think i recall you mentioning somewhere your love for this:

typedef enum
{

[tulip sleet]

no, I don't mind nrfx; it's better than ASF4 :). I don't remember saying anything positive or negative about enums, or are you talking about the indenting style?

[raven canopy]

the style. "better than ASF4"...it gets worse? 😆

[tulip sleet]

no, I like open curlies on the same line:
if (x) {
not

if (x)
{

[tawny creek]

@tulip sleet nope not at all

[raven canopy]

agreed. and my C (or any curly language) experience is infant...

[tulip sleet]

@tawny creek - I don't see anything conflicting in the trinket board def, but that error is due to some python code, prob, so I'd need to look at the whole .py

[tawny creek]

when you say whole .py , as in the board directory and it's contents?

[tulip sleet]

@raven canopy when I was using 24x80 terminals I was not found of extra line breaks, 'cause you could see even less of your code at once

[raven canopy]

@tawny creek do you have any SPI in your code? i think there is an open issue on SPI + UART...

[tulip sleet]

@tawny creek no I mean the python code that generates the pin in use

[tawny creek]

import board
import busio
import digitalio

uart = busio.UART(board.A6, board.A5, baudrate=9600)

while True:
    data = uart.read(32)  # read up to 32 bytes
    # print(data)  # this is a bytearray type

    if data is not None:
        # convert bytearray to string
        data_string = ''.join([chr(b) for b in data])
        print(data_string, end="")```

this is it

[tulip sleet]

good memory @raven canopy https://github.com/adafruit/circuitpython/issues/699

[raven canopy]

well...its not SPI causing it. 😄

[tawny creek]

Traceback (most recent call last):
  File "code.py", line 7, in <module>
ValueError: A5 in use



Press any key to enter the REPL. Use CTRL-D to reload.```

[raven canopy]

or I2C.. part of my brain is working. 🤣

[tulip sleet]

@tawny creek I think I'd have to see your circuitpython repo. I encourage you to fork adafruit/circuitpython, make a branch, and push your changes back to github

@tawny creek if you just reverse A6 and A5 does it work?

[tawny creek]

  File "code.py", line 7, in <module>
ValueError: A6 in use```

[raven canopy]

@tawny creek according to the PMUX table, PA05 is on SERCOM[0]. is anything else in your firmware using that? (flash, perhaps?)

PA06 is on the same SERCOM, as well.

[tawny creek]

@pulsar ferry re: do you have any SPI in your code -- - no but I am using SPI flash!

gonna look at the configuration and see :}

[raven canopy]

i should mention, i barely understand the SERCOM workings

[tawny creek]

// safer 8mhz.
#define SPI_FLASH_BAUDRATE (8000000)

#define SPI_FLASH_MOSI_PIN          PIN_PA16
#define SPI_FLASH_MISO_PIN          PIN_PA19
#define SPI_FLASH_SCK_PIN           PIN_PA17
#define SPI_FLASH_CS_PIN            PIN_PA11
#define SPI_FLASH_MOSI_PIN_FUNCTION PINMUX_PA16D_SERCOM3_PAD0
#define SPI_FLASH_MISO_PIN_FUNCTION PINMUX_PA19D_SERCOM3_PAD3
#define SPI_FLASH_SCK_PIN_FUNCTION  PINMUX_PA17D_SERCOM3_PAD1
#define SPI_FLASH_SERCOM            SERCOM3
#define SPI_FLASH_SERCOM_INDEX      3
#define SPI_FLASH_MOSI_PAD          0
#define SPI_FLASH_MISO_PAD          3
#define SPI_FLASH_SCK_PAD           1
#define SPI_FLASH_DOPO              0
#define SPI_FLASH_DIPO              3   // same as MISO pad
#define SPI_FLASH_CS PIN_PA11```

I believe the configuration should be the same as the trinket m0 haxpress

the only difference in my board is the SPI flash part but the configuration and board stuff should be the same

probably why I havent bothered to fork/push anything on github @tulip sleet , not different enough to the haxpress variation

only other difference is physically being able to access these pins via headers

[manic glacier]

jepler

[adafruit/circuitpython] New comment on pull request #1144: Add pulseio\.FrequencyIn

This is relevant to my interests. I have a metro m4 express and a GPS referenced oscillator, is there some useful testing I can do?

sommersoft

[adafruit/circuitpython] New comment on pull request #1144: Add pulseio\.FrequencyIn

@jepler,

Test? Yes, please! That would be much appreciated.

Usage is pretty easy:

import board
import pulseio

freq_in = pulseio.FrequencyIn(board.PIN)

# returns the reading in hertz
tc.value

[raven canopy]

@onyx hinge mind you, at current state, it will not detect >~512kHz...

[onyx hinge]

@raven canopy hmm is that limitation due to hardware? I can do anywhere from circa 1Hz to 12MHz. I suspect it'll just measure the error in the onboard oscillator of the me!

[raven canopy]

not hardware, really. more firmware. the interrupts get so fast it locks up...

[onyx hinge]

Locking up is certainly not good.

[raven canopy]

tends to not be... 😄

[onyx hinge]

I'll give it a try later when I'm home, it's really easy to generate an accurate frequency in to the M4 with my setup. Are there specific pins that have to be used?

[raven canopy]

any pin that has a timer (TC, not TCC) available. which...is most if not all. (guess I should have a better answer for that...)

[onyx hinge]

Will it give me an error if I make a bad choice?

[raven canopy]

actually...wait. All pins are a go. it uses the EIC. i abandoned direct TC capture moons ago... 😵

[onyx hinge]

Thanks for answering my questions, trying to read the patch on mobile wasn't too useful

[raven canopy]

github mobile...is unfriendly sometimes.

[onyx hinge]

From reading the m4 datasheet it looks like there is a mode to use a "reference clock" instead of the microcontroller's crystal. Looks like that isn't supported yet with your changes? That mode would eliminate the error or the internal crystal, as long as you have a high quality frequency source. That is probably a fairly rare situation though (having a frequency reference better than a microcontroller crystal)

[raven canopy]

are you looking at the FREQM peripheral?

[onyx hinge]

My actual goal is to discipline a 10MHz Crystal oscillator (vc-ocxo) to gps with circuit python.

Yes I think it was called freqm.

[raven canopy]

ahh...that only measures the internal clocks' frequencies. i had hopes for that in the beginning. but, for the M4 it ended up being useful, since the DFLL48 can only run in open mode so it isn't always a stable 48MHz. FrequencyIn uses the TC peripheral to capture the incoming signal and compares it to DFLL and spits out the amount of cycles the rise-to-rise takes.

[onyx hinge]

Without an accurate frequency source you are at the mercy of whatever is available

[raven canopy]

right. M0 is pretty rock solid. M4 drifts a little more. I don't think we'll be giving Saleae any competition. 😄

[onyx hinge]

Interesting. You'd expect improvements in the newer products!

[raven canopy]

M4 drove me crazy for a couple months, until Dan pointed out an errata about the DFLL open/closed mode. that i had read a few times but never connected the dots on.. 😄

[onyx hinge]

I'll let you know what I learn and I'll steer clear of frequencies higher than .5Mhz or so.

[raven canopy]

i do appreciate it. always hard to determine what you've done to make something work... it begs to be broken! 🤕

@onyx hinge with that broken bit, do try above 512k...that escape hatch has been spotty at times.

[onyx hinge]

I assume a tap of the reset button will rescue me?

[raven canopy]

it will.

[onyx hinge]

Great

For my application I ultimately need to measure at an accuracy of 1milliHz or less over time periods of an hour or more. it might just not be a good match for a "generic" frequency measurement module.

[raven canopy]

just noticed my example code snafu in the PR comment. don't introduce new object names that don't exist!

in the comment, i used freq_in to create the object. then used tc.value to get the result. tc is what i've been typing in the REPL to test this whole time. muscle memory...it'll get ya! 😆

[onyx hinge]

Gotcha

[tawny creek]

Is there a way to 'release' or determine what is using up a pin?

[manic glacier]

eighthree

[adafruit/circuitpython] Fork created: eighthree/circuitpython

[onyx hinge]

@tawny creek ooh not that I'm aware of but wouldn't that be awsome! Now I'm imagining that hypothetical.whatuses(board.D1) -> <spi object @...>

[meager fog]

@prime flower did you end up doing any neopixel stuff

[ruby atlas]

<stares at pixelbuf vscode windows sadly> (I really need to find a way to have more free time)

i scrolled up a bit and saw the neopixel_write traceback. what platform was that? Linux/Pi3?

[meager fog]

pi3

im working on it now

@prime flower whats the best library you found?

[prime flower]

@meager fog I did not get to it, the aio_a work took longer than I expected.

the one by jgarff (https://github.com/jgarff/rpi_ws281x) is the one I've been using. I might have done something wrong, but I've checked everything. It's the most up-to-date lib.

[ruby atlas]

Fun. 😃 Pi 3 B+ is a pretty decent device.

(I don't like the Wifi on any prior Pi)

[meager fog]

https://github.com/jgarff/rpi_ws281x ?

oops yeah

ok lemme get htat going

[prime flower]

I searched pimoroni thinking would have something but they just suggested the Arduino lib

[meager fog]

yeah it looks like its not on pypi

which makes things a little annoying

[prime flower]

https://github.com/rpi-ws281x/rpi-ws281x-python

[ruby atlas]

Fork and submit to PyPi?

i've had to do that a few times for projects that appeared abandoned and never submitted.

[prime flower]

@meager fog so there's a wrapper for it, sudo pip install rpi_ws281x, but it uses a fork (richardghirst's fork) of the main lib, which is 140 commits behind

[ruby atlas]

oy

[meager fog]

ughh

[ruby atlas]

time to fork both things.

[meager fog]

it is also called 'neopixel.py'

that sux

i wish i'd seen that, i would have told them to rename it

[tawny creek]

@tulip sleet I have forked and made a branch, but not pushed : https://github.com/eighthree/circuitpython/commit/bfe181a41d9e257d8b06553711ec23bca0b8952d

[meager fog]

https://github.com/rpi-ws281x/rpi-ws281x-python

[manic glacier]

jepler

[adafruit/circuitpython] New comment on pull request #1144: Add pulseio\.FrequencyIn

I gave this a whirl with an Adafruit Metro M4 Express together with the this GPS referenced RF generator.. In one terminal I connect to the M4, and in the other I connect to the RF generator and send frequency commands.

With the frequency set to 1kHz,

>>> sum(f.value for i in range(1000)) / 1000
958.077

With the frequency set to 2kHz,

>>> values = [f.value for i in range(1000)]
...

[meager fog]

@prime flower we're trapped in a name collision so yeah we can't use circuitpython neopixel at this time

[prime flower]

😦 dang

I'll just stick with the rpi-ws281x for the guide.

[pastel skiff]

HI, I tried to read before I posted but didn't see anything in this channel or on line... I am currently working with a CPX/CircuitPython and am experimenting with the sound sample playback. I have a simple stripped down playback program that I cobbled together from the adafruit examples working where I play a sound sample for each the left and right button. I'm a total noob and am attempting to add some code to turn on/off the onboard red LED with the buttons building on what i haveworking but when I simply add the import line "from adafruit_circuitplayground.express import cpx" the program freezes/refuses to run. I don't even have code referencing the "cpx", just the import line and the program dies...I am using adafruit-circuitpython-bundle-3.x-mpy-20180820 and have moved the adafriut_circuitplayground/express.mpy file to the lib folder. Any advice or insight on what is failing?

[tawny creek]

@pastel skiff are you using Mu? and is Mu freezing?

[pastel skiff]

no I'm using a multitude of text editors,...pycharm, Notepad ++, Idle to edit the file then drag and drop to update the main.txt

[tawny creek]

main.py you mean?

[pastel skiff]

haven't got MU running, should I work on that? yeah main.py

[tawny creek]

@pastel skiff If you're starting out I highly recommend it.

[manic glacier]

sommersoft

[adafruit/circuitpython] New comment on pull request #1144: Add pulseio\.FrequencyIn

When I reprogram the frequency source to 200kHz, it still works. But at 300kHz -- even without actually querying the frequency -- the board hangs.

☹️ Escape hatch fail. And possibly a little something else, too. I might need to re-visit emulating how @tannewt changed PulseIn to keep from endless looping.

For values below 1kHz, the results don't seem very good either.
...

[pastel skiff]

Thanks cascade... got MU running, it is giving me an error ValueError: Pin PA30 in use when I try to turn on the "on board" speaker with spkrenable = digitalio.DigitalInOut(board.SPEAKER_ENABLE)

looking at the pinout PA30 seems to be a flash mem pin, not sure why that would be conflicting

[upbeat plover]

neopixel doesnt work for hallowing m0?

neopixel_write?

[raven canopy]

@pastel skiff when you from adafruit_circuitplayground.express import cpx, it enables the speaker when it initializes. see here: https://github.com/adafruit/Adafruit_CircuitPython_CircuitPlayground/blob/master/adafruit_circuitplayground/express.py#L97

also, according to the pin definitions in the firmware, SPEAKER_ENABLE is mapped to PA30.

[upbeat plover]

GRB order...

[raven canopy]

and looking at the PigHixx pin outs...that looks wrong. 🤷

[tawny creek]

@onyx hinge sounds good to me 😃

[upbeat plover]

import board
import busio
import digitalio
import time
from adafruit_rgb_display import color565
import adafruit_rgb_display.st7735 as st7735

spi = busio.SPI(clock=board.SCK, MOSI=board.MOSI, MISO=board.MISO)
display = st7735.ST7735(spi, cs=digitalio.DigitalInOut(board.TFT_CS),
dc=digitalio.DigitalInOut(board.TFT_DC), rst=digitalio.DigitalInOut(board.TFT_RESET))
backlight = digitalio.DigitalInOut(board.TFT_BACKLIGHT)
backlight.direction = digitalio.Direction.OUTPUT
backlight.value = True

while True:
    display.fill(color565(0, 0, 255))
    time.sleep(5)
    display.fill(color565(0, 255, 0))
    time.sleep(5)
    display.fill(color565(0, 255, 0))
    time.sleep(5)

how do you get the display to work on hallowing m0? this isnt work

[meager fog]

tft test

@upbeat plover ^

[tawny creek]

@tulip sleet @raven canopy fiiiigured it out ❤ !

@raven canopy that comment on sercom was right.. thank you!!

[manic glacier]

turbinenreiter

[adafruit/circuitpython] New comment on pull request #1114: WIP: Add German translations

This is a lot of work and it's quite hard to find fitting translations. If anyone want's to pitch in, be my guest. Announce yourself and start from the bottom.

[upbeat plover]

Does Mu work with hallowing m0? the serial cant find it

[manic glacier]

dhalbert

[adafruit/circuitpython] New comment on issue #996: Add board files for PCA10059 \- the nRF52840 dongle

I am including this in a revamp of pin definition files and also #1040.

dhalbert

[adafruit/circuitpython] New comment on issue #1040: nrf: microcontroller/Pin\.c does not handle pin claiming

Closing in favor of #1089, which includes this.

dhalbert

[adafruit/circuitpython] Issue closed: #1040 nrf: microcontroller/Pin\.c does not handle pin claiming

[raven canopy]

@tulip sleet you around?

[upbeat plover]

@meager fog that test only worked after changing the size of the screen to 130x130 from 128x 128 , with 128 x 128 it gave it a boarder

[raven canopy]

@tidal kiln or @solar whale, would either of you be willing to test ladyada's PR for DHT on the RPi? My RPi hasn't been powered on in like 2 years, so I'm sure it would take like a week for me to get to testing. 😄 https://github.com/adafruit/Adafruit_CircuitPython_DHT/pull/14

[solar whale]

@raven canopy sure -- I'll give it a try.

[tidal kiln]

Thanks. Im AFK til Monday.

[raven canopy]

sweet. you might as well do the review too, if you're up to it. my only concern is the assumption that the DHT will respond with the start bit according to plan. we've seen how it likes to do that. 😄

[solar whale]

gathering bits and pieces now

[tulip sleet]

@raven canopy I'm here but was afk earlier

[raven canopy]

@tulip sleet its all good. just wanted to verify something while reviewing a PR. slicing seems to be universally available now, yes? i vaguely remember non-Express builds not having it...

[tulip sleet]

Just checked: MICROPY_PY_ARRAY_SLICE_ASSIGN is on for all our boards

also MICROPY_PY_BUILTINS_SLICE_ATTRS is on as well

[raven canopy]

yep. that was what I looked at. and, of course, i just remember that we have history with git...novel idea. 🤣 https://github.com/adafruit/circuitpython/commit/15dcc7011cf871c32828d1b28d520b210d47cc90#diff-b1656db6ed8d6cf4907e7f2fc300833d

[tulip sleet]

i was not sure about nrf but it is indeed on there too

[raven canopy]

i didn't see it on nRF, but that's still in work so...eventually i'm sure.

ahh...the SLICE_ATTRS works too? i wasn't sure about that.

[tulip sleet]

ports/nrf/mpconfigport.h
100:#define MICROPY_PY_BUILTINS_SLICE_ATTRS          (1)

ports/nrf/mpconfigport.h
98:#define MICROPY_PY_ARRAY_SLICE_ASSIGN            (1)

default values we override to (1)
from py/mpconfig.h

// Whether to support slice subscript operators and slice object
#ifndef MICROPY_PY_BUILTINS_SLICE
#define MICROPY_PY_BUILTINS_SLICE (1)
#endif

// Whether to support slice attribute read access,
// i.e. slice.start, slice.stop, slice.step
#ifndef MICROPY_PY_BUILTINS_SLICE_ATTRS
#define MICROPY_PY_BUILTINS_SLICE_ATTRS (0)
...
// Whether to support slice assignments for array (and bytearray).
// This is rarely used, but adds ~0.5K of code.
#ifndef MICROPY_PY_ARRAY_SLICE_ASSIGN
#define MICROPY_PY_ARRAY_SLICE_ASSIGN (0)
#endif

[solar whale]

@raven canopy slick - DHT worked "out of the box"

getsoccasional invalid reading -- nothing new

[raven canopy]

@tulip sleet that clears that up. Thanks!

@solar whale cool. thanks for testing. you want to do the review? you did all the work... 😄

[solar whale]

sure -- after a few more tests. it just stopped working 😦 -- may be poor connections

3V pulled out 😉

[raven canopy]

that darn 3V. always the rebellious one..

[solar whale]

should I alos test on - non-Pi since the changes are for all, correct?

oh - I see -- the ifdefs only come into play for the RPI

[raven canopy]

wouldn't hurt. it looks solid to me though.

[solar whale]

It does crash occasionally -- I'll comment inthe review

[meager fog]

the pi stuff is hacky

i am doing more research on faster GPIO, it requires us using libgpiod

which is non-trivial, but we'll do it

but other non-pulseio things that can sample at ~100KHz will be good

@solar whale by chance have you ever put a library on readthedocs?

i wanted to put PN532 dox up but im not sure how to make it appear under 'circuitpython'

[solar whale]

@meager fog not that I am aware of...

[meager fog]

all good ill ask scott when he's around

[raven canopy]

yeah, Scott is the one to talk to about creating the sub-project on RTD. but, you can go ahead and add it here...it just won't build. https://github.com/adafruit/circuitpython/blob/master/docs/drivers.rst

[solar whale]

@meager fog the bad readings are not a big problem, but I am puzzeld why it crashes with ```Traceback (most recent call last):
File "<stdin>", line 1, in <module>
File "/home/pi/projects/blinka/dht/dht_simpletest.py", line 11, in <module>
temperature_c = dhtDevice.temperature
File "/home/pi/projects/blinka/dht/adafruit_dht.py", line 213, in temperature
self.measure()
File "/home/pi/projects/blinka/dht/adafruit_dht.py", line 164, in measure
pulses = self._get_pulses()
File "/home/pi/projects/blinka/dht/adafruit_dht.py", line 149, in _get_pulses
pulses.append(int(1000000 * (transitions[i] - transitions[i-1])))
OverflowError: unsigned short is greater than maximum

[meager fog]

hm

what is the type of 'pulses'

[solar whale]

array.array('H')

that'll do it

so this will fail any time (transitions[i] - transitions[i-1]) is nonzero

[manic glacier]

ladyada

[adafruit/circuitpython] New comment on issue #1101: need driver for PN532

https://github.com/adafruit/Adafruit_CircuitPython_PN532

ladyada

[adafruit/circuitpython] Issue closed: #1101 need driver for PN532

[meager fog]

woot https://pypi.org/project/adafruit-circuitpython-pn532/

@solar whale could also be the transitions are over 65535 microseconds

want to add a patch/check to 'max' it to 65535

[solar whale]

@meager fog oh -- I see - transitions is a float ( time.monotonic) I thought it was in integer

[meager fog]

even 10,000 is max

we turn it into ints, but they should be like 20-100 us

[solar whale]

I see taht now -- OK -- if it is > 65536, what error should be thrown?

another RuntimeError("Invalid transition length detected") or something like that

[meager fog]

well it wil fail later

i think just max it at 65535

[solar whale]

ok

@meager fog -- seems to be running OK -- still lots of bad checksums and 'full buffer not returned' but no integer overflows....

[meager fog]

yeah

thats cause we're sampling GPIO by hand, its 'fun' and 'exciting' and 'totally unreliable' 😃

and on pi zero does not work at all

even with libgpiod it wont work

so long term i do have to find some other solution, but at least for pi 3 this will work

[solar whale]

at tleast thats something ...

this was the change I made ```
diff --git a/adafruit_dht.py b/adafruit_dht.py
index f795f4d..928c50f 100644
--- a/adafruit_dht.py
+++ b/adafruit_dht.py
@@ -146,7 +146,11 @@ class DHTBase:
transitions.append(time.monotonic()) # save the timestamp
# convert transtions to microsecond delta pulses:
for i in range(1, len(transitions)):

•                pulses.append(int(1000000 * (transitions[i] - transitions[i-1])))
  

•                pulses_micro_sec = int(1000000 * (transitions[i] - transitions[i-1]))
  

•                if pulses_micro_sec > 65535:
  

•                    pulse.append(65535)
  

•                else:
  

•                    pulses.append(pulses_micro_sec)
     return pulses
  

  def measure(self):

oops need to fix it no need to recompute - fixed

@meager fog I'm not quite sure how to submit a patch to your PR.. if you want me to do that

I'm happy to learn how, but it may be faster and simpler for you to just implement yourself when you have time.

I have to go AFK for awhile -- I can follow up on this later tonight or tomorrow.

[meager fog]

oh didnt see this

i just did it myself

i just cant repro the problem, i dont get it on my pi?

[solar whale]

I have some oter stuff running on mine -- may cause some delays....

@meager fog before the fix, it sometimes ran for a minute or so before the overflow error.

[meager fog]

oh good point lemme run for a while

[solar whale]

I have some I2c (reading sht31) and adafruit.io stuff going on as well

[meager fog]

cool , blinka stuff all working out so far?

sht31 is a pretty stable i2c sensor

[solar whale]

yez - works great - I just wondered if the background processes were causing some delays fpr the DHT taht you are not seeing.

[meager fog]

yep totally

that will cause oddness. i think libgpiod will be more stable there too, cause its a kernel call

[solar whale]

sounds good -- BTW -- I am seeing some bogus readings that "slip" through occasionally

[magic sky]

Is there any plan to make framebuf.mpy work with CircuitPython 3?

[meager fog]

jp, we're looking to possibly replace framebuf altogether in circuitpython4

[meager fog]

@gentle bronze hiya just so you know i have some time now for nrf52 so i am re-doing the guide for programming it into various boards and make it public - i am also going to test all the bootloader stuff!

[manic glacier]

ladyada

[adafruit/circuitpython] New comment on issue #1129: NRF52 build fails on MACOS

oof i hit this building on windows too

[magic sky]

Thanks for the reply milady (corny, I know. won't happen again.) I had a dream of a gemma m0 and a DHT and an OLED with circuitpython. Alas, it's not to be. From what I understand, even with framebuf working, there's no character output, but I was still playing with the OLED and ran into the compatibility issue. Unfortunately, my skills at this time aren't up to rectifying the issue on my own.

[meager fog]

jp, yah the gemma m0 with circuitpython may not be able to do it - but try arduino!

it will work for sure with l0ts of room

[raven canopy]

@meager fog are you building nRF in a vagrant?

[meager fog]

no, win10

[raven canopy]

k. i'm trying it right now inside the VM... 🤞

[meager fog]

it used to work in msys2 - but now doesnt :/ so doing win 10 linux

ill let you know how that goes

im also writing/documenting the process 😃

are u also on win?

[raven canopy]

yep

[meager fog]

7 or 10?

[raven canopy]

10

well, that was quick:

FREEZE freeze
In file included from supervisor/port.c:32:0:
./common-hal/microcontroller/Pin.h:33:23: error: unknown type name 'pin_obj_t'
 #define mcu_pin_obj_t pin_obj_t
                       ^
./common-hal/pulseio/PWMOut.h:36:11: note: in expansion of macro 'mcu_pin_obj_t'
     const mcu_pin_obj_t *pin;
           ^~~~~~~~~~~~~
../../py/mkrules.mk:55: recipe for target 'build-pca10056-s140/supervisor/port.o' failed
make: *** [build-pca10056-s140/supervisor/port.o] Error 1

[meager fog]

yeppp

look at us! error twinsies 😃

[raven canopy]

as usual, jerryn was the first... ⚒

i was just about to start some nRF this week too... 😦

[meager fog]

well, lemme try it in win 10 linux

[raven canopy]

k. i'm looking up the vagrant equivalence to Scott's macOS magic.

[meager fog]

i definitely see the problem

theres Pin.h and pin.h

[raven canopy]

yep. case-sensitive file system is what Scott pointed out. which locks out Windows entirely (in my reading so far).

[meager fog]

https://www.howtogeek.com/354220/how-to-enable-case-sensitive-folders-on-windows-10/

did ya try that by chance?

meanwhile ill try renaming pin.h

i think renaming worked

if you dont mind trying my fork

[raven canopy]

did you just rename it to "Pin.h", and change the #include in common-hal/microcontroller/Pin.h?

[meager fog]

nonoo

i renamed it to nrf_pin.h

one moment tho, its failing in another place 😃

[raven canopy]

yeah...that makes more sense. derr. 😄

[meager fog]

https://github.com/ladyada/circuitpython/commit/ded38d30b6bd9621ab5254f849f131946dfccbf4

[raven canopy]

and that fsutil trick to turn on case-sensitive could work. but...only applies per folder, not subfolder. that could get unwieldy.

[meager fog]

you can try the ^ lemme know does that help

[raven canopy]

running it now.

[meager fog]

yay for me it helped

[manic glacier]

ladyada

[adafruit/circuitpython] Pull request opened: #1145 rename pin\.h to nrf\_pin\.h to avoid conflict with Pin\.h

fix for https://github.com/adafruit/circuitpython/issues/1129 on windows

[raven canopy]

success in VM as well!

FREEZE freeze
LINK build-pca10056-s140/firmware.elf
   text    data     bss     dec     hex filename
 226728    1272   15476  243476   3b714 build-pca10056-s140/firmware.elf
arm-none-eabi-objcopy -O binary build-pca10056-s140/firmware.elf build-pca10056-s140/firmware.bin
arm-none-eabi-objcopy -O ihex build-pca10056-s140/firmware.elf build-pca10056-s140/firmware.hex
Create firmware.uf2
../../tools/uf2/utils/uf2conv.py -f 0xADA52840 -c -o "build-pca10056-s140/firmware.uf2" build-pca10056-s140/firmware.hex
Converting to uf2, output size: 456192, start address: 0x26000
Wrote 456192 bytes to build-pca10056-s140/firmware.uf2.

[manic glacier]

sommersoft

[adafruit/circuitpython] New comment on pull request #1145: rename pin\.h to nrf\_pin\.h to avoid conflict with Pin\.h

Ran this myself. Works on Win10 with a Vagrant VM. 🎉

tannewt

[adafruit/circuitpython] New comment on pull request #1114: WIP: Add German translations

@turbinenreiter Thanks for you hard work on this! How about taking a break, fixing the build and getting what you have so far checked in? Thanks!

[meager fog]

@raven canopy yay look at us!

ok im gonna redo the bootloader page

and then once that works, get back to main flashing

blah bootloader is on but isnt enum

@raven canopy are you using an nrf52840 DK?

oh weird it works if i run the other command

🤷

[magic sky]

@meager fog I know I can do it with arduino. That's my blink for new boards. I was just wanting to play with the circuitpython to see if it would squeak in, then it was just seeing if would do anything with the oled at all, the sample 3 pixel program for example. But no sweat. One more tiny question and I'll leave you alone for a week whilst you tend to more important stuff. Is it possible to use an OLED (since I haven't mentioned it before, the ssd1306 variety) with a circuit playground express mounted on a Crickit? I've fiddled with it a bit with no luck and am not sure if the special build prohibits it.

[meager fog]

i think we took framebuf out of that build too!

:/ you may want to try a Feather M4 on a featherwing crickit

you'll be golden there

samd21 (gemma/CPX) is a little constrained

[raven canopy]

@meager fog was away. yeah, using the DK. i haven't done anything with it yet though. I still need to walk through that README to burn the bootloader.

[magic sky]

That's cool. I'll find other projects. Thanks.

[meager fog]

@raven canopy want me to add u to this guide

you can check it out

[raven canopy]

i'd be happy to take a peek!

[meager fog]

https://learn.adafruit.com/admin/guides/2074/editor/12646

first 2 pages are good

[raven canopy]

hehe. it started me on page 2...didn't even realize it. 😄

@meager fog on "Syncing Your Fork", is advising git pull --rebase intended? i'm still noobish with git, but rebasing has usually caused me headaches. 🤕

[meager fog]

yeah i use this https://stackoverflow.com/questions/41283955/github-keeps-saying-this-branch-is-x-commits-ahead-y-commits-behind

to remind myself how to sync forks

[raven canopy]

👍 k. just checking that it wasn't an errant paste. 😄

[meager fog]

@gentle bronze hmm i think our adafruit-nrfutil needs a new release?

$ make BOARD=pca10056 SERIAL=COM3 dfu-flash Use make V=1, make V=2 or set BUILD_VERBOSE similarly in your environment to increase build verbosity. adafruit-nrfutil --verbose dfu serial --package build-pca10056-s140/dfu-package.zip -p COM3 -b 115200 --singlebank --touch 1200 Error: no such option: --touch make: *** [Makefile:333: dfu-flash] Error 2

[manic glacier]

dhalbert

[adafruit/circuitpython] New comment on issue #1129: NRF52 build fails on MACOS

I am redoing the pin files and this will go away as a side effect of my changes (or I will make sure it goes away).

[raven canopy]

@meager fog for the USB step/picture, might mention the nRF power source slider switch. mine came in the VDD position, but it should be on the USB position, no?

[manic glacier]

dhalbert

[adafruit/circuitpython] Pull request closed: #1145 rename pin\.h to nrf\_pin\.h to avoid conflict with Pin\.h

dhalbert

[circuitpython:master] 2 new commits

ded38d3 rename pin.h to nrf_pin.h to avoid conflict wit... - ladyada
9a10849 Merge pull request #1145 from ladyada/master - dhalbert

[meager fog]

@raven canopy the switch is in the center for me

and that works fine

[manic glacier]

ladyada

[adafruit/circuitpython] New comment on pull request #1145: rename pin\.h to nrf\_pin\.h to avoid conflict with Pin\.h

yay thx!

[raven canopy]

hmm. well, not what i expected. 🤷

[meager fog]

did you get farther?

ok i got thru all the steps

https://learn.adafruit.com/circuitpython-on-the-nrf52/prepare-to-build?preview_token=dLeOvimZKk0sF-y2V0KBAw

i found a few hiccups but nothing killer

[tulip sleet]

@raven canopy yeah, "USB" is what you'd think, but that's not what it is

[manic glacier]

dhalbert

[adafruit/circuitpython] New comment on issue #1007: USB serial

@hathach both @ladyada and i can't get to the REPL from Windows, LInux or Mac, when loading the latest master on pca10056. Is this supposed to work yet? Serial connects but doesn't respond.

[tulip sleet]

also the nrf-only/default switch on the upper right of the board should be on "default". I switched mine by accident once, and had a heck of a time figuring out what was wrong

[raven canopy]

i can only guess that it turns off the JLink?

[tulip sleet]

it seemed to shut down everything for me -- confusing

[raven canopy]

i didn't even realize the DK had one built-in until looking at this guide

[tulip sleet]

but maybe not

[raven canopy]

is the tarball the only way to install nrfjprog on linux? i think i'm going to have to passthough to the vagrant; can't get to building the hex to copy over to the windows side.

[tulip sleet]

@raven canopy that's how I installed it

@gentle bronze @meager fog made a new release for adafruit-nrfutil 0.5.3.post9, should be in pypi shortly. The version number was already bumped - I just created the github release

[raven canopy]

ohh...wait. there is a genhex flag in the makefile...

foiled. still need nrfjprog... mergehex: command not found.

[gentle bronze]

@meager fog yeah, let's me know if I could help in any ways with your testing.

@tulip sleet thanks for releasing the post9 for adafruit-nrfutil

[tulip sleet]

sure - i need to build the executables soon and upload them too, but maybe not tonight

you just got up, right?

[manic glacier]

hathach

[adafruit/circuitpython] New comment on issue #1007: USB serial

Hmm, weird, I don't have any issues connecting on linux, only a bit of delay for the first connection. Which is the previous known bug. Maybe it is time to troubleshoot it as well.

[meager fog]

@gentle bronze good morning!

yes i updated this guide, i can release it to get more people trying out nrf52840 on DK

[gentle bronze]

@meager fog good morning 😄

[meager fog]

https://learn.adafruit.com/circuitpython-on-the-nrf52/prepare-to-build?preview_token=dLeOvimZKk0sF-y2V0KBAw

that way you don't have to update readme with how-to

[gentle bronze]

"Download Nordic SDK and SoftDevice files," --> should be "Download Nordic SoftDevice files" only

sicne we don't download anything from SDK, only SD file

[meager fog]

ok no problem i will fix it now 😃

ok fixed

I tried UF2, it works great!

good work with it 😃

[gentle bronze]

pip3 install --upgrade adafruit-nrfutil --> you probably mean --user instead of --upgrade 😄

[meager fog]

ok

no i mean --upgrade

that way it gets the latest

i can add --user too?

[gentle bronze]

I am not sure @tulip sleet know it better 😄

[meager fog]

ok i ll do both

[raven canopy]

that installs in your user path, versus the system path. iirc...

[meager fog]

it doesnt work as well for most people tho

[gentle bronze]

uf2 works even more reliably than good old cdc, I am thinking to allow it to upgrade bootloader + SD as well, with a set of requirement https://github.com/adafruit/Adafruit_nRF52_Bootloader/issues/24

[raven canopy]

yeah, it can get dicey. all about permissions...

[meager fog]

ok great i am on post9 now

[raven canopy]

nrfjprog + vagrant != happy.

[meager fog]

yeahh

you cant really do that, you should run nrfjprog from msys or command

[raven canopy]

hehe. thats the rub. i can't get the hex to build without nrfjprog...and i (currently) can't make on the windows side. 😄

[meager fog]

@gentle bronze ok great i checked and tested --touch 1200 and it works

so that part is fixed

[gentle bronze]

@meager fog great, did you test while it is in CP ( without manually enter bootloader mode )

[meager fog]

no let me try now

it should auto-reset into bootloader mode?

i can try with a plain terminal first

oh you know, i could not get CDC working in CIRCUITPY mode at all

do you want a USB trace? im on windows 10

[gentle bronze]

yeah, that is weird, it works last time, I didn't make many changes to circuitpyhon lately, only usb. I will troubleshoot it today

yeah, it should auto reset into bootloader mode.

[meager fog]

yah i cannot open the port at all

:/

let me know if you want USB trace i have a Beagle

[manic glacier]

dhalbert

[adafruit/circuitpython] New comment on issue #1007: USB serial

On Linux, if I type ^C at the connection before it responds (which takes at least 6 seconds), then the terminal program exits without showing any output. I think maybe that's why I thought it wasn't connecting.

On Mac, with screen, if I pull the USB cable out after waiting 10 seconds or so, then the initial REPL string appears when screen exits.

On Windows, the COM port appears immediately in the Device Manager, but does not seem to be available for a number of seconds after resetti...

[gentle bronze]

@meager fog I will troubleshoot it, I got an usb analyzer from Lecroy as well. It should do

[meager fog]

ok i know it worked before so hopefully nothing complicated 😃

[gentle bronze]

hopefuly 😄

[manic glacier]

hathach

[adafruit/circuitpython] New comment on issue #1007: USB serial

@dhalbert thanks for analyzing it, I will try to fix it asap

dhalbert

[adafruit/circuitpython] New comment on issue #1007: USB serial

It seems erratic, which may be due to the delay and the host becoming confused.

[tulip sleet]

@gentle bronze would it have anything to do with the softdevice update? is the softdevice part of the .uf2 image? or I need to upload it separately? Maybe we should make a .uf2 for the softdevice?

[gentle bronze]

@tulip sleet possibly not related to softdevice upgrade. I guess. I will troubleshoot and got you update.

[tulip sleet]

yah, seems unlikely

[gentle bronze]

uf2 image currently does not contain SD, and uf2 msc couldn't upgrade softdevice. It is discussed a while ago since we afraid it may break the OTA feature or easily brick device with incorrect uf2 file.

it is before uf2 introduced the family option, it is much safer now

I open an issue for it https://github.com/adafruit/Adafruit_nRF52_Bootloader/issues/24 here

[meager fog]

ok all the stuff i was doing was so i could test https://github.com/adafruit/Adafruit_nRF52_Bootloader/issues/7#issuecomment-415735906

@gentle bronze can you tell me how to test the driver above?

i mean, what should i do to make sure that when i add the driver it did the right thing 😃

[gentle bronze]

@meager fog you can try to upload sketch with arduino, I just got it worked last friday

[meager fog]

do i need the touch-1200 to work?

sorry there's so much 😄

[gentle bronze]

yes, although CDC will show up, and you could upload, the Serial monitor is not redirect to it yet

just upload any sketch to test with 😄

[meager fog]

ok with BSP is it?

which

[gentle bronze]

the develop branch https://github.com/adafruit/Adafruit_nRF52_Arduino/tree/develop

[meager fog]

ok let me change to develop now!

by the way i am using PC10056 but this week i will test the feather nf52

once i can test peripherals/pins 😃

[gentle bronze]

👍

[meager fog]

select this?

[gentle bronze]

yeah

[gentle bronze]

that one, we will remove it later when having the final feather

[meager fog]

sure no problem 😃

'develop' 😄 😄 😄

we can keep it, the DK is handy!

[gentle bronze]

it is up to you ;D

[meager fog]

ok ill try simple blink sketch let me look at variant file

C:\Users\ladyada\AppData\Local\Temp\arduino_build_153447/core\core.a(usb.c.o): In function usb_init':

C:\Users\ladyada\Dropbox (Personal)\ArduinoSketches\hardware\Adafruit\Adafruit_nRF52_Arduino\cores\nRF5\usb/usb.c:81: undefined reference to `nrfx_power_init'

C:\Users\ladyada\Dropbox (Personal)\ArduinoSketches\hardware\Adafruit\Adafruit_nRF52_Arduino\cores\nRF5\usb/usb.c:85: undefined reference to `nrfx_power_usbevt_init'

C:\Users\ladyada\Dropbox (Personal)\ArduinoSketches\hardware\Adafruit\Adafruit_nRF52_Arduino\cores\nRF5\usb/usb.c:87: undefined reference to nrfx_power_usbevt_enable'

hmm

do i have to do a submodule for nrfx?

[gentle bronze]

op sorry, I may miss some file

[meager fog]

no problem i will wait!

let me test nrf52832 on Macbook in the mean while

i tested on windows but not mac

with the new BSP

[gentle bronze]

just pushed the missing file, please try again. We upgrade some of the hal/driver to nrfx

[meager fog]

ok let me try again

[gentle bronze]

but only a small subset and will include it directly into the core folder. Couldn't do the submodule thing, since Arduino will try to compile everything

[meager fog]

ok compiled!

what pin # would make an LED blink?

so i can know its running 😃

[gentle bronze]

on pca10056, it will blink led1 as LED_RED, and led2 as LED_BLUE

[meager fog]

kk

[gentle bronze]

you can try upload bleuart, it should work as well.

[meager fog]

ok good news it works

i just had to manually start the bootloader with Button1+reset

it enumerated as VID 0x239A PID 0x8029

i can open the COM port here, which is odd, because circuitpython doesnt

[gentle bronze]

it has something with CP, I may miss some glues between cp and usb

probably background task is not running at the right place

[meager fog]

all good

ok no data and also i notice in arduino it doesn't use touch

[gentle bronze]

can you try to upload another sketch to that serial port while it is still in arduino sketch

[meager fog]

so it doesnt auto-update

?

[gentle bronze]

Arduino has builtin 1200 touch

[meager fog]

hmm

Global variables use 15092 bytes (29%) of dynamic memory, leaving 35980 bytes for local variables. Maximum is 51072 bytes.
Forcing reset using 1200bps open/close on port COM5
C:\Users\ladyada\Dropbox (Personal)\ArduinoSketches\hardware\Adafruit\Adafruit_nRF52_Arduino/tools/adafruit-nrfutil/win32/adafruit-nrfutil.exe --verbose dfu serial -pkg C:\Users\ladyada\AppData\Local\Temp\arduino_build_153447/Blink.ino.zip -p COM5 -b 115200 --singlebank 
Upgrading target on COM5 with DFU package C:\Users\ladyada\AppData\Local\Temp\arduino_build_153447\Blink.ino.zip. Flow control is disabled, Single bank, Touch disabled
Opened serial port COM5
Starting DFU upgrade of type 4, SoftDevice size: 0, bootloader size: 0, application size: 49944
Sending DFU start packet
Timed out waiting for acknowledgement from device.

Failed to upgrade target. Error is: No data received on serial port. Not able to proceed.
Traceback (most recent call last):
  File "nordicsemi\__main__.py", line 294, in serial
  File "nordicsemi\dfu\dfu.py", line 232, in dfu_send_images
  File "nordicsemi\dfu\dfu.py", line 200, in _dfu_send_image
  File "nordicsemi\dfu\dfu_transport_serial.py", line 177, in send_start_dfu
  File "nordicsemi\dfu\dfu_transport_serial.py", line 237, in send_packet
  File "nordicsemi\dfu\dfu_transport_serial.py", line 276, in get_ack_nr
nordicsemi.exceptions.NordicSemiException: No data received on serial port. Not able to proceed.```

let me try with x-ctu

[gentle bronze]

while it is in dfu cdc only mode

you need to install nother driver for it

the CDC only driver

[meager fog]

right now im in win10 so i shouldnt have to

but i cant seem to get it to reset with touch-1200

i opened the terminal @ 1200 baud and nothing happened 😃

[gentle bronze]

open and disconnect

[meager fog]

yeah it doesnt go away

i can open the terminal @ 1200 bps, close it, COM5 is still there 🤷

do you want usb trace?

i bought a serial tracer software too

[gentle bronze]

yeah if you could, the os may attempt to set it back to 9600, I saw it with minicom

can you try with the adafruit-nrfutil

maybe using the make BOARd=pca10056 dfu-flash in circuitpython

[meager fog]

hmm it looks like it opens it at a higher baud rater

then changes it?

let me look at the trace

[gentle bronze]

If it doenst work as well, I will test more with winodws, only get it worked last friday, didn't test it with windows or mac though 😦

the code will check the baudrate at the time of disconnection

[meager fog]

no dfu-flash aslso didnt reset it

[gentle bronze]

😦

[meager fog]

$ make BOARD=pca10056 SERIAL=/dev/ttyS4 dfu-flash
Use make V=1, make V=2 or set BUILD_VERBOSE similarly in your environment to increase build verbosity.
adafruit-nrfutil --verbose dfu serial --package build-pca10056-s140/dfu-package.zip -p /dev/ttyS4 -b 115200 --singlebank --touch 1200
Upgrading target on /dev/ttyS4 with DFU package /c/Users/ladyada/Dropbox (Personal)/micropython/circuitpython/ports/nrf/build-pca10056-s140/dfu-package.zip. Flow control is disabled, Single bank, Touch 1200
Touched serial port /dev/ttyS4
Opened serial port /dev/ttyS4
Starting DFU upgrade of type 4, SoftDevice size: 0, bootloader size: 0, application size: 231264
Sending DFU start packet
Timed out waiting for acknowledgement from device.

COM5 is still here 😄

its a tough COM port, it wont back down!

[gentle bronze]

😄

give me a bit of time, I will test it more properly with windows and mac 😄

[meager fog]

ok no problem

im going to 💤

[gentle bronze]

good night

[meager fog]

many things worked, so dont worry about it

we'll figure it out

we'll get it working again!

please let me know when you're ready for more testing!

[gentle bronze]

I am will ping you when we are ready, also please update the driver of the new PID 0x002A # Arduino bootloader (CDC)

as well 😄

see you tomorrow ( or eveninig :D)

[meager fog]

ok i will do that when i get to work tomorrow since i have my windows 7 machine there

and it uses drivers

[gentle bronze]

👍

[raven canopy]

giving up for the night... nrfjprog: 1, me: 0. 😄

[gentle bronze]

@raven canopy the mergehex is also part of the nrfjprog, did you see it after downloaded it

[manic glacier]

ScottEP

[adafruit/circuitpython] Fork created: ScottEP/circuitpython

[raven canopy]

@gentle bronze yeah. i'm working inside Vagrant/VM. i've got nrfjprog's path set; need to do the same for mergehex. or...just use SES. 😄

[gentle bronze]

@raven canopy yeah, SES is more convenient 😄

[manic glacier]

tannewt

[adafruit/circuitpython] Pull request closed: #1143 Fix os\.stat\(\) epoch and use RTC for file timestamp

tannewt

[circuitpython:master] 3 new commits

39ee12d Fix os.stat() to use 1970 epoch - notro
b61cf8d samd: Use RTC to set file timestamp - notro
569a050 Merge pull request #1143 from notro/stat_epoch_fix - tannewt

[manic glacier]

hathach

[adafruit/circuitpython] New comment on issue #1007: USB serial

analyzer show a huge delay for set baudrate request, after the serial is connected(). I guess the serial_connected() return true, trigger some function that didn't run the background task to handle the baudrate request !!!. Linux seems to be more patient, windows and mac seems to timeout and drop connection. Looking more closely at it now

[gentle bronze]

@raven canopy I seperated mergehex from genhex, you should be able to compile now. merghex only needed for generated pre-built binaries. Shouldn't be part of that genhex 😃

[gentle bronze]

@tulip sleet nrf52 build failed with DEBUG=1, due to the printf not implemented

CC ../../main.c
../../main.c: In function '__assert_func':
../../main.c:454:5: error: implicit declaration of function 'printf'; did you mean 'mp_printf'? [-Werror=implicit-function-declaration]
printf("Assertion '%s' failed, at file %s:%d\n", expr, file, line);
^~~~~~
mp_printf

should we add nanolib for debug build ?

[indigo wedge]

@gentle bronze you need to add include <stdio.h> in main.c

[gentle bronze]

@indigo wedge printf may need to have a bit of work to retarget it with _write() and _read() as well ? Did we implement those hook in nrf52 port ?

I just comment it out for now to build with debug=1 😄

[indigo wedge]

I think that's done by the circuitpython/micropython code because all I had to do to make it build and work was to add the include.

[manic glacier]

hathach

[adafruit/circuitpython] New comment on issue #1007: USB serial

@dhalbert @ladyada
More on troubleshoot. If we comment out these message function in main.c::run_code_py(), it will connect instantly. My guess is the below code try to write data to the cdc before the above set baudrate request is complete. it is still valid since the LINE STATE request is complete indicating terminal is connected.

Due to the nrf's usb easy DMA resitrction that only 1 DMA transfer can be active at a time, this seems to block the control response which is serve i...

[gentle bronze]

@indigo wedge ah you are right, maybe @tulip sleet knows where it is hidden 😄

[fluid helm]

EduBlocks for Circuit Playground using Circuit Python is shaping well! @slender iron @idle owl @tulip sleet

[tulip sleet]

@gentle bronze You can replace printf(...) with mp_printf(&mp_plat_print, ...). Some files have a macro to do that, I think.

[fluid helm]

is there a known bug with uploading files to CPX in Ubuntu?

The same code uploads in Windows but not in Ubuntu

Scott mentioned something about a serial mode that could stop it working at PyCon?

Can't remember what I need to to do come out of that mode

[tulip sleet]

@tall yacht could you be more specific about what you're doing?

could you point to code that's not working?

[gentle bronze]

thanks @tulip sleet , since it is locatedin main.c ,I will leave it to you since it will affect all the ports https://github.com/adafruit/circuitpython/blob/master/main.c#L454

[fluid helm]

Ahoy @tulip sleet , So what I'm trying to do is upload main.py to the CIRCUITPY drive. I'm trying to upload the basic neopixel example ( https://github.com/adafruit/Adafruit_Learning_System_Guides/blob/master/Introducing_CircuitPlaygroundExpress/CircuitPlaygroundExpress_NeoPixel.py )

The file uploads and runs fine when it's uploaded in Windows

but when I upload the file in Ubuntu, doen't run 😦

[tulip sleet]

are you just uploading it manually, or do you mean code in EduBlocks is uploading it?

@fluid helm looks like you dropped off!

[fluid helm]

Nope, I'm doing it with Mu

Sorry I went, I broke my websever 😄 @tulip sleet

[tulip sleet]

@fluid helm when you do the upload, does anything show in the REPL? It won't run automatically if the REPL is open and at the >>> prompt.

[fluid helm]

Nope, nothing. The program runs when it's connected to a PC running windows, but as soon as it's connected to Ubuntu it wants to do nothing @tulip sleet

however, when i manually type stuff into the REPL, when i've opened the REPL, it does stuff

[tulip sleet]

i'll try it here

[fluid helm]

My laptop is weird

[tulip sleet]

work for me, but with the REPL ("Serial") closed

if Serial is open, then I have to type ctrl-D at the >>> prompt

[fluid helm]

Odd, even when I do it without Mu open and via the text editor

[tulip sleet]

which text editor are you using?

[fluid helm]

ooooooo

CTRL+D fixes it

looks like the REPL auto loads for some reason on my PC?

[tulip sleet]

main.py / code.py won't run automatically if you're sitting at the >>> prompt.

[fluid helm]

Ah, ok. Weird as it does it even when Mu has never been opened

[tulip sleet]

do you have a repl connection via screen or picocom or something like that?

[fluid helm]

Nope

[tulip sleet]

i'm starting up the win10 laptop; will test that with Mu

[fluid helm]

Cool, at least I know how to get it working if it doesn't run now. Thanks Dan!

[tulip sleet]

@fluid helm I don't see a difference in behavior between win10 and Ubuntu 18.04, but certainly do file an issue if you have a scenario that causes the issue. Happy to help!

[fluid helm]

Cool, it might just be my laptop, it plays up a lot with this sort of stuff. @tulip sleet

[plucky flint]

@fluid helm what version of Mu are you using?

[fluid helm]

V 1.0.0

[plucky flint]

that's completely odd behaviour. BTW, I'm on Ubuntu 18.04 right now and it works for me. 😕

riiiiiight

have you tried rebooting your laptop? 😉

[fluid helm]

Yup, done that

[plucky flint]

@tulip sleet 👋

[fluid helm]

It's only this laptop that it plays up on

[plucky flint]

What do the (Mu) logs say?

[fluid helm]

Let me put it in a gist

it's too big for discord 😉

[tulip sleet]

@fluid helm and are you saying that even if mu is closed and you plug in the CPX, it does not start up automatically?

[fluid helm]

Yup @tulip sleet

[plucky flint]

But it works OK on Windows?

[fluid helm]

Yep

Mu isn't open so the log will be gone from before 😕

[plucky flint]

@fluid helm Mu saves the logs for the day. (If you look in the directory where the log files are, you'll actually see the past 5 days worth of log files).

[tulip sleet]

yes, that is really odd. is it regular ubuntu 18.04, and what brand/model laptop is it?

[fluid helm]

https://gist.github.com/JoshuaLowe1002/43bd07e33cf03399804bc0c31837bf9a

There we go

[plucky flint]

👀

[fluid helm]

It's a HP 15-ac152sa @tulip sleet

[plucky flint]

OK... I see two things happening here....

[tulip sleet]

i'm not sure this is mu-related at all. are you in the dialout group? type groups at the shell?

[fluid helm]

Yup

I'm in dialout

It happens with the normal text editor too

so not possible mu related

but it seems to be something to do with the REPL

as when that gets CTRL + D 'ed it works

[plucky flint]

1. On some of the saves you save the file to /home/josh/mu_code/main.py (i.e. into your home directory on the laptop).

2. At the end, you appear to be saving to (the expected) /media/josh/CIRCUITPY/main.py file.

As I understand it, the CPX/CircuitPython detects a write to the filesystem on the device and restarts. Right @tulip sleet ?

[fluid helm]

Ah, that's odd. Sometimes it says "device not found". Does that make it save to /home/josh/mu_code/main.py?

[tulip sleet]

right, it waits 500ms after the last write (to wait for quiescence) and the does an auto-reload (equivalent to typing ctrl-D)

[plucky flint]

@fluid helm yes, as a safe fallback

[fluid helm]

So maybe that is what's happening

[tulip sleet]

yes, it seems to be hiccuping a lot not finding and then finding the CPX, and also finding it on ttyACM1 instead of ACM0

[plucky flint]

This doesn't sound like a Mu thing, rather a timing based issue (i.e. the device isn't yet registered as connected to your Ubuntu laptop)

@fluid helm you're typing TOO FAST. 😛

[fluid helm]

Ahahaha 😄

[tulip sleet]

are there other USB devices, and are you using a USB hub or going direct? USB3 or USB2 port

[plucky flint]

yeah... USB hubs (like @tulip sleet says) could also cause an issue.

[fluid helm]

USB hub 😄

[plucky flint]

I've seen that with micro:bit

aha

[tulip sleet]

take out the hub and go direct. Is it a powered or non-powered hub?

[fluid helm]

non powered

[tulip sleet]

other things plugged into it?

[fluid helm]

yup

[plucky flint]

OK.... try connecting the device directly to your laptop

[tulip sleet]

yes, you may be straining the power capability of the laptop port supplying the power (or it's just a junky hub)

[fluid helm]

Done that

direct works now without having to cancel the REPL

[plucky flint]

OK... so we probably have the source of the problem. No idea why the hub may be causing the problem (although @tulip sleet 's suggestion sounds sensible)

[tulip sleet]

what else was connected to the hub?

[fluid helm]

Argh, that's the thing with modern laptops with a lack of ports 😕

[tulip sleet]

try a powered hub if you have to use a hub

[fluid helm]

I remember when we used to get 5 in built USB ports

Cool, I think i've got one somewhere, shall try it @tulip sleet

Another question, is the weekly on today?

[plucky flint]

one day, I'll make another weekly, but dad-taxi duties always get in the way. 😕

[tulip sleet]

yes, 2pm US eastern time as usual

[prime flower]

@fluid helm shameless adafruit.com plug but I've been using this one we sell for a long time and it's pretty great for a powered hub: https://www.adafruit.com/product/961

(I have one of those macbooks with 2 ports)

[tulip sleet]

you can always use the discord phone app 😃

[plucky flint]

not sure connectivity from a muddy football pitch surrounded by trees is all that great. 😉

[fluid helm]

Oooo that looks good @prime flower

[tulip sleet]

parenting is higher priority

[fluid helm]

Football and muddy pitches is not for me. I'd rather be coding python that that 😄

[prime flower]

@plucky flint Has anyone done Mu on Raspi, which is also running blinka, yet?

I'm considering doing that for my newer blinka-guides

buuut I'm not really sure how the board detection would work, or if it'd even work at all.

[tulip sleet]

@prime flower but you mean with a board attached, or just using blinka and the raspi pins?

if no board, just blinka, you can switch it to "Python 3" mode. it doesn't have to have a board attached to work. Mode button in upper left

[fluid helm]

Any ideas of what I could use to replace the background on this? It's a bit fuzzy and doesn't really work well 😦

[prime flower]

from the board itself

@fluid helm not a UX designer but possibly apply a blur to the CPX a bit more?

add more tiling and then a blur. MakeCode uses a grey background that I kinda like, or "Adafruit Black" could also work.

[fluid helm]

maybe have it just as grey but have the taskbar another colour?

[prime flower]

I really like the cpx background lol

[fluid helm]

Will try a few things

and see what work best

the one at http://microbit.edublocks.org works really well

[prime flower]

select-ible backgrounds would be kinda cool for educators hosting a workshop too, like I was presenting at OSHWNYC I could change it to their logo

[tulip sleet]

@fluid helm do you need it even lighter? Or maybe a greyed blinka?

[fluid helm]

We have a themes menu, that changes the header colour

[prime flower]

or the school could put their logo/emblem, it's a wishlist item ive had for makecode for a while

[fluid helm]

blinka woud work

[prime flower]

RFC: add <blinka> as a recognized html element

[tulip sleet]

the micro:bit background looks good because it's a bunch of small gray bits, not too contrasty. Could repeat blinka and/or various other mascots

[prime flower]

Tiling really small, with adafruit icons and blinka

[tulip sleet]

like the printing on the new Adafruit cardboard boxes

[fluid helm]

not so sure?

[fluid helm]

oh now that's nice

[tulip sleet]

that's a little busy, could just repeat blinka and adabot or something and space them out more, and really light gray

[fluid helm]

Yeah, that's what I'm gonna try

[tulip sleet]

leave off adafruit star and name - we keep those for official - i'll double-check w/mgmt later today

[fluid helm]

Cool, after speaking to pt he said I could use the circuit python logo and blinka @tulip sleet

[tulip sleet]

excellent

[fluid helm]

Too busy?

I think just having a grid works best

[sick creek]

nice

[pulsar bloom]

Anyone have any luck with the rfm9x library using the feather M4? Setup works with M0 express and rfm965 featherwing but fails on the M4 with error saying check wiring.

[turbid radish]

Just a reminder that Adafruit has a spam-free newsletter for Python on Microcontrollers https://blog.adafruit.com/2018/08/27/python-on-microcontrollers-newsletter-sign-up-today-adafruit-circuitpython-python-2/

[raven canopy]

@pulsar bloom there is an issue filed with a possible fix: https://github.com/adafruit/Adafruit_CircuitPython_RFM9x/issues/8

[compact hornet]

@fluid helm Why put a pattern there at all? There's hardly any pattern that isn't gonna reduce readability. IMNSHO

[fluid helm]

We do it on https://microbit.edublocks.org and it works fine

It's getting the right tone

[manic glacier]

davea

[adafruit/circuitpython] New star added

[solar whale]

@slender iron sorry - I have to miss the meeting today -- group hug! I'll listen to it later.

[fluid helm]

Now recreating the examples on the learning system to go inside the samples menu

Will be back in 15 mins @slender iron so if I can go last that would be great

[tulip sleet]

@slender iron Guess I should take notes, since kattni's off? Not sure of the style, I need to look at an old one. Can you drop me a link to a new google doc?

[slender iron]

Thanks @tulip sleet ! https://docs.google.com/document/d/1wa9VL-hmmZgqOT87Qy-nJajIxm-C5Dl9rftL0D9d6ww/edit?usp=sharing

@fluid helm sounds good!

<@&356864093652516868> meeting in 5 minutes

[fluid helm]

cool

Can I test?

[slender iron]

@inland tusk any luck

[fluid helm]

brb

[meager fog]

hiiii im here in text

[fluid helm]

Hey! @meager fog

[meager fog]

hi josh!

[slender iron]

k, feel free to post hug reports

[meager fog]

mega hugs to sommersoft for being nrf52 buddy

[stuck elbow]

my heart goes to all the people who struggled with neopixels again

[meager fog]

also big round of applause for kattni & brennen who did 45 guides and tested/updated them for raspberry pi suppoty

support

they are on well deserved break, and we'll be back to do the remaining sensors/device

[fluid helm]

nope 😦

[meager fog]

(but if people have requests i can probably do one or two, open up an issue on circuitpython)

[stuck elbow]

I can't

[fluid helm]

https://learn.adafruit.com/adafruit-circuit-playground-express/circuitpython-quickstart

[raven canopy]

@jepler for running some tests on my PR. @jerryn for testing and troubleshooting the DHT-on-RPi PR. @ladyada for working on a learn guide for nRF bootloader install. Group hug to infinity and beyond!!

[tidal kiln]

:buzz_lightyear:

darn. no emoji. 😦

[slender iron]

https://github.com/tannewt/circuitpython/tree/hallowing

[fluid helm]

Really really cool!!! @prime flower

[prime flower]

thx josh! keep your 👀 on the learning system!

[stuck elbow]

The final prototype for PewPew Standalone is working and I have sent inquiries for quotes to the fabs to start the production soon. I also wrote a bouncing ball tutorial for it: https://pewpew.readthedocs.io/en/latest/tutorial.html

[turbid radish]

I am here, will go next

[prime flower]

that looks so good, I like the left hand side color-coding 😃

you're welcome mike! thanks for moderating my guides with your 🦅 eyes

(It's really cool, yeah makecode)

[slender iron]

@meager fog status update?

[raven canopy]

FrequencyIn: PR is in. Worked one change already thanks to @jepler's testing (bail out of endless interrupt loop). Will continue to work on things as the review progresses.

nRF DK: Finally kicked the tires on it last night. Win10 + Vagrant is proving difficult for the bootloader. Will push on it a little more, but may proceed with using Segger Embedded Studio. I'll document & screen grab either/both for possible inclusion to ladyada's learn guide.

.pylintr: patch successful except one outlier which has a PR in.

[slender iron]

@inland tusk type in if you like

[fluid helm]

Right so a temporary domain name for EduBlocks CPX is here: http://adafruit.edublocks.org not finished/fully pushed to server yet.

[tidal kiln]

i do

version numbering and bundle release process related

[tulip sleet]

https://semver.org/

[meager fog]

@slender iron sorry was in a meating

status update is PN532 driver is done, plz try it if you want RFID card support in circuitpython

[turbid radish]

🍖

[meager fog]

im currently investigating libgiod for circuitpython support on Pi - if anyone is excited to help lemme know 😃

[fluid helm]

People seen this? it's a camp badge powered by micropython/circuitpython (can't remember which) https://t.co/oQTrMSu5Mc

[meager fog]

and getting back to nrf52840 - i started a guide for building/burning

will publish once we have USB workin' so people can experiement

[fluid helm]

@hybrid scarab from Pimoroni might. Sounds like something he's done @slender iron

[meager fog]

also did some crickit HAT testing, so far so good

lol

[fluid helm]

Is micro:bit Crikit going to be a thing?

[meager fog]

mostly crawling out of a Pi party

micro:bit crickit is being manufactured today 😃

[fluid helm]

yay!

[inland tusk]

Thanks I heard you

[meager fog]

watch the blog, will try to get it at least for signups on wednesday

[fluid helm]

Cool, will have to get one for micro:bit magazine!

[meager fog]

sweet! yes makecode support works for sure

micropython is not as much yet - not sure if we can fit it

[fluid helm]

Yeah, it's pretty pushed for space

[meager fog]

we'll have to wait till a next-gen microbit - hopefully they'll make one with an '832 or '840

[raven canopy]

Thanks everyone!! Be well and Monday on!!

[meager fog]

for a spacious build

[fluid helm]

Hopefully, would be nice

also the library upload is not as nice as CircuitPython

[meager fog]

with '840 yes 😃 that would be ideal

[inland tusk]

browser version Newest version of firefox

[meager fog]

anywho thats my update

[fluid helm]

At PyCon UK they said if they delayed the micro:bit by 2 months it would have the '840

but they were delayed as it is i think

[raven canopy]

Dropping out. Salt mines await!

[slender iron]

thanks @meager fog just wrapped up the meeting

[tulip sleet]

@inland tusk what OS?

[raven canopy]

👋

[meager fog]

thanks folks!

@fluid helm did they mean the gen 1 or the next gen?

[inland tusk]

OS linux ubuntu 16.04

[meager fog]

i dont know they've released any details on if/when the next gen bit will come out or what it would use

[fluid helm]

Don't think it's gonna be soon

gen 1 @meager fog

[inland tusk]

ssounds familar

[tulip sleet]

@inland tusk working well for me on 18.04 with the app (not browser). had to set audio input in discord settings to default: specifying particular mic didn't always work

[fluid helm]

Also trying to build some EduBlocks hardware

this will be on order soon

[meager fog]

oh cool!

nice silkscreen 😃

is that a HAT/

[tulip sleet]

there's a download link on the home page for discord

[fluid helm]

Yup 😃

[inland tusk]

I will be updating my system this week. I will load the app.

[meager fog]

nice

[slender iron]

@inland tusk want to try now or just ping us later?

[fluid helm]

That's the back of it

[meager fog]

aww yeah, self documenting!

[inland tusk]

I am going to try to set up the app

[fluid helm]

That ground wire to the buzzer needs to be sorted out a bit

Should hopefully be on sale with Farnell by December!

Is anyone at the Blockly User Summit in California in a few weeks?

[tulip sleet]

@inland tusk turn down the gain a little more, still overdriving

[tidal kiln]

lower left?
https://moderndevice.com/featured/three-new-fluxamasynth-variations/

[fluid helm]

I forgot to say @slender iron https://github.com/AllAboutCode/pixelzero/blob/master/examples/colourtest.py

I've ported the CIrcuiPython neopixel lib to the Pi

for standard Python 3

[prime flower]

@fluid helm What Pi have you tested this on?

[manic glacier]

ravipurigen

[adafruit/circuitpython] New star added

[fluid helm]

3B+ @prime flower

[prime flower]

This is SUPER useful

I'm going to fwd the repo to @meager fog

Do you need to run it as root?

[fluid helm]

Yep, if it doesn't detect sudo, it will print a friendly error and not import it @prime flower

It's not fully completed yet and needs lots of work

[prime flower]

Yeah

[fluid helm]

but proof is there

I've been using it for personal projects for a while now, thought it might be useful for the wider community @prime flower

[prime flower]

fwd'd

thanks for sharing!

[fluid helm]

No worries, would be nice to get it used a bit more 😃

Bye!

Ahoy @slender iron , could you locate me to the circuitpython neopixel library file. I see to have lost where it is 😦

[tidal kiln]

@fluid helm should be neopixel.mpy on your board under /lib

[timber mango]

Hi!

I'm almost stuck with my Feather M4 Express : no CIRCUITPY, no REPL but I can get FEATHERBOOT. I try to update to 3.0.1, without effect.

As I was working on big change in my script, I think my main.py make the board unresponsive.

I have seen UF2 firmware to erase the flash or ignore boot.py & main.py, but only for other boards. Anything for Feather M4 Express ?

[tulip sleet]

@timber mango

import storage
storage.erase_filesystem()

[slender iron]

<@&356864093652516868> Meeting recording is here: https://youtu.be/pyew9CN2hDM

[tulip sleet]

@timber mango missed that it was no repl. I'll be back soon

[timber mango]

yep, no REPL... when I can connect, nothing happend and it do nothing, even with [Ctrl]+[D] or [Ctrl]+[C]...

[tulip sleet]

i can get you a CPy .uf2 that does NOT run main.py / code.py on startup. give me a few minutes

[hybrid scarab]

@fluid helm who, what, why, when, where? 😄

[tulip sleet]

@timber mango This .uf2 is 3.0.1, but starts in safe mode and does not run main.py or boot.py

[hybrid scarab]

I really should give CircuitPython some time- feel it might be a good antidote for the not-so-great experience of writing Python for the micro:bit

[timber mango]

@tulip sleet Thank you !

[stuck elbow]

@hybrid scarab what was wrong with the micro:bit, out of curiosity?

[hybrid scarab]

@stuck elbow the hardware is just too limited to do anything fun, even when you go to painful extremes to optimise code, and the shipped Python doesn't support binary distributions of libraries

It's hard to ship concise libraries for it with well structured code because you hit stack depth limits, or blow the parser memory, etc

And that's no slight against the platform/MicroPython itself- it's just a professional observation. I'm very much a fan of PXT (even though I dislike JavaScript) because it compiles down to something less resource intensive and doesn't need an embedded interpreter.

[stuck elbow]

@hybrid scarab well, I agree with your assessment here

[hybrid scarab]

I think the tooling surrounding MicroPython - at least on the micro:bit - has a little way to go, too, since there's no official way to distribute libraries even if you do manage to write them

Really wanted to get our eInk displays running but it was so limited that even with an FRAM memory buffer for the display, the code limit would be blown just implementing the drawing logic for the display and handling of the FRAM. I could probably have optimised it to the nth degree, but then there'd be nowhere for user code anyway 😄

TBF you could question the sanity of shipping an eInk display for the micro:bit in the first place- but we don't really do sane ;D

[stuck elbow]

@hybrid scarab I once made a "shield" for the micro:bit with a small display and dpad+fire buttons, and tried to make games for it, but even though I had my display code very well optimized, I just ran out of memory for any reasonable game logic

[main meteor]

I didn't know the micro:bit had FRAM, I thought that was just the MSP430FR and related chips.

[hybrid scarab]

Nah I bodged the FRAM onto our add-on board for prototyping, it wasn't onboard

[main meteor]

Oh, cool!

[hybrid scarab]

Wrote a really nice class that maps all the FRAM IO to standard array indexing calls too sigh

[stuck elbow]

well, that used extra code space

[hybrid scarab]

So you could "memory = FRAM()" and then "x = memory[0:30]" or "memory[0:3] = [1,2,3,4]"

[timber mango]

@tulip sleet same result : FEATHERBOOT => ok, I flashed with the safe-mode UF2 you gave, but no CIRCUITPY

[stuck elbow]

it's the too short blanket problem

[main meteor]

I didn't know FRAM was available as a separate item. Do you happen to know which chip you used?

[timber mango]

and no REPL

[hybrid scarab]

But yeah that was certainly not optimised 😄

[tulip sleet]

@timber mango can you get to repl with this .uf2? then you can clear the filesystem

[timber mango]

no REPL

[stuck elbow]

@hybrid scarab I'm now working on a "microbit done right" device that I plan to use for workshops myself

[tulip sleet]

i'll get you another .uf2 that clears CIRCUITPY on startup

[timber mango]

ok

[hybrid scarab]

@main meteor not off the top of my head, but I can grab my laptop and look it up

[timber mango]

My script may have written a lot of data on flash memory...

Full flash can produce that ?

[hybrid scarab]

@stuck elbow something pin-compatible with a better micro would be nice

Even PXT has problems- can't use Bluetooth alongside some more demanding add-ons for example

Which is a shame, because- for example- a Bluetooth LED display would be cool

Gotta respect whoever was crazy enough to rewrite the MicroPython parser to even run on the micro:bit at all, though 😄

[tulip sleet]

@timber mango this erases CIRCUITPY on on every hard reset

it shouldn't break if you fill up flash, but there could easily be a bug

[slender iron]

@hybrid scarab micropython on micro:bit was damien, the micropython creator

we've tried to establish best practices for circuitpython libraries. we're bringing those libraries to raspberry pi now as well

[hybrid scarab]

Ah that would explain why it works at all then 😄

https://www.winbond.com/resource-files/w25x40cl_e01.pdf Ah it was this serial RAM chip, not FRAM as it were

@main meteor ^

[fluid helm]

Oh yeah, we were talking about Python libs and C in the meeting earlier @hybrid scarab

Talk to @slender iron

[hybrid scarab]

But there are such things as: https://uk.rs-online.com/web/p/fram-memory-chips/1242933/

Ahhhh @split ocean

Not being able to write micro:bit Python modules in C is definitely one of my bugbears with the platform- but I'm so frazzled keeping up with things that I... well can't keep up with things!

[slender iron]

ya, @meager fog wants to package a gpiod library in pypi

[timber mango]

@tulip sleet new colors ! and CIRCUITPY is back !

[hybrid scarab]

I have all the problems and none of the solutions- like a good consumer 😄

[tulip sleet]

the new colors may be a bug, but glad you've recovered!

[timber mango]

and the REPL

Thank you !

😃

[hybrid scarab]

A gpiod library for the micro:bit, or? I'm not sure I'm on the right page

[slender iron]

micropython can't dynamically load c modules yet. our approach is to bake in just the critical C functionality and use python for everything else

gpiod for CPython on raspberry pi

[hybrid scarab]

Oh it was byte-compiled Python I was looking into, rather than C specifically

[fluid helm]

Yeah, the CircuitPython way of doing libraries is so nicer. There is a libs folder on the flash drive of the device where you drag and drop the library @hybrid scarab

[hybrid scarab]

Since it being byte-compiled would, presumably, avoid the interpreter explosion that some larger libraries encounter

[slender iron]

ah, we bundle everything together

https://github.com/adafruit/Adafruit_CircuitPython_Bundle

[main meteor]

Ah, that Windbond serial flash is similar to the one used on the Metro, Feather, etc.

[hybrid scarab]

I was under the impression some genius had shipped a "pip" for flavours of MicroPython that just downloaded dependencies on the fly, too?

[stuck elbow]

@hybrid scarab pin-compatible is not an option for me, as one of the problems I have with micro-bit is the connector

[slender iron]

ya, they have it but we don't use it

[stuck elbow]

@hybrid scarab though there is an esp32-based board that is compatible with microbit

[slender iron]

we use pip for the libraries on rpi only

[timber mango]

@tulip sleet I'll look at my code and what goes wrong tomorrow.

Good night !

[tulip sleet]

good night, glad to help!

[slender iron]

https://pypi.org/search/?q=adafruit-circuitpython

[stuck elbow]

@hybrid scarab https://www.aliexpress.com/store/product/Banana-PI-Bit-board-with-EPS32-for-STEAM-education/302756_32861694375.html

[hybrid scarab]

Ooh yeah I've seen that one, in retrospect

@stuck elbow yeah the connector is... special

The edge-connector part to mate with it is expensive and cumbersome to say the least

But I do kind-of like how it makes the micro:bit a cartridge to a larger host board

[stuck elbow]

yeah, so I instead used standard pin header holes, but staggered, so you don't have to solder

you can just insert a male pin header into them, and plug it directly into a breadboard

[hybrid scarab]

We use that same technique to hold headers for soldering, is it really good enough for a robust connection?

[stuck elbow]

I do 400kHz i2c over it without problems :)

[hybrid scarab]

That's pretty decent! I used to jam strip header and stick blue-tac under my boards so it'd make a connection 😄

[stuck elbow]

I suppose it will wear off after a while

then you can solder it

[hybrid scarab]

True! We're currently shipping shims that use really, really tight tolerance holes to mate with the Pi header without soldering- they're still pretty hit and miss it seems

[stuck elbow]

the point here is that the plastic part of the header strip acts as a spring

[hybrid scarab]

Good zero solder connectors is the pipe dream 😄

That's a good point

Or maybe self-soldering headers- like self-heating coffee cans, pull cord to solder

[stuck elbow]

great idea, lets give those to children

what could possibly go wrong

[hybrid scarab]

Explosives are fun!

[stuck elbow]

(I would rather use some kind of conductive glue, I guess)

[hybrid scarab]

Ooh I tried that once, ugh

Used conductive silver paint to try and fix a header I'd made a total and complete bodge of- made such a miserable mess, suffice to say that was long before I'd learn to solder properly

The little surface mount 40-pin push-through headers the Pi Foundation mount on stuff like Sense HAT are kinda cool

[stuck elbow]

then again, during the maker faire workshop in here with have 7yo children using soldering irons without supervision

[hybrid scarab]

But an expensive extra part versus just staggered (and free) PCB holes

[stuck elbow]

and there has been no accident in 2 years

[hybrid scarab]

That's pretty impressive- I burned myself the first time I ever touched an iron, I have no idea how

[stuck elbow]

if it smells like chicken, you are holding it wrong

[hybrid scarab]

I can't have seen too many stock photos back then, they weren't really a thing

[stuck elbow]

it might be because those are swiss children, they walk to the kindergarten alone

in the snow

uphill

both ways

[hybrid scarab]

across the Alps, fighting off bears?

[slender iron]

@hybrid scarab definitely check out circuitpython if you haven't. We've tried to make it better than MicroPython

[raven canopy]

@main meteor Adafruit has two FRAM breakouts.
I2C: https://www.adafruit.com/product/1895
SPI: https://www.adafruit.com/product/1897

and i'll eventually get to the library: https://github.com/adafruit/circuitpython/issues/665

@slender iron that .pylintrc PR i mentioned earlier...ladyada merged it two days ago, which my brain totally blocked earlier. 😄

[slender iron]

kk 😃

[stuck elbow]

@slender iron looking at your display code, I think it's important to remember that different st7735 displays have different RGB/BGR defaults and different offsets of the physical display in the memory

[slender iron]

@stuck elbow yeah, I think its ok for that. The init code there is explicitly for whats on the Hallowing

[raven canopy]

@gentle bronze thanks for de-mergehexing the genhex! in a total 🤦 moment, i just realized that mergehex has its own directory that i didn't put into PATH.

[stuck elbow]

@slender iron it's spi, not parallel, right?

I never had bit order problems with that...

[slender iron]

ya, spi. I think its my bitpacking code

the byte order is wrong but not the bits

gotta finish the newsletter before I can look at it more

[stuck elbow]

by the way, is there a good way to look at all changes to a branch without having to wade through all commits?

on github, I mean

[raven canopy]

the only way, on github at least, i could think of would be to fork, and then use compare against the master branch. basically, act like you're going to submit a PR.

actually, you don't need to fork...

[slender iron]

compare https://github.com/adafruit/circuitpython/compare/master...tannewt:hallowing

[stuck elbow]

thanks

so the fourwire functions include CS so that you can use DMA?

(later on, I mean)

[slender iron]

ya and data/command

[main meteor]

@raven canopy Thanks! I'm boggled that I somehow missed those!

[split ocean]

@hybrid scarab sup?

[stuck elbow]

well, dma mostly makes sense for data

[slender iron]

ya

[fluid helm]

Think he might have tagged you instead of @slender iron maybe @split ocean

[slender iron]

its still faster to do from C than python

[manic glacier]

sommersoft

[adafruit/circuitpython] New comment on pull request #1144: Add pulseio\.FrequencyIn

Through additional brainstorming, and discussion over in samd-peripherals, I'm wondering if I should setup the TC using a different clock than DFLL.

I could use DPLL1, set below 100MHz to ensure that TC4-7 don't exceed that limit. Then, all the kludgeyness of using the FREQM to adjust for DFLL drift wouldn't be necessary. In my quick-looks, DPLL1 isn't used, but I still need to roll through the common-hal to verify that.

I los...

[raven canopy]

@meager fog got through my bootloader Vagrant woes! Success! Might be worth it to update the "if you have hex" to show that you can use the genhex flag to create the hex. If you're interested in putting it in the guide, I can run through using the Segger Embedded Studio option...

[solar whale]

@raven canopy @pulsar bloom I have reproduced your issue with the rfm9x on a featherM4_express and I also can confirm taht the fix in https://github.com/adafruit/Adafruit_CircuitPython_RFM9x/issues/8 solves the problem.

[meager fog]

oh man i was totally eating while y'all talked about ligpiod

[solar whale]

@pulsar bloom @raven canopy aha -- found the rfm9x issue with the featehr_m4 --- it just needs a delay between the command to enter sleep mode and the rest of the setup. The commentedlines are not the real issue. HAving trouble with github at the moment -- I'll update the issue when it comes back.

[meager fog]

@raven canopy yah wanna edit the guide?

@hybrid scarab i missed ya while 🌮 - i opened an issue on rpi_ws281x if you have a chance to look at it, just before i send a PR

@solar whale we havent done a ton o testing with the rfm9x but ill take a PR if ya want - i just dont have anything wired up

[solar whale]

@meager fog sure -- I'll wrok up a PR -- Just want test ti on the M0 and RPi first - adding the delay "should" be harmless. The commented out line was actually setting it to "low_frequencey" mode as weel then it gets unset a few lines later.... The delay just makes sure it really goes to sleep mode.

[meager fog]

totez

thank ya!

[solar whale]

glad to help!

[meager fog]

we did almost all i2c and simple-spi devices

github's back up so i guess i gotta get back to work 😃

[solar whale]

@meager fog just to clarify, the issue was on the m4 -- not RPi -- I have a small change to the rfm9x_simpletest for the RpI -- I'll add it to the PR.

[meager fog]

interesting

it was a speed thing?

like we sent commands too fast?

[solar whale]

yes -- it needs some time to get into sleep mode

it has back to back sleep mode - then set long range. Ok on M0, not on M4

[meager fog]

oooh totally that makes lots of sense

[solar whale]

yup -- just needed to stare at it a minute!

on the M4 if I readback the op_mode, it was not in sleep Mode

[raven canopy]

@meager fog guide edited.

[meager fog]

thanx 2 x

[meager fog]

@tidal kiln do you want to try splitting pn532 into a module to see if that helps with memory?

[tidal kiln]

@slender iron ok. finally actually running through this.
https://learn.adafruit.com/creating-and-sharing-a-circuitpython-library/sharing-in-the-community-bundle
you've got this sequence of commands for adding the module:

git submodule add <github https clone link> libraries/<category>/<library name>
./update-submodules.sh
python3 build-build.py

i think the last one needs an update?

[meager fog]

we haven't written any dox for it so im ok changning api now

[tidal kiln]

@meager fog sure. pretty straight forward. also - were you going to add it to bundle?

[meager fog]

yah once i had more testers

[slender iron]

@tidal kiln ya, its not needed

[meager fog]

@tidal kiln so now's a good time to do the split

we'll need 4 files

i think that will work. not 100% sure tho

[tidal kiln]

@meager fog yep yep. saw your comment. easy enough to branch / refactor / test. i'll include a trinket m0 in the hw testing.

[meager fog]

thanx ya i that is a good test

[tidal kiln]

@slender iron thanks. is second command still needed? the update script? looks like it's just a few git commands.

[slender iron]

ya, it makes sure its on the latest tag rather than the latest commit

[tidal kiln]

i nominate this for most useful error message of the year award:

Failed to generate versions file. Its likely a library hasn't been released yet.

esp. the second sentence.
🤦

indeed. now travis is happy.

@slender iron thanks. pr'd to the bundle for the tca9584a library. you ok with me closing:
https://github.com/adafruit/circuitpython/issues/1093
?

[slender iron]

😃

sure! always feel free to close. can always reopen if I disagree

will look tomorrow, thanks!

[tidal kiln]

sure. np.

[manic glacier]

caternuson

[adafruit/circuitpython] New comment on issue #1093: need driver for TCA9548A I2C Multiplexer

Closing this for now. See here for driver:
https://github.com/adafruit/Adafruit_CircuitPython_TCA9548A

Thanks @deshipu for the excellent idea on this one - to make the switching happen behind the scenes automagically. Thanks to @tannewt and @ladyada for comments on how best to implement.

caternuson

[adafruit/circuitpython] Issue closed: #1093 need driver for TCA9548A I2C Multiplexer

[pulsar bloom]

@solar whale thanks can confirm it works

[meager fog]

@gentle bronze good morning! im around for a little bit

if you'd like me to try or test something

[gentle bronze]

@meager fog good evening. Ah, not yet, I am still troubleshoot the serial issue,. It should be ready when you waked up though 😄

[meager fog]

ha ok 😃

[manic glacier]

hathach

[adafruit/circuitpython] Pull request opened: #1146 update tinyusb to fix \#1007 serial issue

The issue is a bit complicated, but require only a few lines to fix as expected. The issue is EP0STATUS to send control status also require Easy DMA access. This seems to be missing from Nordic documents

https://github.com/hathach/tinyusb/compare/7b35cd0203bc409d7c1aefc075672103cb4a913e...583326e535454f16b06ebdb9cc06869602a5564c#diff-9b5241cf0c81e9eb639b442c58170472R220

hathach

[adafruit/circuitpython] New comment on issue #1007: USB serial

The issue is a bit complicated, but require only a few lines to fix as expected. The issue is EP0STATUS to send control status also require Easy DMA access. This seems to be missing from Nordic documents

https://github.com/hathach/tinyusb/compare/7b35cd0203bc409d7c1aefc075672103cb4a913e...583326e535454f16b06ebdb9cc06869602a5564c#diff-9b5241cf0c81e9eb639b442c58170472R220

hathach

[adafruit/circuitpython] Issue opened: #1147 nrf52 build failed with DEBUG=1 due to printf function

As @dhalbert and @arturo182 suggested a couple of ways to resolve this. However, since this is in the main.c
https://github.com/adafruit/circuitpython/blob/master/main.c#L454

I better leave this to those who have more knowledge on the CP, since main.c changes will affect all port

make V=1 BOARD=pca10056 DEBUG=1 all
GEN build-pca10056-s140/genhdr/mpversion.h
CC ../../py/modsys.c
CC ../../extmod/modwebrepl.c
CC ../../main.c
../../main.c: In function '__assert_func':
../../ma...

[hybrid scarab]

@meager fog I'll take a look!

Ooh, that's a stickler!

[manic glacier]

dhalbert

[adafruit/circuitpython] New comment on issue #1147: nrf52 build failed with DEBUG=1 due to printf function

I will fix this soon in a separate PR or as part of another.

[gentle bronze]

@meager fog Some correction for bootloader tutorial https://learn.adafruit.com/circuitpython-on-the-nrf52/nrf52840-bootloader .
$ make BOARD=feather_nrf52840_express VERSION=6.1.0r0 flash will flash the board with pre-built binary. "VERION=" only needed for flashing specific binaries without compiling the code.

$ make BOARD=feather_nrf52840_express flash without VERION will compile the current code and flash the newly generated hex. The code doesn't match your screenshot, sorry if i didn't doc it well to cause some confusion.

[manic glacier]

hathach

[adafruit/circuitpython] New comment on issue #1147: nrf52 build failed with DEBUG=1 due to printf function

yeah, close this when you see appropriate.

[gentle bronze]

just to be clear, flashing binary specifcied withVERSION=, user can skip git submodule update --init as well since it doesn't compile the code at all, just use the board and the version to look for the correct binaries.

[slender iron]

@raven canopy when you have a chance: https://forums.adafruit.com/viewtopic.php?f=60&t=140262

[meager fog]

@hybrid scarab yay thanx!

[manic glacier]

tannewt

[adafruit/circuitpython] New comment on pull request #1133: \[WIP\] Spanish translation\.

Would you like me to polish this up so we can get it checked in?

[hybrid scarab]

@meager fog no worries! This has somehow evolved into me trying to get Travis CI working on the library, since I can't pip it until tomorrow 😄

[gentle bronze]

@tulip sleet Just a quick question for USB serial string, currently for samd port, is the serial different between the bootloader mode & the circuitpython mode ? I am not sure if we could use the same serial or have to make them different.

[manic glacier]

tannewt

[adafruit/circuitpython] New comment on pull request #1144: Add pulseio\.FrequencyIn

Thank you for your continued work on this. I think it should only be added on express boards with space available. We don't have room for new things in the non-express builds.

What is the lower bound of what you are trying to measure? What's the high bound? Interrupting on every pulse seems very limiting to the top end.

Why not interrupt at a fixed frequency based on the best calibrated clock and count the number of clock cycles over that duration? The TC event action would then be COUN...

[tulip sleet]

@gentle bronze there are different USB PIDs for bootloader and circuitpython. There is one PID for UF2 bootloader, onen for Arduino, and one for CircuitPython.

[manic glacier]

tannewt

[adafruit/circuitpython] New comment on pull request #944: WIP/RFC: Introduce a display module for driving ILI9341 over SPI

Any conclusions from this? I'm working on a different API here that is more like _stage than framebuf.

[gentle bronze]

@tulip sleet yeah, the pid is different but i wonder that if the serial needs to be different as well. The one you help me out with the nipple to hex. Though, don't be borthered. I will try to find a samd board to test with. I think ktwon did sent me one a long time ago. May need it for other ref as well.

[tulip sleet]

no, serial does not need to be different; we haven't done that on atmel

some devices don't even report serial. I think we are using serial to distinguish MSC because that helps in some way; i forget the details

[gentle bronze]

Ok, got it. Then we could just use one serial for both modes. Thanks.

Btw I submitted a pr that should fix serial connection issue. Let me know if you still have any issues, I will check it out tomorrow.

[tulip sleet]

i'll build new one - thanks for debugging!

[gentle bronze]

NP, It is my bug :(, though it is a bit complicated. I should have troubleshoot ed it earlier :D

[manic glacier]

dhalbert

[adafruit/circuitpython] Pull request closed: #1146 update tinyusb to fix \#1007 serial issue

dhalbert

[circuitpython:master] 3 new commits

ba26ca9 update tinyusb to fix #1007 serial issue - hathach
87811ec update readme board name feather_nrf52832 - hathach
93100a7 Merge pull request #1146 from hathach/nrf52_usb... - dhalbert

dhalbert

[adafruit/circuitpython] Issue closed: #1007 USB serial

dhalbert

[adafruit/circuitpython] New comment on issue #1140: CircuitPython 3\.0\.1 breaks built in NeoPixel on Feather M4 Express

Fixed by #1141.

dhalbert

[adafruit/circuitpython] Issue closed: #1140 CircuitPython 3\.0\.1 breaks built in NeoPixel on Feather M4 Express

C47D

[adafruit/circuitpython] New comment on pull request #1133: \[WIP\] Spanish translation\.

Hi,

Yes if possible, but i would also like to know why the Travis build is failing, i checked the first failing test and it's marking the following error:

So if you can polish it it would be great, i'm not having as much free time as i would like to.

[raven canopy]

@slender iron forum post answered.

[slender iron]

Thanks!

[hard loom]

New to Circuit Python anyone 1. Where do I go to get started with this code. 2. Is it an interpretive language, or something I can compare it with???? Thank you in advance for the Newbie 😳

[solar whale]

@hard loom If you are specifically interested in CircuitPython -- look here https://learn.adafruit.com/welcome-to-circuitpython?view=all CircuitPython is an implementation of Python3 that executes on microcontrollers. If you are completely new to Python, then you may want to start with a general tutorial on Python. Are you using a micro-controller board?

[hard loom]

Yes I am using microcontroller boards but pitifully. An old-timer just getting into this open source coding. To give you an idea of my age my first computer language was Fortran and Pascal in high school. Before the 8086, but had a great time with it then it was basic ,basic +, C ,C+, C++, you get the idea. Thanks for the tip really appreciate it and I'm really need it. Have a good day my friend and I'm not going to be the first one to turn that switch on), Circuit-On

[solar whale]

@hard loom We are contemporaries 😉 -- Glad to help -- keep asking if you need more.

Actually - I think I'm considerably older than you are ...

[hard loom]

Thank you very much, great to hear I'm not the only one that remembers the C: and hated the rise of Windows GUI. HA HA the good old days when your program ran all night

Signing off now have a good night everyone. Need my beauty sleep

[meager fog]

@hard loom welcome and nite!

[solar whale]

@hard loom good night !

[storm spoke]

ok, noobist-noob here! I decide to order some toys from Adafruit to make a wearable and "learn through experience"... I purchased some Gemma MO boards (circuit python) and some LED matrix's and was loosely following a wearable guide on Adafruit that explains how to wire up the device and import the code! I got it working! however, I want to generate my own effects/displays from the matrix and here is where I'm stuck. Adafruit offers "librarys" for the matrix, but they are all audrino (i think) and my controller is Circuit Python. I think it's capable of Audrino, but what I've read is that I should start with Circuit python... I guess my question is: should I try to do anything with the matrix-libraries, or do I have what i need with Circuit python.

My challenge with modifying the existing code, is that it was written for mono-color matrix, and i bought bi-led matricies, so I don't know how to address the code in order to access the other color capabilities of the matrix

[main meteor]

If you're happy with CircuitPython, may as well stick with it. It does support the bicolor matrix. There's some example code here: https://github.com/adafruit/Adafruit_CircuitPython_HT16K33/blob/master/examples/bicolor24.py

Oops, that example code is for the bicolor bargraph. The library does support the bicolor matrix too, but I don't see example code for that.

[storm spoke]

Ah a breadcrumb, thank you madbodger!

[main meteor]

The matrix example code is here, and a commented section shows how to set up the bicolor 8x8 matrix: https://github.com/adafruit/Adafruit_CircuitPython_HT16K33/blob/master/examples/matrix.py

[storm spoke]

so the HT16k33 library, is it a set of code that acting as a driver for the the matrix. IE do i need to load a library that is specific to the matrix im trying to drive?