circuitpython-dev

[tough flax]

Odd... looks like adafruit_imageload is included in the nightly build, but not the bmp submodule

so it won't actually decode the file

So... copying the bmp folder over successfully loads a BMP file (yay)

But displayio is not in the build

I can't setup the build env on this machine so I might be done for the night 😕

So, the latest build (en_US-20180910-9ace50a) doesn't seem to have this displayio in it. I'm OK with waiting on this, but is there a way to kick off a Travis build?

[high oracle]

Is the inline assembler available in circuit python port?

[slender iron]

@high oracle nope. too advanced for what we're going for

[high oracle]

Thanks @slender iron

[slender iron]

np

[manic glacier]

ATMakersBill

[adafruit/circuitpython] Issue opened: #1183 Adafruit\_LoadImage Buffer Management

Per @tannewt and @ATMakersBill's Discord Chat, loading multiple image files will be needed by many scenarios.

Right now, loadimage() allocates the pallete and bitmap objects for each file load. At best this means that the user will need to delete the bitmap object (which scales with widthXheight) before loading the next image. That will cause memory fragmentation depending on Python's memory management.

Scott is looking at reading directly from files to the Display's screenbuffer, ...

[tough flax]

FYI I did get one image loaded 😃

[manic glacier]

deshipu

[adafruit/circuitpython] New comment on issue #1170: Support scrolling by changing the portion of the framebuffer read by the display

Yes, the datasheet is not clear about this, but it controls how the pixels are written by the memory write commands, not how they are sent to the display. At least that's what I figured out with my experiments, I might remember it wrong, it was a year ago.

[manic glacier]

deshipu

[adafruit/circuitpython] New comment on issue #1170: Support scrolling by changing the portion of the framebuffer read by the display

Actually, that wasn't PTLAR. PTLAR is for only using a part of the display (for example, when the rest is covered by the bezel).

[high sluice]

Hello,
I'm using a Trinket M0 with CircuitPython to run a virtual mouse and it works great.
But is there a way to not mount the Trinket as a USB storage device? (presumably a different bootloader?)
Use case: I'm happy with my CircuitPython code and I don't anticipate changing it and I want to use my device as strictly a plug-n-play and not have it do anything besides the virtual mouse functions.

of course it's not out of the question that I would eventually want to change my code so I'm guessing I can just reload the normal boot loader, test with the mass storage device as normal and then switch back.

[tawny creek]

@high sluice I'm not sure if there is a flag to not mount CIRCUITPY on the bootloader: https://github.com/adafruit/uf2-samdx1

[high sluice]

thanks for the link. 👀

[manic glacier]

tannewt

[adafruit/circuitpython] New comment on issue #519: More accurate time reporting

@CBMalloch how is the light beam being interfaced to the microcontroller? We may be able to add a C helper for it.

It's highly unlikely we'll ever allow for predictable timing in Python execution. There are too many critical things happening around the VM to do it.

tannewt

[adafruit/circuitpython] New comment on issue #1175: Stack size problems

@dhalbert Increasing works for me. Its possible this is the error when the parse stack runs out of space since it's not actively running python code.

[slender iron]

@high sluice I don't think you want a new bootloader. That only controls the *BOOT drive. I think you want this: https://github.com/adafruit/circuitpython/issues/1015

[manic glacier]

ATMakersBill

[adafruit/circuitpython] New comment on issue #1015: Dynamic USB device enabling/disabling

This would be helpful in some of our projects as well. One way I could see this working is to check for the presence (or absence) of a file or config option in the BOOT filesystem. So, when CP's UF2 is loaded, it checks for that file (maybe a config file?) and checked which USB descriptors to present.

This would make it possible to enter the bootloader mode, tell it to not share the mass storage device or the HID (or the Serial port?) and restart. If you ever wanted to change that, just...

ATMakersBill

[adafruit/circuitpython] New comment on issue #1182: Support streaming pixel data off the filesystem

After an evening of thinking about it, I think this is the only way to do the "load any image" approach.... and it really needs to be memory aware (like only having a single scan-line buffer and everything else static).

[manic glacier]

tannewt

[adafruit/circuitpython] New comment on issue #1015: Dynamic USB device enabling/disabling

The idea I had was to control the descriptors from boot.py which runs before usb enumeration.

ATMakersBill

[adafruit/circuitpython] New comment on issue #1015: Dynamic USB device enabling/disabling

Where does that live? On the BOOT drive or on the CIRCUITPY drive? How does a user edit it?

[high sluice]

@slender iron yes, that seems to be what I want.

[ruby lake]

are there any numbers on how long a pin toggle takes in cp? Say something like somepin.value = True; somepin.value = False; somepin.value = True --how long does the pin actually stay false/low in milliseconds or microseconds

I can measure it later of course

on an m0, I figure the m4 will be faster because m4

[manic glacier]

notro

[adafruit/circuitpython] New comment on issue #1175: Stack size problems

I need 8k to run the time_os and time_time unittests.
For the tests that test unittest itself 8k isn't enough so I use 12k.

[stuck elbow]

@ruby lake I'm afraid this can change from version to version and from device to device

if you need to toggle pins fast, use the pulseio module

or busio

[ruby lake]

@timber mango it is not time-critical for me, I am just curious

I know the duration is fine for latching a regiter

~register

[stoic marsh]

in the Mu 1.0.0 enviroment I am using Circuit Playground express. and Crtl-d nor Crtl-c are working after I save the program

[arctic heron]

In my (limited) experience, something (possibly out of memory) has made your app go south. If you reset does it come back?

[stoic marsh]

@arctic heron If i shut down Mu and restart it works for 1 time

[arctic heron]

Sounds like something in there is hanging. I'd start by commenting out some and trying to narrow down what is killing it. Or, you could try REPL to execute some of the statements individually

[arctic heron]

And lots of print of course 😀

[manic glacier]

dhalbert

[circuitpython:pwmout\-tcc\-channel\-claiming] 1 new commit

66a7e45 PWMOut was not claming channels on shared TCCs - dhalbert

dhalbert

[adafruit/circuitpython] New branch created: pwmout\-tcc\-channel\-claiming

dhalbert

[circuitpython] Branch pwmout\-tcc\-channel\-claiming was force-pushed to `c3918ba`

Compare changes

dhalbert

[adafruit/circuitpython] Pull request opened: #1184 PWMOut was not claming channels on shared TCCs

Fixes #1106.

@deshipu - you may want to test this. I did try a test case similar to yours. The code was allocating the same TCC channel to multiple pins. This means that you may run out of timers earlier previously, because we were allocating the same TCC and channel to multiple pins by accident.

[manic glacier]

dhalbert

[circuitpython:pwmout\-tcc\-channel\-claiming] 1 new commit

f289863 bump gcc-arm-embedded to 7-2018q2 to save flash... - dhalbert

dhalbert

[adafruit/circuitpython] New comment on pull request #1184: PWMOut was not claming channels on shared TCCs

Some builds didn't fit any more by a tiny amount, so I bumped the compiler up to the version I was already using anyway.

[manic glacier]

dhalbert

[adafruit/circuitpython] New comment on pull request #1184: PWMOut was not claming channels on shared TCCs

Didn't mean to do a PR from adafruit/circuitpython. I had to reclone and recloned from adafruit/ instead of dhalbert/

[tough flax]

So... my trinket's Dotstar just suddenly switched from RGB to GBR? Is that like a thing?

Actually it's BGR... I'm quite confused. This was just working

This is bizarre - This code comes up RED # Used if we do HID output, see below

One pixel connected internally!

dot = dotstar.DotStar(APA102_MOSI, APA102_SCK, 1, brightness=0.8)

RED = (255, 0, 0)
YELLOW = (255, 150, 0)
ORANGE = (255, 40, 0)
GREEN = (0, 255, 0)
TEAL = (0, 255, 120)
CYAN = (0, 255, 255)
BLUE = (0, 0, 255)
PURPLE = (180, 0, 255)
MAGENTA = (255, 0, 20)
WHITE = (255, 255, 255)

dot.fill(BLUE)
dot.show()
while True:
time.sleep(1)

I honestly think I have a bum Trinket M0

[manic glacier]

sommersoft

[adafruit/circuitpython] New comment on pull request #1184: PWMOut was not claming channels on shared TCCs

This is the direction I was leaning in my "investigation" as well. In that since an existing TCC set at the desired frequency was looked for and used, any subsequent changes to it applied to all latched pins. I had stopped short of digging into the channel aspects before doing some domestics. Nice work @dhalbert!

[tough flax]

Ok, I found the problem... it's a problem with the examples in the Learning System. https://learn.adafruit.com/mini-pinball-table-with-gemma-m0/code-the-gemma-m0-with-circuitpython

MOSI and SCK are reversed in the example - which leads the DotStar to go nuts

Where do I enter issues for the examples?

[raven canopy]

@tough flax since those code examples are not hosted on GitHub, i would use the Feedback? Corrections? link on the left side of the guide.

[tough flax]

I created an issue https://github.com/adafruit/Adafruit_Learning_System_Guides/issues/327

I figure there's a project or the learning system guides, so that is where it should live

[raven canopy]

@slender iron working the meeting notes. i made these requested changes a while back, but forgot to ping you for a re-review and/or merge: https://github.com/adafruit/Adafruit_CircuitPython_MCP230xx/pull/1

[tough flax]

The thing is I'm really not sure I copied that code... I think it might have been on the Trinket when I opened it. - I wish I had copied the code as it was

[manic glacier]

tannewt

[adafruit/circuitpython] Pull request closed: #1184 PWMOut was not claming channels on shared TCCs

tannewt

[circuitpython:3\.x] 3 new commits

c3918ba PWMOut was not claming channels on shared TCCs - dhalbert
f289863 bump gcc-arm-embedded to 7-2018q2 to save flash... - dhalbert
2cb7039 Merge pull request #1184 from adafruit/pwmout-t... - tannewt

[sturdy furnace]

Has anyone recorded (and played) audio with a CPX?

e.g. produced a .wav or equivalent file. I'm getting some weird results

[manic glacier]

dhalbert

[adafruit/circuitpython] New comment on issue #1056: "make long lived" adversely affects UART on Metro M4 Express

Fixed by #1080.

dhalbert

[adafruit/circuitpython] Issue closed: #1056 "make long lived" adversely affects UART on Metro M4 Express

[manic glacier]

dhalbert

[adafruit/circuitpython] New comment on issue #1106: pulseio\.PWMOuts share duty cycle values when many of them are created

Fixed by #1184.

dhalbert

[adafruit/circuitpython] Issue closed: #1106 pulseio\.PWMOuts share duty cycle values when many of them are created

dhalbert

[adafruit/circuitpython] branch deleted: pwmout\-tcc\-channel\-claiming

deshipu

[adafruit/circuitpython] New comment on issue #1106: pulseio\.PWMOuts share duty cycle values when many of them are created

Awesome, thank you!

[lone sandal]

mew-cx

[adafruit/awesome-circuitpython] New star added

[manic glacier]

mew-cx

[adafruit/circuitpython] New star added

[manic glacier]

tannewt

[adafruit/circuitpython] New comment on issue #1015: Dynamic USB device enabling/disabling

CIRCUITPY just like code.py. You can edit it the same way but it won't be rerun automatically. Instead, one needs to eject the drive and then reset the board (using the button or unplugging).

ATMakersBill

[adafruit/circuitpython] New comment on issue #1015: Dynamic USB device enabling/disabling

I understand how that would be helpful for the HID descriptors, etc. But shouldn't the Mass Storage descriptor be controllable by something not on that drive?

[meager fog]

@umbral dagger hey hey im checkin your guide

would it be possible to get a gif of your or a friend wearing them and tiling your head?

[umbral dagger]

@meager fog Sure thing

@meager fog Where are you thinking of putting it?

@meager fog I replaced the one at the top of the overview page. Let me know if that's good or you want it as the thumbnail instead/as well.

[manic glacier]

dhalbert

[adafruit/circuitpython] Pull request opened: #1185 usb\_write\(\) output\_len was uint8\_t instead of uint32\_t

usb_write() was only using 8 bits to represent buffer length. If buffer was larger, length would be wrong.

Fixes #1181, and might fix some other mysterious USB output bugs.

dhalbert

[adafruit/circuitpython] Pull request opened: #1186 Uart enhancements

• Cherry-pick 02b3f62 UART timeout zero fix from master to 3.x.
• Add UART.in_waiting -- characters in buffer waiting to be read (duplicating PySerial API)
• Add UART.reset_input_buffer() -- clear input buffer, discarding unread characters (PySerial API)
• Remove unused fields in UART object.
• Clear board.{UART,SPI,I2C} singletons on soft reset. Previously a ctrl-D wouldn't clear these objects, and trying to create them again and use them would cause a crash.
• Clean up `#if...

ladyada

[adafruit/circuitpython] New comment on pull request #1186: UART enhancements

awesome. do you want me to test with hardware tonite?

[tulip sleet]

@slender iron hmm sphinx in travis is now 1.8.0, was 1.7.7, and the docs build is failing. will investigate. Might just force the version if necessary

[slender iron]

https://github.com/sphinx-doc/sphinx/issues/5417

on it 😃 can merge anyway

[manic glacier]

tannewt

[adafruit/circuitpython] Pull request closed: #1185 usb\_write\(\) output\_len was uint8\_t instead of uint32\_t

tannewt

[circuitpython:3\.x] 2 new commits

2bd7040 usb_write() output_len was uint8_t instead of u... - dhalbert
beb9446 Merge pull request #1185 from dhalbert/usb_writ... - tannewt

[slender iron]

nrf is failing for the other build

[tulip sleet]

yeah, I forgot there were nrf builds in 3.x; i did do esp8266

have to add one routine

[manic glacier]

ladyada

[adafruit/circuitpython] New comment on pull request #1185: usb\_write\(\) output\_len was uint8\_t instead of uint32\_t

great eyes!

[slender iron]

yup yup, no problem

[manic glacier]

tannewt

[adafruit/circuitpython] Pull request opened: #1187 Add OnDiskBitmap which loads pixel data straight from disk\.

Also, renamed Sprite's palette to pixel_shader so it can be
anything that produces colors based on values (including color values).
Added a ColorConverter that converts RGB888 (found in bitmaps) to
RGB565 for the display.

Fixes #1182

[raven canopy]

@tidal kiln HT16K33 merged. I'll leave the LSM9x PR reviews/merges for @idle owl, so that she can update the learn guides at the same time.

[tidal kiln]

@raven canopy saw that. thanks! one more HT16K33 PR coming in about a minute. 😃

[raven canopy]

👀 👍

@tidal kiln i don't have a segment to test...i'm assuming you did though?

[tidal kiln]

yep

[raven canopy]

either way, it looks like the solution code-wise

k

waiting on Travis...

[tidal kiln]

yah. it's been slow today.

[charred lake]

Hello all

I have a quick question; hopefully this is the right place... I'm using Mu Editor on Raspberry Pi to control my Circuit Playground Express... I don't see a Plotter button on my Mu Editor, is it because I'm using the Raspberry Pi or what? I have Mu Editor 1.0.0-1~rpt1 (if this helps)

[tidal kiln]

@raven canopy thanks!

[manic glacier]

tannewt

[adafruit/circuitpython] Pull request closed: #1186 UART enhancements

tannewt

[circuitpython:3\.x] 4 new commits

02b3f62 When UART timeout of zero is given, make read()... - PaulKierstead
6a046f5 UART fixes and enhancements; default board obje... - dhalbert
6a72084 fix nrf builds; sphinx 1.8.0 crashing: use lowe... - dhalbert
2dd9407 Merge pull request #1186 from dhalbert/uart-enh... - tannewt

[acoustic pollen]

is there a function in circuit python for the cpx for it to detect a sudden stop in acceleration?

[tough flax]

Thank you all for you help this week!

[charred lake]

Well, I got my answer, it's def due to PyQtChart not on Raspberry Pi, sigh, that's a bit sad

[tidal kiln]

@charred lake bummer. but that's what i'd figure also.

@acoustic pollen there's shake detect that might work

@acoustic pollen https://github.com/adafruit/Adafruit_CircuitPython_CircuitPlayground/blob/master/adafruit_circuitplayground/express.py#L221

[charred lake]

@cater Yeah, I did some research reading through github after some frustration. Hopefully PyQtChart can be added in the future to Raspberry Pi. Perhaps there's a workaround, I shall try to find one.

[compact hornet]

My Feather M4 just arrived. I'm always a little shocked at how small these things are. 😃

[charred lake]

Nice! Always exciting to get a new toy

[ruby lake]

hah, well THAT will do it. My MIDICV shield thinks SPI is in "arduino position" on the headers.

this being a metro m0

[manic glacier]

dhalbert

[adafruit/circuitpython] New comment on pull request #1187: Add OnDiskBitmap which loads pixel data straight from disk\.

Initialize self->needs_refresh in common_hal_displayio_group_construct() ?
(or are you depending on it starting as zero?)

[charred lake]

@tidal kiln HAHA! I got the Plotter (on Mu Editor) to work on Raspberry Pi!!! Yeeehaaaawwww - After hours and hours of research.

[tidal kiln]

@charred lake nice! was there an issue in the mu repo for that? you should post how you solved it.

[charred lake]

@tidal kiln I can't take full credit for it, but I can help facilitate where to find the solution (and what NOT to do). It has nothing to do with the mu repo though; it's just that Raspberry Pi doesn't have PyQtChart nor any packages of it for an easy install. So you have to install a few dev packages and manually compile/build a few others to get it to work.

You also have to make sure you build SIP 4.18.1 and not any other version (doesn't work with any newer versions)

That's a small price to pay, I'm so happy Plotter is now working on my Raspberry Pi!

[ruby lake]

ok, so the question becomes, can I use digital pins D11 and D13 as MOSI and SCK and have circuitpython form a hardware SPI port on them?

(on the Metro M0 Express)

[raven canopy]

@ruby lake looking at the mux real quick...

@ruby lake they both have SERCOM available, so you should be fine.

[ruby lake]

@raven canopy yeah it runs a but better now using the D11/D13 SPI 😉

~bit

there are still a couple glitches but I'll sort those out

[manic glacier]

hathach

[adafruit/circuitpython] Pull request opened: #1188 Nrf52 qspi flash

It is still work in progress but current code works with 8 MB MX25R6435F on pca10056. The read speed is ok, but the write is painfully slow (will improve later). I would like to have as many review/suggestion as possible.

• added flash_api directory and move all the flash there
• flash_api.c implement generic read/write blocks API with page caching along with python glueing.
• flash_api use flash_hal_erase/read/program which are implemented by internal or qspi flash.
• flash hal will...

hathach

[adafruit/circuitpython] New comment on pull request #1188: Nrf52 qspi flash

the awful write speed really need some improvement !!

[manic glacier]

yk9

[adafruit/circuitpython] New comment on issue #1005: SPI baudrate of 100000 does not work reliably

im also using the nrf52832 with the STMPE610 . i can read the ID but im getting no data from the registers.
anyone had this problem?

jerryneedell

[adafruit/circuitpython] New comment on issue #1005: SPI baudrate of 100000 does not work reliably

Are you using the latest version of the stmpe610 library - the default baud rate was changed to 1MHz. Can you post and example of the code you are using?

yk9

[adafruit/circuitpython] New comment on issue #1005: SPI baudrate of 100000 does not work reliably

i think im using the last version , yes the baud is 1M
here is the example of the init:
void _SMTPE610_init(void)
{
uint8_t d ;
ts_hardware_init();
writeRegister8(SYS_CTRL1, SYS_CTRL1_RESET); // 0x03 / 0x02
for ( d=0; d<65; d++)
{
readRegister8(d);
}
writeRegister8(SYS_CTRL2, 0x00); // Enable TSC and ADC Blocks
writeRegister8(TSC_CTRL, 0x04); // X, Y mode
writeRegister8(TSC_CTRL, 0x05); // Enable Device, No tracking index,...

jerryneedell

[adafruit/circuitpython] New comment on issue #1005: SPI baudrate of 100000 does not work reliably

ah - you are using the Arduino IDE -- My experience has been with the CircuitPython Library. I have used it on an nrf52832 with CircuitPython.

yk9

[adafruit/circuitpython] New comment on issue #1005: SPI baudrate of 100000 does not work reliably

not arduino - IAR

jerryneedell

[adafruit/circuitpython] New comment on issue #1005: SPI baudrate of 100000 does not work reliably

Ah - sorry --- Still , the discussion above was with respect to the CircuitPython implementation.

jerryneedell

[adafruit/circuitpython] New comment on issue #1005: SPI baudrate of 100000 does not work reliably

I really am not sure what your code is doing and what you mean by

i can read the ID but im getting no data from the registers.

does readregister8(d) -- display the value?

yk9

[adafruit/circuitpython] New comment on issue #1005: SPI baudrate of 100000 does not work reliably

yes im getting some values . but its the initialization

jerryneedell

[adafruit/circuitpython] New comment on issue #1005: SPI baudrate of 100000 does not work reliably

Are you using the SPI or I2C interface? The SPI can be tricky since it can sometime use either clock edge when it initializes. the the cod in the Arduino Library detects this and switches modes:
see https://github.com/adafruit/Adafruit_STMPE610/blob/master/Adafruit_STMPE610.cpp
I'm not sure if that is related to your problem or not. - You may want to post your questions to that repository to see if anyone has any suggestions.

jmroceaneer

[adafruit/circuitpython] Fork created: jmroceaneer/circuitpython

[fierce girder]

Howdy, I'm Drew Fustini, the maintainer of Adafruit_BBIO and @meager fog suggested I look at getting CircuitPython drivers to work on the BeagleBone, similar to the work that has been done for Raspberry Pi.

[manic glacier]

dhalbert

[adafruit/circuitpython] New comment on issue #1128: uart 4\.0

Added a to-do list to initial comment. UART.in_waiting and UART.reset_input_buffer() implemented by #1186.

dhalbert

[adafruit/circuitpython] New comment on pull request #1186: UART enhancements

Adds some features requested in #1128.

[manic glacier]

yk9

[adafruit/circuitpython] New comment on issue #1005: SPI baudrate of 100000 does not work reliably

using SPI. im controlling the SS so data is control. since im getting the correct data (ID and VER) i guess that the SPI is correct. am i wrong?

jerryneedell

[adafruit/circuitpython] New comment on issue #1005: SPI baudrate of 100000 does not work reliably

Agreed, If you can read the ID then it is not clear why you can't read other registers. I'm not sure what else to suggest.
I suggest posting this in the Adafruit Forums https://forums.adafruit.com/viewforum.php?f=19

From what you have posted, it is still not completely clear to me what is not working. Can you post an example of the code that is not working?

[tulip sleet]

@fierce girder Great! Have at it and let us know how we can help.

[fierce girder]

Thanks!

[tulip sleet]

@slender iron I have prepared release notes and a blog post for 3.0.2 and will await Limor's final testing. Also will merge 3.x changes into master when it's cooked.

[slender iron]

sounds good @tulip sleet !

[meager fog]

@ruby lake for D11/D12/D13 yes i think i put them purposefully on a SPI sercom

@fierce girder hiii

im here!

[slender iron]

welcome @fierce girder !

[meager fog]

thanks for comin' by

i have to get my BBB going

im a bit under the weathe today, but i dont expect bumpy codin'

wanna start by forkin blinka?

[manic glacier]

tannewt

[adafruit/circuitpython] New comment on pull request #1188: Nrf52 qspi flash

Write speed isn't critical to using CircuitPython. Yes, it'd be nice if it was faster but let's ignore it for now and focus on getting all of the APIs going.

I suggest filing a separate issue for it for the full 4.0.0 release milestone and moving onto other tasks for now.

[meager fog]

ok i just grabbed my FoB BBB

@fierce girder BeagleBoard.org Debian Image 2016-01-24

[fierce girder]

Sorry to hear you aren't feeling well.
Do you mean fork this repo?
https://github.com/adafruit/circuitpython

[meager fog]

nope

this one 😃

https://github.com/adafruit/Adafruit_Blinka

[ruby lake]

@meager fog Yeah it works much better now 😉

[meager fog]

yeah i also put an i2c sercom on a4/a5

mabye thats the m4 metro tho

m0 i stuck to zero compat

drew, is there a quickstart for upgrading to latest

like 'sudo apt-get upgrade'?

[fierce girder]

@meager fog I would recommend grabbing latest image from http://beagleboard.org/latest-images

Debian 9.5 2018-08-30 4GB SD IoT

[meager fog]

thx

downloooooadin

[fierce girder]

Earlier this we made the change to handling device tree overlays in the bootloader

[meager fog]

np ok will burn this to a 4GB card and stick into tf slot

[fierce girder]

This simplified many ways in which users interact with the pin modes and peripherals

[meager fog]

ok cool

[fierce girder]

Yes, I usually run from microSD, instead of eMMC, as I'm always updating to our weekly bb.org snapshot images

[meager fog]

ok i will try it

ok ya forked?

[fierce girder]

Fyi https://elinux.org/Beagleboard:BeagleBoneBlack_Debian#Where_did_the_slots_file_go.3F

Yes, forked https://github.com/pdp7/Adafruit_Blinka

[meager fog]

gr8

now git clone that to your BBB

im behind ya cause im still burning this SD

is there an easy way to set up samba for editng on desktop

(i know how to do it on pi, but maybe BBB has something more elegant)

[fierce girder]

I'm not too familiar... I've always just edited in vim via ssh

@meager fog you will need to update the bootloader on the eMMC
https://elinux.org/Beagleboard:BeagleBoneBlack_Debian#U-Boot_Overlays

Can you boot the BeagleBone from eMMC while you waiting for the flashing SD card?

Run this /opt/scripts/tools/developers/update_bootloader.sh

The reason is that the SoC ROM bootloader looks at eMMC first. If you have a pre 2018 uboot on your eMMC, then it will cause uboot overlays not to work

Despite the fact that a new image is on the microSD card

The alternative is once booted from the new SD card, you could run dd to erase the bootloader partition on the eMMC. The SoC ROM bootloader will then skip the eMMC uboot and go directly to microSD uboot and uboot overlays will work OK.

[meager fog]

hi back, sorry gotta take litl breaks

ok sd burned

ill run bootload update first

[fierce girder]

Fyi, when running from SD, this erases the eMMC's version of u-boot:
sudo dd if=/dev/zero of=/dev/mmcblk1 bs=1M count=10

[meager fog]

ok well its been a while so ill just do whacha recommend

Are you 100% sure, on selecting [am335x_evm] (y/n)? y
-----------------------------
log: dd if=/tmp/tmp.HcSbEPex7d/dl/MLO-am335x_evm-v2018.09-r2 of=/dev/mmcblk0 seek=1 bs=128k
0+1 records in
0+1 records out
90932 bytes (91 kB) copied, 0.00391849 s, 23.2 MB/s
log: dd if=/tmp/tmp.HcSbEPex7d/dl/u-boot-am335x_evm-v2018.09-r2.img of=/dev/mmcblk0 seek=1 bs=384k
1+1 records in
1+1 records out
447372 bytes (447 kB) copied, 0.0182809 s, 24.5 MB/s
-----------------------------
Bootloader Updated
Bootloader Recovery Complete

ok done reboot with uSD in slot?

[manic glacier]

hathach

[adafruit/circuitpython] New comment on pull request #1188: Nrf52 qspi flash

@tannewt phew, I am trying to increase usb msc buffer size to 8KB but it doesn't help as well. probably need to dive deeper into the flash datasheet timing etc ... I will try to wrap it up now.

[arctic heron]

What would be the best way to discuss a proposal to add write_timeout to UART? An issue?

[fierce girder]

@meager fog yes reboot with uSD

[meager fog]

rebootin

this is the hardest part of portin'

[fierce girder]

After reboot, run:
/opt/scripts/tools/version.sh

[meager fog]

@arctic heron yes please! we have a 'uart 4.0' issue

thats a good place 😃

drew - waitin for usb enumeration...

[arctic heron]

Alright. How to implement it is straightforward enough (follow pySerial, mostly), but how to make it not break current implementations who are thinking the write_timeout will be the same as the read timeout is the trick. I'll try to write something up with a proposal

[meager fog]

paul, thanks - we did just add two new UART elements

in_waiting (inbuffer size) and reset_input() (clear input buffer)

if you wanna try em!

[arctic heron]

Yah, I say that. in_waiting makes sharing code with pySerial a lot easier, and allows non-0 timeouts on read, which can alleviate the problem of no seperate timeout for write

[meager fog]

cool - @tulip sleet can you give paul a build?

its 3.0.2 RC

[arctic heron]

I have a dockerfile that'll make a build np

so long as I know the branch

[meager fog]

@fierce girder hmm i see a COM port but connecting no longer gives me login

[tulip sleet]

@arctic heron Pick a ...2dd9407... build from here: https://adafruit-circuit-python.s3.amazonaws.com/index.html?prefix=bin

[fierce girder]

@meager fog do you have an FTDI Cable handy?

[meager fog]

yah

[tulip sleet]

@arctic heron branch is 3.x

[fierce girder]

@meager fog pins https://elinux.org/Beagleboard:BeagleBone_Black_Serial

[meager fog]

ok i need a pinmap now

[fierce girder]

If it is a problem with eMMC, then holding down the USER BOOT push button will force the SoC ROM bootloader to look only at uSD

If it is problem with uSD, then the serial console should provide some hints

[meager fog]

oh wait theres an ftdi slot

ok hold on

no errors i can see

debian@beaglebone:~$ uname -a
Linux beaglebone 4.14.67-ti-r73 #1 SMP PREEMPT Thu Aug 30 00:08:52 UTC 2018 armv

🤷

but im logged in so whoc ares

[fierce girder]

Please run After reboot, run:
/opt/scripts/tools/version.sh

[meager fog]

ok done

apt-get updatin'

[fierce girder]

Do you have Ethernet connected?

Could you paste the output of version.sh?

[meager fog]

oh hold on

i started installing samba

yah im on ether

root@beaglebone:/home/debian# /opt/scripts/tools/version.sh
git:/opt/scripts/:[73593ebe3b7d3cc381eeb502d45ccb33a6ec5e78]
eeprom:[A335BNLT00C01618BBBK17E0]
model:[TI_AM335x_BeagleBone_Black]
dogtag:[BeagleBoard.org Debian Image 2018-08-30]
bootloader:[microSD-(push-button)]:[/dev/mmcblk0]:[U-Boot 2018.03-00002-gac9cce7c6a]:[location: dd MBR]
bootloader:[eMMC-(default)]:[/dev/mmcblk1]:[U-Boot 2018.09-00002-g0b54a51eee]:[location: dd MBR]
kernel:[4.14.67-ti-r73]
nodejs:[v6.14.4]
uboot_overlay_options:[enable_uboot_overlays=1]
uboot_overlay_options:[uboot_overlay_pru=/lib/firmware/AM335X-PRU-RPROC-4-14-TI-00A0.dtbo]
uboot_overlay_options:[enable_uboot_cape_universal=1]
pkg check: to individually upgrade run: [sudo apt install --only-upgrade <pkg>]
pkg:[bb-cape-overlays]:[4.4.20180803.0-0rcnee0~stretch+20180804]
pkg:[bb-wl18xx-firmware]:[1.20180517-0rcnee0~stretch+20180517]
pkg:[kmod]:[23-2rcnee1~stretch+20171005]
pkg:[librobotcontrol]:[1.0.2-git20180829.0-0rcnee0~stretch+20180830]
pkg:[firmware-ti-connectivity]:[20170823-1rcnee1~stretch+20180328]
groups:[debian : debian adm kmem dialout cdrom floppy audio dip video plugdev users systemd-journal i2c bluetooth netdev cloud9ide gpio pwm eqep admin spi tisdk weston-launch xenomai]
cmdline:[console=ttyO0,115200n8 bone_capemgr.uboot_capemgr_enabled=1 root=/dev/mmcblk0p1 ro rootfstype=ext4 rootwait coherent_pool=1M net.ifnames=0 quiet]
dmesg | grep pinctrl-single
[    1.099695] pinctrl-single 44e10800.pinmux: 142 pins at pa f9e10800 size 568
dmesg | grep gpio-of-helper
[    1.111742] gpio-of-helper ocp:cape-universal: ready
END

[slender iron]

@tulip sleet you ok reviewing #1188?

[tulip sleet]

I am reviewing some of it but if you could look at the QSPI-specifc stuff given your experience that would be great.

[slender iron]

looks like most of it was copied over

[tulip sleet]

yeah, so maybe just qspi_flash.c, which is the actual impl

[meager fog]

@fierce girder ok made a new 'bone' account, set up samba

im ready (finally) to code!

oh wait we gotta setup python3

[slender iron]

@tulip sleet I take it back. its not copied 😦

[pastel panther]

does anyone know if there has been any effort to do something like FadeCandy/FastLED style dithering and/or general led animation helpers for CP?

[tulip sleet]

@pastel panther in C or in Python?

[pastel panther]

In C presumably

[tulip sleet]

https://github.com/adafruit/circuitpython/pull/943 (pixelbuf) is ongoing but not done

[pastel panther]

python would be a good starting point but I imagine it would cap out fairly fast

[tulip sleet]

that's at a lower level, but functionality could be added there

[pastel panther]

Cool I thought that might be a good foundation

[meager fog]

drew, ok add to .bashrc

export PYTHONPATH=/home/pi/Adafruit_Blinka/src:.

[fierce girder]

Ok thanks

[meager fog]

er instead of /home/pi put whatever

so that you at least get this

Python 3.5.3 (default, Jan 19 2017, 14:11:04)
[GCC 6.3.0 20170118] on linux
Type "help", "copyright", "credits" or "license" for more information.
>>> import digitalio
Traceback (most recent call last):
  File "<stdin>", line 1, in <module>
  File "/home/bone/Adafruit_Blinka/src/digitalio.py", line 10, in <module>
    from adafruit_blinka.agnostic import board_id
  File "/home/bone/Adafruit_Blinka/src/adafruit_blinka/agnostic/__init__.py", line 23, in <module>
    from Adafruit_GPIO import Platform
ImportError: No module named 'Adafruit_GPIO'

when you try to import digitalio

[fierce girder]

Should be possible to install Adafruit GPIO from pip

Though I usually install from the repo since I'm doing development

[manic glacier]

tannewt

[adafruit/circuitpython] New comment on pull request #1187: Add OnDiskBitmap which loads pixel data straight from disk\.

Just an oversight. I've added it to displayio_group_construct. Thanks!

[meager fog]

use pip

because we put it in requirements.txt

[manic glacier]

dhalbert

[adafruit/circuitpython] New comment on pull request #1188: Nrf52 qspi flash

@hathach We'd like to make this implementation be as similar as possible to the atmel-samd QSPI and external flash implementation, so that we can factor out the common code as much as possible and only implement the lowest level routines in a chip-specific way. For instance, that code has parameterized chip support, and it already has caching logic. Did you try copying that impl? This impl is more different than we thought it would be.

[meager fog]

drew, now you can run https://learn.adafruit.com/circuitpython-on-raspberrypi-linux/digital-i-o#blinky-time-3-16

and see it breaks

[manic glacier]

tannewt

[adafruit/circuitpython] New comment on issue #1015: Dynamic USB device enabling/disabling

I think its ok. You can change files through the REPL or use the REPL to reboot to safe mode which skips all user code.

A common technique for the similar problem of flipping read-only for the file system is to read an external line connected to a switch to determine whether to do it or not on boot.

[meager fog]

ok heres code to detect BBB

https://github.com/ladyada/Adafruit_Blinka/commit/ae336b5c0f6707cf6a2284c9edff9ce39c07f018

@fierce girder lets add LEDs first

and any user buttons

whats the user led/button pin names

pins arent indicated by number right, just by string name>?

@fierce girder you around still?

ok well ping me when you're back

[fierce girder]

Got disconnected. Back

The pins names are normally refered to by header and pin number

Such as P8 or P9 header, and pins 1-46

[manic glacier]

hathach

[adafruit/circuitpython] New comment on pull request #1188: Nrf52 qspi flash

@dhalbert many of the spi std command for flash is handled by nrfx qspi driver. Since nrf52840 has enough memory (4K) to cache the whole page. I just refactor both caching for internal + external flash giving it to only basic hal erase, read, program. At anywhere you could use block API flash_read_blocks/flash_write_blocks instead of internal/external_flash_write_blocks. The flash_ will use either internal or external depending on the config. Those can be changed to function pointer for conc...

[fierce girder]

@meager fog for example, P9_12 is by default set to GPIO mode

What actually happens though, is there a mapping to the gpio pin number in Linux GPIO subsystem

[meager fog]

ok cool

im back too! with 🍵!

ok for now we'll use PX_XX notation

raspi has 'bcm names' which are the underlying gpio #'s

ok so almost ready for first tests, what method should we use for GPIO togglin

like what python library/binding is best

[fierce girder]

https://github.com/adafruit/adafruit-beaglebone-io-python/blob/master/source/common.c

[meager fog]

in python plz 😃

[fierce girder]

Has the mapping to GPIO from the perspective of Linux

[meager fog]

Adafruit_BBIO ?

yah - we need a way to make python do the gpio

[fierce girder]

Yes, Adafruit_BBIO.GPIO uses those mappings

[meager fog]

we can add more names later

ok

whats a good first pin to try

[fierce girder]

For example

import Adafruit_BBIO.GPIO as GPIO GPIO.setup("P9_12", GPIO.OUT) GPIO.output("P9_12", GPIO.HIGH)

[meager fog]

invalid permz

do i have to be root? is there a way to add me to the group for gpio access

[fierce girder]

No, with newer kernels, udev works

[meager fog]

ok hit me with the udev

[manic glacier]

dhalbert

[adafruit/circuitpython] Issue closed: #1182 Support streaming pixel data off the filesystem

dhalbert

[adafruit/circuitpython] Pull request closed: #1187 Add OnDiskBitmap which loads pixel data straight from disk\.

dhalbert

[circuitpython:master] 4 new commits

f21cf15 Add OnDiskBitmap which loads pixel data straigh... - tannewt
fb833b2 Update translations and limit Sphinx to <1.8.0 - tannewt
75d497b More explicit init in Group - tannewt
101531b Merge pull request #1187 from tannewt/bitmap_disk - dhalbert

[charred lake]

Talking about Beagle Boards, I need to get around to repairing mine, I fried it 😛

[fierce girder]

You run that as a standalone program and what was the output?

[meager fog]

ValueError: Set gpio direction failed, missing file or invalid permissions.

[fierce girder]

Can you try running with sudo to determine that it is in fact permission

[meager fog]

i have a pyenv thing

so its not easy

hold on

yah it runs

with sudo

[fierce girder]

ok

so debian@beaglebone:~$ strace -e file -f python gpio.py 2>&1 | grep /gpio/
access("/sys/class/gpio/gpio60", R_OK|W_OK|X_OK) = -1 ENOENT (No such file or directory)
open("/sys/class/gpio/export", O_WRONLY) = 3
open("/sys/class/gpio/gpio60/direction", O_WRONLY) = 3
open("/sys/class/gpio/gpio60/value", O_WRONLY) = 3
open("/sys/class/gpio/gpio60/value", O_WRONLY) = 3
open("/sys/class/gpio/unexport", O_WRONLY) = 3

[meager fog]

nope nothin comes up

[fierce girder]

do you have this?
debian@beaglebone:~$ ls -ltar /sys/class/gpio/export
-rw-rw---- 1 root gpio 4096 Sep 10 04:21 /sys/class/gpio/export

[meager fog]

yah

[fierce girder]

debian@beaglebone:~$ id debian
uid=1000(debian) gid=1000(debian) groups=1000(debian),4(adm),15(kmem),20(dialout),24(cdrom),25(floppy),29(audio),30(dip),44(video),46(plugdev),100(users),101(systemd-journal),108(i2c),111(bluetooth),112(netdev),115(cloud9ide),999(gpio),998(pwm),997(eqep),996(admin),995(spi),994(tisdk),993(weston-launch),992(xenomai)

so that is in group gpio

[meager fog]

im user bone

[fierce girder]

so that is why the debian user is able to do it

ah... did you create a new accout?

[meager fog]

so just add me to gpio group?

cause i like to test with a clean account

[arctic heron]

I shoulda brought my BBB to work so I could play along 😃

[meager fog]

it helps me find problems, like this one

[fierce girder]

the bb.org images are geared towards the debian user being configured properly

[meager fog]

totally

[fierce girder]

so new user accounts would probably want to be in the same groups

[meager fog]

that is not what other people do

i have to be in the soul of someone who does not follow the default path

[fierce girder]

sure

[meager fog]

gr8 added to guide instructions

[fierce girder]

thanks

[arctic heron]

I doubt it is hard for you to get into the soul of someone who does not follow the default path.

[meager fog]

it is practice

thats how you write good docs

very zen

you have to be the fool that makes things foolproof 😄

@fierce girder much progress, ok are there any gpio connected to LEDs i can toggle

[arctic heron]

I find for experienced devs understanding the default path is often harder.

[fierce girder]

other important groups

debian@beaglebone:~$ ls -latr /sys/class/pwm/pwmchip0/export
-rw-rw---- 1 root pwm 4096 Jan 1 2000 /sys/class/pwm/pwmchip0/export
debian@beaglebone:~$ ls -latr /dev/i2c* |head -1
crw-rw---- 1 root i2c 89, 0 Sep 10 04:21 /dev/i2c-0
debian@beaglebone:~$ ls -latr /dev/spidev* |head -1
crw-rw---- 1 root spi 153, 0 Sep 10 04:21 /dev/spidev1.1

[meager fog]

ok will do those too!

[fierce girder]

gpio, pwm, i2c, spi

[meager fog]

gr8 added to my list

[stuck elbow]

oh, hi Drew!

[fierce girder]

@meager fog https://github.com/adafruit/adafruit-beaglebone-io-python#on-board-leds

GPIO.output("USR%d" % i, GPIO.HIGH)

[meager fog]

lovely!

hold on

ill give you test code momentarililily

oh interesting

i cannot set those pins to be inputs?

are they hardcoded somehow?

[fierce girder]

the USR LED pin?

[meager fog]

yah

  File "gpio.py", line 4, in <module>
    GPIO.setup("USR1", GPIO.IN)
ValueError: Set gpio mode failed, missing file or invalid permissions.

[fierce girder]

well a USR LED is a hardware circuit

[meager fog]

iz a gpio 😃

[fierce girder]

not on P8/P9

[meager fog]

not a gpio?

is it on an expander chip?

[fierce girder]

it is controlled with GPIO.setup() and GPIO.output()

but it is a not on the P8/P9 header

there is no way for it to be an import

USR0-3 are LEDs on the beaglebone

it is special case https://github.com/adafruit/adafruit-beaglebone-io-python/blob/master/source/event_gpio.c#L199

so nice to use since built-in to the hardware

[meager fog]

exciting

ok

[fierce girder]

but not useful as a true GPIO pin

P9_12 is a good pin to wire up an LED or a button

[meager fog]

yeah but this is something we have to handle

hold on

[fierce girder]

http://beagleboard.org/static/images/cape-headers.png

[meager fog]

adding a Thing

thinks

[fierce girder]

P9_12 and P9_15 are good to use for examples since default pin mode is GPIO

technically USR0-3 could be abstracted as LEDs

[meager fog]

thinkingthinking

linux does have this concept of GPIO that have mode restriction

[fierce girder]

Linux kernel uses LED subsystem for them /sys/class/leds/

[meager fog]

i think i will have to add it later

[fierce girder]

versus

[meager fog]

as an exrta option on Pin creation

right now we default all new GPIO to INPUT

[fierce girder]

versus the GPIO subsystem (/sys/class/gpio)

[meager fog]

but it should check what is the default state required

ok ill skip the USR LEDs

good edge case tho

[fierce girder]

yes

[meager fog]

didnt bump into this one with raspi

[fierce girder]

can come back to it... Adafruit_BBIO is based on the sematics of RPi.GPIO which is not great (I think the original author [Ben?] would admit to that... it was a just a quick project back in 2012 😃

the proper abstration for USR0-3 is as LEDs

this is how the Linux kernel treats them

but in the world of Adafruit_BBIO / RPi.GPIO, there was no abstraction for LED

[meager fog]

we dont hae to use BBIO

i want to use libgpiod in the long run

since the way gpio is done in linux is a crime

[fierce girder]

yes, for GPIO, Adafruit_BBIO will need to move to the new character device, and the python bindings for libgpiod make sense to use

[meager fog]

ok well im blinking an led

wanna try?

[fierce girder]

although Adafruit_BBIO is a collection of modules beyond just GPIO

The biggest issue potentially with Adafruit_BBIO is basically just a C extensions

is that an acceptable dependency for CircuitPython?

[meager fog]

yeah i mean as long as we can pip it

i dont care

we use rpi_ws281x for neopixel support on raspi

and thats really terrible C code

[fierce girder]

Ultimately, I think Adafruit_BBIO.GPIO should transition to libgpiod bindings. But there is other functionality in Adafruit_BBIO. ADC and PWM being the most important modules after GPIO.

[meager fog]

drew - review &ill merge https://github.com/adafruit/Adafruit_Blinka/pull/36/files

[fierce girder]

will do

[meager fog]

i just invited u to the repo

[fierce girder]

thanks

[meager fog]

htne you can fastforward your fork

and try it

then we have to add the GPIO pin name bindings

but thats ez, just a lot on BBB

i2c and spi should just work once we get the names in

[fierce girder]

this is the test?

src/adafruit_blinka/board/beaglebone_black.py

[meager fog]

nope

thats the code

test

this is the test code

but it wont work until i do the merge

and i cant assign you as s reviewer until you accept the invite

[fierce girder]

Ok thanks

[meager fog]

let me know when you accept it

[manic glacier]

tannewt

[adafruit/circuitpython] New comment on pull request #1188: Nrf52 qspi flash

There is additional logic in the atmel-samd port that isn't in this one because it wasn't started from a copy. For example, the SPI flash chip is detected rather than hard coded. Ideally this logic would be shared but for now it can be copied if that's easier.

Ideally you'd only need to replace the implementations [here](https://github.com/adafruit/circuitpython/blob/master/ports/...

[fierce girder]

Ok

Thanks... was dealing with some open hardware summit badge issues momentarily

[manic glacier]

hathach

[adafruit/circuitpython] New comment on pull request #1188: Nrf52 qspi flash

@tannewt ah yeah, I am adding support device lookup table with JEDEC manufacturer and device ID. Instead of using 3 bytes response from 0x9F for manufacture id, memory type and capacity, can we use only 2 bytes response from 0x90 command (manufacture id, device id) ?

[slender iron]

@gentle bronze why do it differently than the samd port?

[gentle bronze]

@slender iron you mean the whole implementation

?

[slender iron]

all but the mcu specific stuff

[meager fog]

@fierce girder np leme know when you've accepted invite

[fierce girder]

accepted

[gentle bronze]

nrf got hardware qspi, it doesn't need to do at spi command except for a few for initialization.

I ended up with something 80% like internal_flash

so I just refactor them to have common flash caching for both internal & qspi flash

[meager fog]

@fierce girder ok review it & approve if you agree 😄

[slender iron]

@gentle bronze I'd rather have one implementation than two even if that means you need to use a lower level nrf api

I don't think you'd need to though

[gentle bronze]

You mean the internal/external _flash_read/write API ?

I would seperate them without issues

[slender iron]

@gentle bronze I expected you to copy internal_flash.*, flash_api.* and external_flash into the nrf port, delete the samd specific parts and replace it with nrf. that way we can factor out the common parts later to share between the ports

[gentle bronze]

I would rename the file and exported API, though, the implementation for nrf is much simpler than samd port in this case, we don't really need to have many stub. When refactoring out, we still only reply on API

[slender iron]

I realize I was never explicit about that though

[gentle bronze]

yeah, if there is share binding, I would follow that

[slender iron]

I'd rather have one more complicated API thats shared than two different APIs even if one is simpler

I want to get to the point where its like shared-bindings but it isn't yet

[gentle bronze]

please make it to share binding, I will know what to follow.

[slender iron]

For now, can you copy the files I listed over and not change function definitions? That will make it easier to refactor later

[gentle bronze]

I will keep the following block API

void external_flash_flush(void);
bool external_flash_read_block(uint8_t *dest, uint32_t block);
bool external_flash_write_block(const uint8_t *src, uint32_t block);

void internal_flash_flush(void);
bool internal_flash_read_block(uint8_t *dest, uint32_t block);
bool internal_flash_write_block(const uint8_t *src, uint32_t block);

[slender iron]

thats way higher than it needs to be though

[gentle bronze]

lower than that is port specific already

[slender iron]

I'm thinking spi_flash_command, spi_flash_read_command etc

https://github.com/adafruit/circuitpython/blob/master/ports/atmel-samd/external_flash/spi_flash_api.h should be enough

I see a couple minor things that do need to be factored out of external_flash.c

https://github.com/adafruit/circuitpython/blob/master/ports/atmel-samd/external_flash/external_flash.c#L50

and

[fierce girder]

@meager fog the hello blinky test works. I have reviewed and approved

[slender iron]

https://github.com/adafruit/circuitpython/blob/master/ports/atmel-samd/external_flash/external_flash.c#L243

but the rest should work

[gentle bronze]

@slender iron I could implement then spi_flash_api.h since it is an abstraction, it would be easy enough.

[slender iron]

ok cool!

[gentle bronze]

the external_flash.c is not needed

[slender iron]

yes it is

[gentle bronze]

it can be done in much simpler way in nrf port

[slender iron]

it does caching and device discovery

we can simplify it later once its shared

[meager fog]

@fierce girder awesome. ok next up

can you assist in adding all the GPIO?

and testin' em

meanwhile i can look at i2c

[fierce girder]

@meager fog sure, I will take a look at that

[meager fog]

sweet, thanx

i may need a hand for the multiple i2c ports

we shall seee

[gentle bronze]

@tannew when you move it to shared, I will change to adapt it. Since many in the external_flash.c already implmented by nrfx driver

[tulip sleet]

@gentle bronze is the point that the actual peripheral is more capable, or is it that the nrfx library is higher level?

[gentle bronze]

the hw prph

@tulip sleet you configure register, it will do the whole read / write data with wren and all the spi level

[tulip sleet]

do you think the slow write is due to the periph or the library or not sure?

[slender iron]

it handles all of the write enable stuff

[gentle bronze]

@tulip sleet the write speed may need to enabled high performance mode or something, it is probably flash device specific config. I did benchtest CP on CPX, the write speed is also around 3x kbps

[meager fog]

@violet stone ping!

[violet stone]

@meager fog pong!

[meager fog]

yay

you just need a cute puppy icon

ok grab yer htu21d

and a metro m4

or other m4 board

[violet stone]

1 sec

[meager fog]

puppy!,!

[gentle bronze]

@tulip sleet @slender iron the nrfx hw prph already done lots of spi cmd level for us. I think we shouldn't implement the lower level spi command based on it, then go one turn around with software !!

SPI command can be useful for generic configuring such as read JEDEC ID etc ..

[tulip sleet]

@slender iron @gentle bronze do you think we might go the other way with spi flash: modify existing atmel-samd Api to be closer to nrf periph, and then implement as necessary in sw for atmel-samd and use nrfx direct for nrf

[violet stone]

I have to solder some headers on the HTU21d

[meager fog]

kk i wait!

[gentle bronze]

@tulip sleet I have no ideas, didn't work with atmel chip much.

[tulip sleet]

maybe we are missing a slightly higher level api in the current samd impl

[slender iron]

@tulip sleet I'm fine reorganizing it later but for now I want it all copied and edited. I don't want them to be different

[gentle bronze]

I think we only need to keep the API clean and clear for all ports.

[tulip sleet]

well i mean that if currently there are say two calls necessary in the atmel api that the nrf can (or must) do in one, then forcing the nrf to do something it shouldn't (or can't?) is weird

[slender iron]

it can always ignore incoming commands if it wants

[tulip sleet]

so if the nrf periph takes care of, say, chip-select, and tjat

tjat

[slender iron]

if it wants to manage write enable you can return early from spi_flash_command

[tulip sleet]

that's broken out in atmel, then just skip that for now, but should be commented that we should refactor

and the write-enable calls are no-ops in nrf?

[slender iron]

they can be if we want to let the peripheral manage it

later we can reorganize it if we want. but its used for normal spi as well

[tulip sleet]

i don't have a detailed enough picture to have a strong opinion right now. @gentle bronze maybe sleep on this - you are up too late as it is

[gentle bronze]

oh yeah, thanks. I will get some sleep 😄

good night

[slender iron]

good night @gentle bronze !

[tulip sleet]

good night!

[arctic heron]

I don't want to stir up a discussion that has almost certainly beat to death (but before I looked), is there anywhere you would consider a good discussion on a plan -- or plan not to -- merge with micropython?

anywhere -> discussion forum, etc

I just want to read, not have the discussion

[slender iron]

we don't have any immediate or medium term plans

and I don't know of where we've discussed it

the projects have different goals

[gentle bronze]

only want to say, nrf hw is currnetly doing higher level API than software in samd port. We should have a higher level software API for both, rather than forcing nrf to follow lower level API

[tulip sleet]

@gentle bronze ultimately yes, i think that's the right solution. with some sleep maybe this will become clearer

[slender iron]

I regularly talk with damien and offer to help upstream anything he is interested in merging

[meager fog]

@violet stone soldered?

[gentle bronze]

💤

[violet stone]

@meager fog Just wiring it up

[meager fog]

k are you using a metro?

[violet stone]

the itsybitsy m4 I just ordered

[meager fog]

ok rad, update to latest 3.0 release too!

[violet stone]

@meager fog Dragged the .uf2 onto the drive, and reset, but file's still there and boot_out.txt still says 3.0.0-beta.0?

[meager fog]

hmm

you dragged onto the ITSYBOOT right?

also check the terminal output for sure

i like Mu because it autodetects the com port

[violet stone]

@meager fog K, updated

Download mu as well, though I'm pretty much wedded for life to Atom at this point 😛

[arctic heron]

i'm hard core hours-a-day pycharm user, still find mu pretty useful. Often use pyCharm with a deployment target which copies to the CPX (etc) and then use MU to tweak and debug

[violet stone]

@arctic heron Cool, just firing it up now to have a look

[manic glacier]

arofarn

[adafruit/circuitpython] Fork created: arofarn/circuitpython

[meager fog]

@violet stone ok hav eyou done circuitpython essentials?

https://learn.adafruit.com/circuitpython-essentials

[violet stone]

Looking through it now

[meager fog]

maybe just digitalio and https://learn.adafruit.com/circuitpython-essentials/circuitpython-i2c for now

[violet stone]

@meager fog The output should show up in the REPL window post save, right? Not seeing anything.

[meager fog]

yeah should

type control D?

[violet stone]

Nothing

[meager fog]

did you click in repl first?

odd, i see repl carets

if you type in there does it echo?

[violet stone]

It did before I saved, restarting

K, after a reset works

[meager fog]

hmm

yay

ok see i2c scan?

[violet stone]

Sees the sensor

[meager fog]

gr8

youre done

time to write lib

[violet stone]

_ pats self on back_

[meager fog]

ok go here

https://learn.adafruit.com/creating-and-sharing-a-circuitpython-library

log in, do you see my hidden checklist

[violet stone]

yup

[meager fog]

ok try following it lemme knwo if you have q

the guide is good to read when you have time

[violet stone]

👍

[meager fog]

but the checklist will get you going

once you have the cookiecut and commit

look at the si7021 library

its nearly the same

[violet stone]

K ... easy sensor so shouldn't be too complicated

[meager fog]

hardest part is getting structure and travis going

[violet stone]

Cool, I'll poke at it a bit and ping if my brain explodes

[meager fog]

@slender iron hi imma test out the hallowing tft bitmap display code

[slender iron]

👍

[meager fog]

is latest s3 the latest build?

[slender iron]

should be

[meager fog]

@slender iron ok got it loaded, next up - is your example sketch with the auto-shuffle somehwere

i found the rtd example

[slender iron]

https://gist.github.com/tannewt/2c898283d8566bb8fe2f264fc9c3ef47

[meager fog]

ok changed to

print(images)```

works gr8

[slender iron]

great!

[manic glacier]

arofarn

[adafruit/circuitpython] Pull request opened: #1189 Add VOLTAGE\_MONITOR pin to feather boards pins

Issue : https://github.com/adafruit/circuitpython/issues/1149

I choose VOLTAGE_MONITOR like the Feather M4 Express, over BATTERY like the Hallowing.
VOLTAGE_MONITOR is longer look less ambiguous to me (for me the BATTERY pin could be the pin where I connect a battery, not the pin to monitor it) but I did not change the hallowing to avoid breaking anything.

I only tested on my Feather M0 Express.

I didn't change the pins of boards I didn't found in the learning system guides.

[meager fog]

@fierce girder ok when do you want to have the remaining GPIO added by?

[fierce girder]

I'm taking care of some Open Hardware Summit badge issues right now, but I will fill in the rest of the GPIO pins this evening. thanks

[manic glacier]

dhalbert

[adafruit/circuitpython] New comment on pull request #1189: Add VOLTAGE\_MONITOR pin to feather boards pins

In all of these, please move this line so it's just under the D_ name of the pin (usually D9), instead of somewhere else (e.g., see D1/TX) example.

dhalbert

[adafruit/circuitpython] New comment on pull request #1189: Add VOLTAGE\_MONITOR pin to feather boards pins

It's called BATTERY (inconsistently) on the Hallowing because the board is labeled "BAT".

[meager fog]

@fierce girder kk i can test/review once you have a PR!

then we'll do SPI/I2C and others

[tulip sleet]

@slender iron the common_hal_..._configure_... routines don't throw ValueError with a msg when they get a param that won't work. Instead they just return false, and the shared-binding routines throw MP_EIO, which is kinda uninformative. Whaddaya think about throwing more informative msgs? E.g. nrf doesn't support 16-bit SPI, and only supports >8MHz spi on one instance. I could report useful msgs in that case.

not sure if there was a strong reason just to return false vs throw

[rustic nymph]

So my Trinket M0 when I first plug it in the RGB turns green, then orange, then white

but if I use the reset button it turns green and stays green

[tulip sleet]

@rustic nymph is it blinking multiple times on the orange and white? That indicates an error at a certain line number. What's in main.py or code.py?

[rustic nymph]

It went solid green, solid orange, solid white

I think maybe I had a conflicting code.py

one was code - copy.py

[tulip sleet]

it doesn't run any other files besides code.py,main.py, code.txt, main.txt, code.py.txt, main.py.txt

[rustic nymph]

main.py was also in root but it was the original file

[tulip sleet]

(latter 4 are for people who have trouble with file extensions)

code.py takes prio over main.py if it exists

[slender iron]

@tulip sleet ideally common_hal impls would be micropython free so an error code would be better

[tulip sleet]

we have lots of mp_raise's in common_hal, but mostly in the constructors

[rustic nymph]

ok still doing it, green for about 5 seconds, orange, off, white

[slender iron]

ya, thats why I said ideally 😛

[rustic nymph]

https://pastebin.com/3vkUtWK2

[tulip sleet]

not sure why ideally though - i mean, common-hal is for pyhton impl, so not sure the motivation

@rustic nymph what version of CircuitPython are you using?

[rustic nymph]

Adafruit CircuitPython 3.0.1 on 2018-08-21; Adafruit Trinket M0 with samd21e18

https://www.youtube.com/watch?v=RXA3bsGyLXY

[slender iron]

ideally shared-bindings would handle all of the python stuff. having raises in common-hal makes it harder to use it in the vm

er, outside the vm

[tulip sleet]

that i understand, yes, ok, I'll think about error codes in this case

@rustic nymph does it run fine after pressing reset?

[rustic nymph]

yes

[tulip sleet]

put a time.sleep(1) before kbd=Keyboard()

[slender iron]

can you not see the error over serial?

[tulip sleet]

they could, theoretically, but it only happens on power-up not on reset, so hard to connect to repl immediately on pwoer up

[rustic nymph]

still does it with that sleep

I changed it to sleep(15) and its still doing it in the timeframe in the video

[tulip sleet]

is there an ir sensor on the back? I don't see antyhing conencted

[rustic nymph]

yes

gnd-3-4

[tulip sleet]

which sensor?

[rustic nymph]

I use 4 to power it

[tulip sleet]

OOH

[rustic nymph]

http://a.co/d/fF32urJ

[tulip sleet]

hold on

don't use 4 to power it, use the 3v pin. it's not ready fast enough when you power it via 4. Also, the pin may not provide enough current

then you don't need sleep

you could put a sleep after turning on pin4, but better to power it from 3V

[rustic nymph]

I have this and it still goes white immediately - https://pastebin.com/RVPLbgGF

and the light on the IR receiver never lights up from usb plug in - white

[tulip sleet]

you'd have to put a sleep after powerPin.value = True, and you don't need any of the others. But like I said, use the 3V pin in preference

[rustic nymph]

why after the pin going true?

[tulip sleet]

because you aren't supply any power to the IR sensor until you bring pin D4 high, so it isn't even on, and it probably takes a bit of time to turn on. when you press reset it's already been on so it may recover faster (e.g. if there's a cap on the IR sensor board)

[rustic nymph]

sleep of 5 didnt work

[tulip sleet]

on line 36?

[rustic nymph]

https://pastebin.com/fPUbwrYi

[tulip sleet]

if you go into the repl and type ctrl-C and then type ctrl-D, does it work?

[rustic nymph]

Mu says it cant connect to device on that com port when its white

[tulip sleet]

so press reset and then connect, and then be in the repl

[rustic nymph]

that works

[tulip sleet]

you mean you send stuff to the IR sensor and you get output?

[rustic nymph]

yes

[tulip sleet]

then I'm not sure - what OS?

[rustic nymph]

windows 10

[tulip sleet]

the video of orange and white is hard to interpret. It's green, then orange, then white. Could you describe the sequence with gaps more explicitly? Does it get stuck on white?

[rustic nymph]

yes stuck on white

doesnt change after that

[tulip sleet]

The colors mean this: https://learn.adafruit.com/welcome-to-circuitpython/troubleshooting#circuitpython-rgb-status-light-18-10

[rustic nymph]

i rewired it to use the 3v pin and it still does it

[tulip sleet]

... (good on power) I don't immediately see anything wrong. Maybe comment out the keyboard stuff and just do print's, and see if it still hangs. Possible it's an HID problem.

[rustic nymph]

nope

[tulip sleet]

could you pastebin the code now?

[rustic nymph]

https://pastebin.com/MdwYXJWF

[tulip sleet]

take out the extraneous sleeps, line 2, 33
you still have layout.write() not commented out line 60, 63, etc.

i think you could get rid of the pulses.clear() and pulses.resume(). I don't see those used in examples.

[rustic nymph]

https://pastebin.com/q6jhU98r

still not working, I rewired it to use usb for power

[ruby lake]

hm, channel outputs still glitchy, I figure I have something broken in the assigner code

[tulip sleet]

if usb power, you might see 5v on sig pin, no? That's not good for the input pin

line 63 is not commented out

[pastel panther]

@ruby lake what are you working on?

[rustic nymph]

still nothing on fresh plugin

[ruby lake]

@pastel panther 4-channel MIDICV, it has worked for years (decades?) using C, and assembly fir the Z8 stuff

[pastel panther]

@ruby lake are you reviving an old project or upgrading it?

[ruby lake]

@pastel panther Making a CP version

[pastel panther]

nice; using your module playground board?

[tulip sleet]

@rustic nymph sorry, I'm at a loss now. If you disconnect the SIG line to 3, does it still fail? Trying to figure out if it's code or the sensor.

[rustic nymph]

@tulip sleet I deleted everything in code.py, on fresh plugin the light pulses green a few times then goes white, on reset it stays pulsing green

[ruby lake]

eventually, doing tests on my Arduino shield first, since I had that laying around

[rustic nymph]

@tulip sleet ive disconnect the sensor entirely at this point

[ruby lake]

when I make the module playground v2 I will use an 8-channel DAC directly, and get rid of the tdm mux and s/h amps

[tulip sleet]

@rustic nymph and it still fails on power-up? OK that's very helpful.

post the latest version of the code and I'll try it here on another board

[rustic nymph]

@tulip sleet I deleted everything in lib folder, cleared code.py entirely, and it pulses green twice then white on fresh boot

[pastel panther]

@ruby lake Nice; have you thought about building in an m4 instead of using a feather? I wouldn't mind giving one a go if you're interested.

[tulip sleet]

there's no code.py and no main.py?

[ruby lake]

I might do that at some point

[rustic nymph]

the contents are - .fseventsd (empty), lib (empty), .metadata_never_index, .Trashes, boot_out.txt, code.py

there is code.py but its empty

[tulip sleet]

that makes sense, it runs code.py (green), and then drops into the repl (white)

[rustic nymph]

ok default main.py goes white almost immediately

https://learn.adafruit.com/adafruit-trinket-m0-circuitpython-arduino/downloads

from there

[tulip sleet]

put

while True:
    pass

into main.py

[rustic nymph]

fresh boot = green, orange, off, white

reset = green

now notepad++ says the file or directory \main.py is corrupt and unreadable

[tulip sleet]

oh, you are using notepad++, that's not a good choice unless you're ejecting after each write. It's doesn't write the file out fully each time

get into the repl, do

import storage
storage.erase_filesystem()

what other editors do you have?

[rustic nymph]

Ill use Mu now

[tulip sleet]

yes, that's safer

[rustic nymph]

ok I did that through Mu

[tulip sleet]

with no main.py it should pulse green when plugged in

with an empty main.py it should also pulse. With an infinite loop in main.py it should be solid green.

if you were making all those edits in Notepad++, you may not have been testing what you thought you were testing. esp pressing reset a lot

windows thinks it knows what's on the drive, but it doesn't write out stuff for 10's of seconds, so what it thinks and what's really on the drive are not the same

see https://learn.adafruit.com/welcome-to-circuitpython/creating-and-editing-code#1-use-an-editor-that-writes-out-the-file-completely-when-you-save-it-6-13 and the paragraphs before that link

[rustic nymph]

so it did pulse after doing the storage.erase

I rebuilt the original file using Mu and saved it, now doing the green, orange, white thing again

[tulip sleet]

with sensor not connected?

[rustic nymph]

correct

[tulip sleet]

ok, drop that file here and I'll try it on my trinket

[rustic nymph]

also doing it with a code.py file made in Mu containing the while True: pass

[tulip sleet]

oh really !!

ok, then that's the better test case

[rustic nymph]

Ok did storage erase again, pulsing green light

[tulip sleet]

i just did main.py while True pass and get solid green on plug-in

[rustic nymph]

fresh plugin, pulses green a few times then white

[tulip sleet]

with empty filesysteM?

[rustic nymph]

yes

well with the while True: pass in code or main

https://i.imgur.com/yH0h2qj.png

[tulip sleet]

are you running any monitoring programs like AIDA64?

[rustic nymph]

no

[tulip sleet]

what's in boot_out.txt?

[rustic nymph]

Adafruit CircuitPython 3.0.1 on 2018-08-21; Adafruit Trinket M0 with samd21e18

[tulip sleet]

is this on a hub or straight to computer?

[rustic nymph]

straight

[tulip sleet]

when you say "white" you mean pure white, not greenish?

[rustic nymph]

well id saying it has a pink tint?

[tulip sleet]

ok, do you have a micro-USB phone charger or something you can try powering it with, or another non Windows computer?

[rustic nymph]

tried a different cable

did storage.erase_filesystem

pulses green a few times then white again

[tulip sleet]

don't write a main.py, just trying plugging into a power supply or another computer. Shouldn't go white, should pulse green indefinitely. If it goes white on a power supply, I'd say it's a hardware issue

you might try measuring the 3v pin to see if it's 3.3v, but if it is, then something is fried in a weird way

[rustic nymph]

I think its dead

I created a main.py with the while: True pass

did ctrl-c, ctrl-d with no issue

I removed power, plugged it back it and now main.py is {

[tulip sleet]

after erase_filesystem() see if it pulses green indefintely, without a main.py. That means it's waiting in the repl

[rustic nymph]

No it was going to white

[tulip sleet]

ok, that's bad news if empty filesystem and going to white. that should never happen. i'd say just try a power supply to rule out computer, but it doesn't sound good

[rustic nymph]

Thanks for your help Dan, I gotta get to bed here

[tulip sleet]

me too!

kinda roundabout, but hw problems are uncommon

[rustic nymph]

I have an itsy bitsy I'll try the code on

[manic glacier]

dhalbert

[adafruit/circuitpython] New tag created: 3\.0\.2

[manic glacier]

ladyada

[adafruit/circuitpython] Pull request opened: #1190 Add trellis m4 variant and little fix for samr21

tested with trellis m4 rev B (purple)

[manic glacier]

arofarn

[adafruit/circuitpython] New comment on pull request #1189: Add VOLTAGE\_MONITOR pin to feather boards pins

OK. Good idea. It's done.

[marble hornet]

Anyone know how to copy and instance of a custom class in cp? I know it works for lists. I tried .copy() . _ _ _copy _ _ _() .deepcopy() and a bunch of googling. ANY help welcome. thanks either way.

?

[solar whale]

@tulip sleet much better. Breadboard compatible! Comes up at 3.3v - running open drain bilnky. Life is good 😉

[marble hornet]

@solar whale that's great it's working !

nrf right?

[solar whale]

Yes - nrf52840 "Dongle" - AKA pca10059

[marble hornet]

ah

[solar whale]

nice board for $10 https://www.nordicsemi.com/eng/Products/nRF52840-Dongle

[marble hornet]

wow that is really cheap ! (to untrained price ears)

[solar whale]

makes it easy to justify getting a few to play with 😉 Still that is the same price as a Raspberry Pi Zero W -- Still amazes me...

[marble hornet]

apple watches and wear os watches. Whole devices in packages smaller than a wall AC-DC brick !!!!!!! it's so amazing

from banging rocks to ...

[manic glacier]

dhalbert

[adafruit/circuitpython] New comment on pull request #1190: Add trellis m4 variant and little fix for samr21

Looks good! Only q is would you like to to add the trellis board to .travis.yml and tools/build_adafruit_bins.sh ?

dhalbert

[adafruit/circuitpython] New release published: 3\.0\.2

[tulip sleet]

https://blog.adafruit.com/2018/09/14/circuitpython-3-0-2-released-adafruit-circuitpython/

[ruby lake]

I've decided my hardware is flaky, the 3.3v logic signals can't quite play nice with the 5v mux chip

[stuck elbow]

@tulip sleet the pwm fix didn't get in?

[tulip sleet]

it did! supposed to!

[stuck elbow]

thanks

[tulip sleet]

https://github.com/adafruit/circuitpython/pull/1184

you scared me 😃

@timber mango i forgot to put it in the release notes. I'll add. I think I wrote it and it got deleted during editng.

@timber mango all updated

[marble hornet]

any thoughts on the object copying question ?

[tulip sleet]

@marble hornet regular Python has copy.copy and copy.deep_copy. MPy/CPy don't. But in general it's not necessarily obvious whether for any particular attribute in an object you want to copy it or just refer to it. So you may as well just write a .copy method for the class in question.

[marble hornet]

may i ask how one would do that,? just return the class calling itself with saved inputs for init?

[tulip sleet]

it depends a lot on whether the __init__() method sets all the attributes you want, or whether you want to copy some attributes that are computed or added later.

[marble hornet]

thanks!

[tulip sleet]

like here's a simple example:

class C:
    def __init__(self, x):
        self.x = x
        self.y = x+2

    def set_list(self, l):
        self.l = l

    def copy(self):
        cpy = C(self.x)
        cpy.l = self.l.copy()

so the init method sets x and y. Later you can set self.l. When copying, you make a copy of the list so it's not shared (so the two objects can change their own copies independently. But if .l were a tuple, then it's immutable, so you wouldn't need to copy it., etc.

[wraith tiger]

I created a simple virtual die on the micro:bit with MicroPython: https://youtu.be/RsWxPQvKYMw

[charred lake]

Adafruit Crickit still out of stock :(

Crickit for Circuit Playground Express

That's pretty cool @wraith tiger

[manic glacier]

ladyada

[adafruit/circuitpython] New comment on pull request #1190: Add trellis m4 variant and little fix for samr21

ok

[manic glacier]

dhalbert

[adafruit/circuitpython] Pull request closed: #1190 Add trellis m4 variant and little fix for samr21

dhalbert

[circuitpython:master] 6 new commits

28aabb4 trellis m4 variant - ladyada
bec4267 missin' reset pin for the radio. tested works. - ladyada
994c066 rev B (minus microphone) - ladyada
521a14e add mic in / out - ladyada
14ad7d3 automate builds - ladyada

[manic glacier]

blonguet

[adafruit/circuitpython] Fork created: blonguet/circuitpython

[onyx hinge]

Woo thanks for 3.0.2 @tulip sleet and all contributors!

[fierce girder]

@meager fog fyi - I have been validating the open hardware summit badges I recently received but will get back to the BeagleBone support in CircuitPython once I know the badges don't need rework 😃 I should be able to get a PR created for all GPIO pins by tomorrow

[meager fog]

totes!

ill do i2c & spi

then we can look at PWMout

[fierce girder]

thanks

[meager fog]

good luck!

[charred lake]

Man I love CircuitPython!

[raven canopy]

@charred lake yay! that's the idea. 😄

[charred lake]

@raven canopy haha! You guys are epic! 😄

[timber mango]

you folks did boocoo work on it. ;)

Simple USB in Atmel Start was plenty for me, and that's a dirt basic implementation (with an example project to work from).

Very clear to me that the ringbuffer in usb.c is in my future. ;)

[charred lake]

@timber mango I commend you sir!

[velvet oasis]

Hello, does anyone have a sort of anti-rebound routine for a switch?

while button.value == 1:

the press of a button generates so many hits that passes by my routine

[raven canopy]

@velvet oasis button debouncing is usually done with delays. the "blocking" method would be to use time.sleep(n) to halt all further execution so that no additional reads are taken. n can usually be around 0.01; adjust as you see fit. example at the top of this learn guide: https://learn.adafruit.com/circuitpython-essentials/circuitpython-digital-in-out

the non-blocking way involves delaying by comparing elapsing time. like so:

import time
interval = 1000
last_execution = time.monotonic()
current_time = last_execution
while True:
    current_time = time.monotonic()
    if (current_time - interval >= last_execution):
        # do that stuff we want to do
        last_execution = time.monotonic()

[fluid helm]

We are hacking with CircuitPlaygrounds and Edublocks at PyCon UK

[charred lake]

Niiiiiiiiiiiiiiiiice!

[manic glacier]

tannewt

[adafruit/circuitpython] Issue opened: #1191 Add subscr to Bitmap

So individual pixels can be read and written.

bitmap[0,0] = 1

[manic glacier]

hathach

[adafruit/circuitpython] New comment on pull request #1188: Nrf52 qspi flash

I tried to enable high performance bit but couldn't get it written any faster. It is probably default mode already. Note: nrf52 SCK clock speed is only up to 32 Mhz.

I post an issue on Nordic devzone to see if there is anyone got more lucky than us.
https://devzone.nordicsemi.com/f/nordic-q-a/38570/write-speed-with-mx25r6435f-on-nrf52840dk

[scarlet fjord]

hey so im having a bit of trouble with my neopixels
ive hand-soldered some on to a board of mine and im not sure if ive managed to kill the chips or if my code doesnt work. This is what I'm doing:

import board
import neopixel

pixpin = board.D5
numpix = 4

strip = neopixel.NeoPixel(pixpin, numpix, brightness=0.3, auto_write=False)

while True:
    strip.fill((255, 0, 0))

and I'm getting nothing on my board.
This is my board:

have i broken something, or am i just doing it wrong?

[tough flax]

Can anyone help me find the implementation of the input() method in CP? Actually any advice about how to trace a python function down to the implementation in C would be great.

[marble hornet]

Anyone else having trouble with jlc PCB and not dimensions or milling layers ?

[tough flax]

Found it - thanks anyway

[meager fog]

@scarlet fjord set auto_write to True

[scarlet fjord]

😮 ok

hmm no that didnt work
could i have damaged the chips?

[pastel panther]

@scarlet fjord It's always possible but unless there is some visible damage or you saw them burst into flames or something, I would say a soldering or wiring issue is more likely.

Do you have another known good neopixel device you can test your code with?

[scarlet fjord]

i do not 😅
i have some spare chips though so i might try soldering some wires to it and see whats going on tomorrow if you dont think ive managed to kill them 😛

[pastel panther]

I mean, where there is a will, there is a way to break them but I've only ever killed one by reworking it too many times and manhandling it in the process

[manic glacier]

notro

[adafruit/circuitpython] Issue opened: #1192 Moving a directory into itself doesn't raise an exception

Discovered while working on Lib/test/test_shutil.py

Adafruit CircuitPython patchbase-7-g216f0f952-dirty on 2018-09-15; Adafruit Metro M4 Express with samd51j19
>>> import os
>>> os.mkdir('/tmp')

>>> os.mkdir('/tmp/foo')
>>> os.listdir('/tmp')
['foo']

>>> os.rename('/tmp/foo', '/tmp/foo/bar')

>>> os.listdir('/tmp')
[]
>>> os.listdir('/tmp/foo')
Traceback (most recent call last):
  File "<stdin>", ...

notro

[adafruit/circuitpython] New comment on issue #1179: How are class function objects created?

I've now been through all the tests that test unittest itself (except mock).
unittest.expectedFailure is adding a property on the function and this is working using the function name as dictionary key.
The other use case is functions marked with unittest.skip. Skip itself raises an exception so that is working. But if all functions in a class is marked skipped, unittest also skips calling setUp, so this functionality is lost. This is no big deal, since the only real loss i...

notro

[adafruit/circuitpython] Issue closed: #1179 How are class function objects created?

[upbeat plover]

i cant flash my adafruit adalogger M0, it exits bootloader mode instantly, then flashes red

ill try throwing something on it with arduino

cant find its port

[meager fog]

hot off the press for people with a hallowing, you can generate & display a QR code!

https://github.com/adafruit/Adafruit_CircuitPython_miniQR/pull/1

[tulip sleet]

@upbeat plover does the red led pulse if you double-click?

[upbeat plover]

once then, turns off and the red flashes every sec

[tulip sleet]

weird, did you load Blink?

the pulsing is a "breathing" look

gotta eat lunch - back in a little while

[upbeat plover]

ive flashed before... it seems broken right now

it auto exits bootloader after 1 sec

it pulse on, then off, then on, then goes to a red flash

will i need to j-tag it or something to fix it?

[manic glacier]

ATMakersBill

[adafruit/circuitpython] Fork created: ATMakersBill/circuitpython

[manic glacier]

ATMakersBill

[adafruit/circuitpython] Pull request opened: #1193 added timeout to input\(\)

I have added a second optional parameter to input() that acts like a timeout.

When input("PROMPT>", 10) is called with a timeout parameter, it will return in that number of milliseconds if there is no data available on the USB serial connection.

This is not a true async serial read, it's just a workaround to let allow other tasks to occur while waiting for serial input to start. Once a keystroke is read, the method drops into readline processing allowing for command history e...

ATMakersBill

[adafruit/circuitpython] New comment on issue #231: 'getch\(\)'

I created a workaround for this and am sharing it via this pull request
https://github.com/adafruit/circuitpython/pull/1193

Not sure it's the right solution, but it does solve our issues.

[tough flax]

(FYI First PR here, please be gentle)

[manic glacier]

PaulKierstead

[adafruit/circuitpython] New comment on issue #1128: uart 4\.0

The addition of in_waiting has made non-blocking timed reads much easier; however, I think it would be very valuable to have the write_timeout separate from timeout, as is the case in pySerial. It is very common to want to wait on writes but not on reads, so separate timeouts are handy. Also useful is out_waiting, in this case so you can implement your own buffering.

I've been trying to do a bunch of asynio type stuff with the UART and those two (along with my 0 timeout fix) would be a ver...

[arctic heron]

Is there a planned date to merge 3.x forward to master?

[manic glacier]

ATMakersBill

[adafruit/circuitpython] New comment on pull request #1193: added timeout to input\(\)

I see from the Travis build that this fails on some ports that don't have usb_bytes_available().

If there is a desire to go forward, I can add checks for particular architectures or features that work. For our (short-term) needs the M0 will work for us (which doesn't imply anything for accepting the PR of course - we can build from our fork)

[manic glacier]

davidperrenoud

[adafruit/circuitpython] New comment on issue #899: Gaming API\!

You can take example from the Arduboy game format:
https://community.arduboy.com/t/arduboy-game-format-guide-arduboy-files/2784

It is a zip file ending in .arduboy which contains a precompiled binary and some metadata:

• game.hex
• info.json
• screenshot00.png
• banner.png

For CircuitPython, this could be as simple as shipping a .uf2 file with one or more games inside it. The advantage of this solution instead of shipping Python files, is that you can evolve the gaming API along ...

[lethal bear]

By the way, this is also how Smalltalk/Squeak works and the images still work more than 20 years later: https://squeak.js.org/run/#url=https://freudenbergs.de/bert/squeakjs&files=[Squeak1.13u.image,Squeak1.13u.changes,SqueakV1.sources]&swapButtons=true

@slender iron Is it possible to beta-test the DMA display rendering? Is the code available on GitHub?

My long term goal is to bring MicroBlocks (http://microblocks.fun) to game consoles. I tried with @stuck elbow's uGame but I got stuck because the rendering to the ST7735 was too slow

[tulip sleet]

@arctic heron I will merge 3.x to master on Tuesday or earlier.

[arctic heron]

👍

[tulip sleet]

@upbeat plover A very early batch of Feathers was shipped with unprotected bootloaders. Depending on its age, yours might be from that batch. You do need a j-link (https://www.adafruit.com/product/3571 is cheapest) or openocd https://learn.adafruit.com/programming-microcontrollers-using-openocd-on-raspberry-pi to re-program

[solar whale]

@tulip sleet @upbeat plover is on an adalogger M0. Were they also shipped with an unprotected bootloader like the M4? They used to be shipped with the Arduino bootloader and you could use bossac to load the UF2 bootloader. Sorry if I came into this late.

[tulip sleet]

@solar whale @upbeat plover this was an issue with early M0s from late 2015, when they were first introduced.

I had some of these that I set fuses on.

[solar whale]

Ah. Thanks for clarifying.

[tulip sleet]

np - it is a little worrying that the bootloader would be damaged, because it takes some setup to write the flash. There may be some kind of latent bug (same for the M4 issue)

[charred lake]

My girls are ready to turn this wooden house into a creepy holloween house with sound, motion detecting, and more. Using CircuitPython on Circuit Playground Express+Raspberry Pi (with my help and supervision of course).

Stay tuned!

[pastel panther]

@charred lake sounds awesome! I look forward to seeing it in action

[runic spindle]

hello! super new to this discord, and really excited to keep learning 😃

anyone here know how to calculate fourier transforms with circuit python on cpx? I see there's arduino code for it here but i'm wondering if there's a library for circuit python as well? https://learn.adafruit.com/circuit-playground-sound-controlled-robot/using-the-microphone

Appreciate any help, thank you so much!

[pastel panther]

@runic spindle I don't know, sorry

[runic spindle]

thank you anyway @pastel panther !

[pastel panther]

There is apparently an audio in call
https://circuitpython.readthedocs.io/en/latest/shared-bindings/audiobusio/PDMIn.html
and the normal CPX library, but it doesn't look like it has mic support yet:
https://learn.adafruit.com/circuitpython-made-easy-on-circuit-playground-express/first-things-first

I have an old version so I could be wrong about the mic support of the cpx lib

[runic spindle]

hmm so i'll have to attach a separate mic i guess

good thing i have one lying around 😄

worst case its a small project so i can just transition everything to arduino

[pastel panther]

Sorry if I wasn't clear; the first link was a way of accessing the mic

[runic spindle]

ohh interesting

[pastel panther]

I'm just not sure if it's integrated into the CPX helper library yet. You should be able to use PDMin directly

The arduino/C library for the circuit python has the FFT support built in but I don't know if it's available in the Circuit Python version yet

[runic spindle]

ah i see so if i really wanted to i could do the FFT calculation myself in CircuitPython using PDMin

?

might just transition to Arduino ahahaha. this is all super helpful though thank you for your help!!

[pastel panther]

no problem, glad to help

[runic spindle]

i made a bongo cat robot plush and i just want to make it bongo when it hears a low pitch sound ahahaha https://twitter.com/chardane/status/1041219760367665153

charlyn gonda (@chardane)

I love #BongoCat so much I made it a robot plush ahahaha!! 😂😍

First time playing around with @adafruit crickit! 🤖 Thank you @StrayRogue for bringing Bongo Cat into the world 👏🏼🙏🏼💖🎉

#12monthsofmakes September ✅ https://t.co/qWx8ui9uZo

Likes

470

▶ Play video

Twitter

magnitude works using RMS but i thought that detecting low tones with FFT would be much cooler hahaha

[raven canopy]

@runic spindle FFT has been discussed here before. IIRC, it would need to reside on the C side; the Python side is not fast enough to be all that useable. MicroPython has a FFT, but requires an FPU-capable MCU (M0 doesn't have FPU). Arduino is probably your best bet...

[runic spindle]

very cool, thank you for that info!

[raven canopy]

you can search this channel for the history. may be some relevant links/discussion...

[manic glacier]

tannewt

[adafruit/circuitpython] New comment on pull request #1193: added timeout to input\(\)

I don't want to add this as is because it breaks compatibility with CPython because input only takes in a prompt. https://docs.python.org/3/library/functions.html#input

You could add this as a custom function in a custom module. It gets it closer to the second usb serial api that would be good in the long term. Let me know if you want help with that.

[manic glacier]

ATMakersBill

[adafruit/circuitpython] New comment on pull request #1193: added timeout to input\(\)

I get it - I have no problem with that

[manic glacier]

ladyada

[adafruit/circuitpython] New comment on issue #1128: uart 4\.0

i think we're slowly going to try and add all pyserial supported functions, but ya some are more useful than others :)

[meager fog]

@runic spindle soooo cuuute

[manic glacier]

tannewt

[adafruit/circuitpython] New comment on issue #899: Gaming API\!

@davidperrenoud Thanks for the tips! I like the idea of packaging in a zip.

I don't really want to use uf2 to ship games though because its more complicated than editing files on a CIRCUITPY drive. Ideally we'd have the fast iteration loop with game files just like code.py.

[slender iron]

@lethal bear all my work is going into the master branch. Its not very optimized yet though

[slender iron]

<@&356864093652516868> Join us tomorrow at 11am Pacific/2 pm in the voice channel for the one year anniversary of the CircuitPython weekly meeting! 🎉 🎉 For a flash back check out the first recording here: https://youtu.be/OX825aegraw

[pastel panther]

parteeeeeeee 🎆

[lethal bear]

@slender iron Are you using parallel, MIPI or SPI interface? I only see mention of SPI in displayio

[manic glacier]

dhalbert

[adafruit/circuitpython] New comment on issue #1174: nrf: Incorrect pin voltages

Fixed by #1177.

dhalbert

[adafruit/circuitpython] Issue closed: #1174 nrf: Incorrect pin voltages

dhalbert

[adafruit/circuitpython] New comment on pull request #1177: use open\-drain capabilities on GPIO; clean up board init; set correct GPIO voltage

Fixes #1174.

dhalbert

[adafruit/circuitpython] New comment on pull request #1177: use open\-drain capabilities on GPIO; clean up board init; set correct GPIO voltage

Fixes #1151.

dhalbert

[adafruit/circuitpython] New comment on issue #1151: Turn on nRF52 instruction cache if it is not already on

Fixed by #1177.

dhalbert

[adafruit/circuitpython] Issue closed: #1151 Turn on nRF52 instruction cache if it is not already on

dhalbert

[adafruit/circuitpython] New comment on issue #1089: nrf: implement checking for pin in use

Fixed by #1040.

dhalbert

[adafruit/circuitpython] Issue closed: #1089 nrf: implement checking for pin in use

dhalbert

[adafruit/circuitpython] New comment on pull request #1158: nrf: rework of pin files; add new boards; add pin claiming

Fixes #1089

dhalbert

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

Fixed by #1158.

dhalbert

[adafruit/circuitpython] Issue closed: #996 Add board files for PCA10059 \- the nRF52840 dongle

dhalbert

[adafruit/circuitpython] New comment on issue #1008: USB mass storage

Fixed by #1074.

dhalbert

[adafruit/circuitpython] Issue closed: #1008 USB mass storage

dhalbert

[adafruit/circuitpython] New comment on pull request #1074: Nrf52840 usb hid

Fixes #1008 as well (and maybe also via other PRs).

dhalbert

[adafruit/circuitpython] New comment on issue #1009: USB HID

Fixed by #1074.

dhalbert

[adafruit/circuitpython] Issue closed: #1009 USB HID

dhalbert

[adafruit/circuitpython] New comment on pull request #1074: Nrf52840 usb hid

Fixes #1009.

dhalbert

[adafruit/circuitpython] Issue opened: #1194 nrf: implement uheap and ustack

These are not incredibly vital but would be nice for consistency with atmel port.
Breaks out task from #526.

dhalbert

[adafruit/circuitpython] New comment on issue #526: nRF Port: Get Remaining Shared Modules Working

Broke out remaining tasks into individual issues and closed in favor of them. (e.g. #1194).

dhalbert

[adafruit/circuitpython] Issue closed: #526 nRF Port: Get Remaining Shared Modules Working

[tulip sleet]

git stuff above is just issue bookkeeping getting the nrf52 task tracking up to date

@indigo wedge of the remaining nrf52 tasks https://github.com/adafruit/circuitpython/labels/nrf52 are there some that you particularly want to do yourself? @gentle bronze and I would plan to do what's remaining by default. @gentle bronze is also working on nrf52 Arduino support.

[gentle bronze]

@tulip sleet there is still lots of nrf52 issues. Please assign me to issue that you feel must have for the feahther launch to implement. Since Arduino nrf52840 basic aren't runnable yet, I would like to make it at least usable for the user. Since many would expect to use it with Arduino as well.

[indigo wedge]

I've been really busy recently so can't contribute much right now, I have a mostly complete ble code rewrite waiting for a rebase and push but not much more than that.

[gentle bronze]

@indigo wedge yeah, ble is the most fun part of nrf52 port, that would be great

[tulip sleet]

@indigo wedge that's fine - you are a volunteer and your contributions have been very helpful! The BLE stuff, especially (from my point of view) the API we want to present, is certainly where your expertise is very valuable.

[indigo wedge]

i have it 95% done, just need time for that 5%

[tulip sleet]

we have lots more to do -- it is not a blocker 😃

[indigo wedge]

I'm pondering making a CP nRF52840 badge for this year's CCC so might spend more time on it soon 😉

[stuck elbow]

@indigo wedge I saw the new rocket badge yesterday

[indigo wedge]

oh? do you have a link?

[stuck elbow]

says blinkenrocket.com on the box

I got a kit, the new one is 2x smaller, with the 20mm matrix

[indigo wedge]

don't think i ever heard of this

[stuck elbow]

you program it from the browser from your phone or computer with an audio jack

[manic glacier]

davidperrenoud

[adafruit/circuitpython] New comment on issue #899: Gaming API\!

This is also how GP Blocks and MicroBlocks works:

• .gpp file (a zip file)
  • objects.gpod
  • thumbnail.png
  • ancestor
  • hash
  • topHash
  • modules/
    • 6b3f57d31cf79cb743a589a0f3862f09493fe246ab02a8b9bf44132136f14f38

The code, sounds and images are all saved in the .gpod file which I assume stands for GP Object Database. You could also save the code as separate .gp files, but then you lose some of the advantages of...

[stuck elbow]

took them 2 years to perfect that technique

[indigo wedge]

nice, I've been looking into that as a way to sync with the phone

[stuck elbow]

@indigo wedge if you decide to do that badge, I would be very interested to get involved

[indigo wedge]

I've been tweeting about it, not sure if you saw it, current idea is to have the badge be also a optional smartwatch with eink

[ruby lake]

audio firmware update has been used in synth gear for a few years

Olivier Gillet (Mutable Instruments) does all his code open-source, for example https://github.com/pichenettes/avr-audio-bootloader

[stuck elbow]

@indigo wedge I toyed with that idea a bit for the pycon.uk badge, I even did some research with the cheap e-inks, I figured out how to drive them fast

https://hackaday.io/project/100853-circuitpython-badge/log/139310-the-look-up-table-holds-the-secret

I also did some brainstorming with the watch design, if you are interested:
https://hackaday.io/project/100853-circuitpython-badge/log/142479-the-shape

[indigo wedge]

Very interesting, I will look at this later after work, thanks

For the shape I already have the one I used and verified last year, it's quite comfortable

[fluid helm]

PyCon UK badge? Did i miss this for this year? @stuck elbow

[stuck elbow]

@fluid helm there should have been a lightning talk for it

I couldn't come, so I'm not sure if it actually happened

[fluid helm]

Hmm, might not have made the random draw

will the badges be for next year?

[stuck elbow]

that's the plan, and we wanted to recruit some people this year to help with it

but it's all very fuzzy still

by the way, I have some ideas about how to cut the cost in half if we need that

[covert oxide]

qemu

Hi all, I'm trying to understand what tests are being executed after each Circuitpython promotion. Rosie executes the tests on board of tests/basics and tests/circuitpython, and travis executes everything else under test/ for Unix port, is this correct? and qemu tests are gone?

[manic glacier]

tannewt

[adafruit/circuitpython] New comment on issue #1194: nrf: implement uheap and ustack

I think uheap can be removed instead. I don't know anyone who uses it.

[slender iron]

@covert oxide I think we turned them off

@lethal bear fourwire spi for now but I'm trying to design it so its easy to swap in a different bus protocol

[covert oxide]

thank you @slender iron , as for the other tests, is my understanding correct? Rosie executes the tests on board of tests/basics and tests/circuitpython for all ports, and Travis executes everything else under test/ for Unix port? sorry, I'm not very experienced understanding how this works.

[slender iron]

ya, I think so

[stuck elbow]

I think the maker faire was a success, I always had at least 3 kids standing by my table and playing the games

[arctic heron]

@indigo wedge The last 5% takes 80% of the time...

[stuck elbow]

wasn't it 20%?

[tidal kiln]

i think it's generally asymptotic

[raven canopy]

@inland tusk yep!

@idle owl 👋 👋 welcome back!

[idle owl]

Thanks!!

[raven canopy]

Hehe. Sure..

<@&356864093652516868> Meeting starting shortly. Anyone else, feel free to join in as well!

[upbeat plover]

@tulip sleet after i get (https://www.adafruit.com/product/3571) debugger is there a very simple tut i can follow, one that doenst require building anything in python?

[slender iron]

https://docs.google.com/document/d/1tLdD8-A_IMqdaapoTXcGDTQTYdizdHZ1MeiKVY39zUw/edit?usp=sharing

[upbeat plover]

for reprogramming my adalogger m0?

[fluid helm]

Won't be able to make it this evening as flying to San Francisco tomorrow, have a good meeting all!

[idle owl]

@fluid helm Have a good one!

[pastel panther]

hiyeeeeeeee

lurkmode activated

[idle owl]

Hooray!

[tidal kiln]

🍰 1 YEAR!!!

[pastel panther]

Yaaaaaaaaaay

💃

🕺

[neat folio]

just lurking, (and restarting PC and it decided to do updates 😦 )

yay origin stories 😃

[errant grail]

yay @idle owl!