circuitpython-dev

[pastel panther]

it's trying to be too smart and can't tell that my trellis is plugged in

[idle owl]

Doesn't work with Trellis, afaik.

[pastel panther]

🙄

back in the trash can it goes

[ruby lake]

I had that issue 😉

[sacred edge]

So I have been playing with the basic functions on the Circuit Python Playground Express, AKA "CPPE", I have been using the "Make Code" web app... Which while it does work, seems (To Me) a bit limited on what one can do... Although, It is amazing how many loops you can squeeze into the screen...

The question is, How do I migrate over to using Mu instead of MakeCode?? Is there a Circuit Python Command Reference manual?

[exotic pumice]

I think this is what you're looking for https://circuitpython.readthedocs.io/en/3.x/

[sacred edge]

Thanks, I'll search it out!!

[exotic pumice]

probably an intro to python too

not sure what the best source for that is

oh I guess you could try Mike Barela's CPX book

https://www.makershed.com/products/getting-started-with-adafruit-circuit-playground-expres

[gusty topaz]

Guys, can I use the Adafruit Mini TFT with Joystick Featherwing with the M0 and CircuitPython?. It doesn't make reference to CircuitPython on the product page: https://learn.adafruit.com/adafruit-mini-tft-featherwing

[manic glacier]

dhalbert

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

Looks good to me, but mpconfigboard.mk is set to SPI, not QSPI filesystem? Does QSPI work now?

tannewt

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

@dhalbert Nope, I'm going to look at it when I get my prototype.

dhalbert

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

dhalbert

[circuitpython:master] 2 new commits

15eeac5 A few fixes for nRF52840 feather QSPI and neopixel - tannewt
292737f Merge pull request #1354 from tannewt/nrf_feath... - dhalbert

dhalbert

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

Merged anyway - I read Slack convo.

[cosmic patrol]

Hey guys, I'm making a custom PCB with a SAMD21 board and I want to have it appear as an M0 express. I'm going to modify the M0 express PCB. I can't figure out how I should upload the bootloader to the chip though. I understand that I can upload the bootloader using a J-link and ATMEL studio, but all the resources mention uploading .bin or .hex bootloaders. Will the same process work for .uf2 bootloader provided for the M0 express?

I tried to read through the makecode blog on .uf2 but I'm not sure I quite get it. I understand that the chips ship with SAMBA bootloader already flashed. Will this allow the device to appear as a mass storage device directly? I imagine I'll have to hold down the reset button instead of double clicking.. Can I just drag and drop the .uf2 file? As you can probably tell, I'm very confused.

My backup option is to buy M0 express boards and get the pre-programmed microcontroller from it. I'll probably do this for the first few boards anyway.

Here are my references: https://www.instructables.com/id/ATSAMD21g18-Bootloader-With-J-link-Debugger/, https://learn.adafruit.com/proper-step-debugging-atsamd21-arduino-zero-m0/restoring-bootloader, https://makecode.com/blog/one-chip-to-flash-them-all

I apologize if this is an already answered question... I couldn't find it though.

Thanks in advance!

[exotic pumice]

you can convert uf2 to bin

[cosmic patrol]

Oh, I knew you could convert bin to uf2, didn't know the opposite was possible as well..

[exotic pumice]

https://github.com/Microsoft/uf2/blob/master/utils/uf2conv.py

I think you just pass it the uf2 and it figures it out based on the filename

[slender iron]

@cosmic patrol the chips do not have a bootloader when shipped afaik

you should expose the SWD pins somehow so you can load through the debug port

[cosmic patrol]

@slender iron yes! I'm planning to expose them via test pads as in the adafruit boards

I'll make a pogo pin jig to program the boards

[slender iron]

if you have the space I like the 2x5 header

[cosmic patrol]

It's going to be an earring, so space is at a premium!

[slender iron]

ah, ya. test pads then 😃

[cosmic patrol]

@exotic pumice Thanks a bunch! I'll give this a shot and share the results

[exotic pumice]

@cosmic patrol happy to help

[cosmic patrol]

I have another question... can I use bossac to upload the bootloader to a fresh chip or does it have to be through an SWD programmer?

[exotic pumice]

I don't think you can because bossac uses part of the bootloader

so it would be overwriting itself basically

[cosmic patrol]

Ah, I figured it would be too good to be true 😛

No worries, J-link will do just fine!

[manic glacier]

tannewt

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

We merged some USB fixes today so make sure and update your repo. Thanks!

[exotic pumice]

do you have a j-link yet?

[cosmic patrol]

Not exactly... I have a board with a J-link dev board that has a microcontroller as well. I plan to cut the swd traces and connect them to the samd21

[exotic pumice]

ok. My solution was a raspberry pi and an swd breakout board

https://www.adafruit.com/product/2743

[cosmic patrol]

Whoa, this is just a passive board, right? No active chips? I can use a raspberry pi as a SWD programmer?!

[exotic pumice]

yep

there's a learn guide on it

it's a little old and I had to adapt the config file but I got it working https://learn.adafruit.com/programming-microcontrollers-using-openocd-on-raspberry-pi/overview

[cosmic patrol]

Whoa! You're blowing my mind!!

[exotic pumice]

it's kinda slow but for the file sizes I'm working with that's alright

[cosmic patrol]

I HAVE to try it now.

[exotic pumice]

lol

I can upload my config file but it's set up for stm32

I think it's just one line to change the board

[cosmic patrol]

I'm just dusting off my Pi. I'll try and follow the instructions first

[exotic pumice]

ok

[cosmic patrol]

Hmm, partial success. I used an external J-link(I will definitely try the raspberry pi approach soon as well) connected to a windows computer. I converted the latest 3.0.3 .uf2 file to .bin using the microsoft uf2conv.py util. I've attached the resulting bin file. Then I followed the instructions posted here: https://www.instructables.com/id/ATSAMD21g18-Bootloader-With-J-link-Debugger/ to program an M0 express board with this bin file. The upload happens without issues. However, when I connect the board to my computer now, I can't detect it at all. The neopixel on board is dead as well. Help please!!

The connections were as such:
M0 Jlink
RST RST
SWDIO SWDIO
SWDCLK SWDCLK

GND GND

M0 was provided with power through a USB cable, as I wasn't sure if the J-link could supply enough current

TLDR; on the instructions: In Atmel Studio, Tools>Device Programming -> Select Chip -> Verify fuses -> Erase -> Program with .bin file

I tried converting to and uploading .hex format as well, but Atmel studio cannot read the .hex file. But I get the following error: "Unable to parse objectfile C:\PROJECTS\hearrings\uf2-master\uf2-master\utils\adafruit-circuitpython-feather_m0_express.hex: Unsupported format."

[exotic pumice]

try this maybe? https://learn.adafruit.com/proper-step-debugging-atsamd21-arduino-zero-m0/restoring-bootloader

oh , you used a circuitpython uf2 not a bootloader

that'd do it

@cosmic patrol

[cosmic patrol]

Oh, um.. I'm not sure what the difference is 😛

[exotic pumice]

circuitpython is a userspace program not a bootloader

[cosmic patrol]

I see. Yes, that makes sense.

So, which bootloader should I use?

[exotic pumice]

you install the bootloader first, then the circuitpython if you want it. There's a bootloader linked in that learn guide, not sure if it's the latest, let me do some digging

here we go https://github.com/adafruit/uf2-samdx1/releases

[cosmic patrol]

Got it! Trying it now.

Woot woot!! SUCCESS!

@exotic pumice I cannot thank you enough! You've been my 👼 tonight!

[exotic pumice]

awesome!!!

you're welcome

[cosmic patrol]

I'm making earring that whisper prerecorded messages in your ear, by the way. I call hearrings 😛

[exotic pumice]

cool

[timber mango]

@exotic pumice is the same true for the SAMD51 (used on the ItsyBitsy M4 Express) - i planing / thinking of using this one for a own board (smaller then the ItsyBitsy M4 Express - less pins accessible) - i need a small half-round outline for one of my props... (if i could fit in the ItsyBitsy i would not think of doing my own - as its only more work and costly overall... 😉 )

[exotic pumice]

@timber mango is what true for samd51?

[timber mango]

@exotic pumice that you need a SWD programmer to flash the UF2 bootloader
iam used to use a arduino as ISP to programm the other emty /new ATmega chips myself ;-)
so would be great if this 'replicate itself' approach would also work for the SAMD51 processors...

[lost moss]

^^^ Agreed.

[exotic pumice]

I think it's the case that you need some form of SWD programmer, as I mentioned to tinkrmind, that could be something you already have on hand like a RPi. Perhaps it is also possible to bitbang swd on a samd chip, but I don't think anyone has tried.

[lost moss]

Mmm LOL Getting a Pi Zero as SWD Programmer would be the best Idea IG

[timber mango]

seems there are more boards you can use with OpenOCD
https://petervanhoyweghen.wordpress.com/2015/10/11/burning-zero-bootloader-with-beaglebone-as-swd-programmer/

[exotic pumice]

yep

it would be interesting to try and make a samd-based programmer

I'm trying to locate a detailed specification of SWD to see if that would be feasible

it should be possible with pretty much any linux soc, all that you need is a openocd target

I think the ST-Link is based on a pretty weak chip like an M3 or something

yeah so it should be possible with an M4

[stuck elbow]

take a look at the black magic probe

[exotic pumice]

I did

maybe I should have a chat with @faint galleon

[main meteor]

Now I'm tempted to see if I can implement SWD with an ItsyBitsy. At first glance, it looks feasible.

[exotic pumice]

yeah

[stuck elbow]

with trinket m0

should be possible, afaik the BMP works with stm32f103, which is a weaker microcontroller

[timber mango]

woo i know that this is something i think its way over my comfort zone... 😉

if anybody a tempts something like this iam happy to try it out 😉

[exotic pumice]

stm32f103 is an m3 though

@stuck elbow

[stuck elbow]

I thought it's an m0

[exotic pumice]

https://www.st.com/resource/en/datasheet/cd00161566.pdf

[manic glacier]

DakotahRedstone

[adafruit/circuitpython] New comment on issue #1330: call for the creation of the first "LEARN SESSION"

Mondays are pretty busy, with pretty much the only three things in my life that happen according to a schedule... circuitpython meeting, a VISTE meeting bit later after that, and then a weekly group gaming session. So other than Monday afternoon/evenings, I can be available pretty much anytime I can manage to be awake... ( a non-trivial effort, but doable - sleep apnea is no fun) ;)

Whatever is convenient for you, will prolly work fine for me. Suggest a time and I'll set some alarms/w...

DakotahRedstone

[adafruit/circuitpython] Issue closed: #1330 call for the creation of the first "LEARN SESSION"

DakotahRedstone

[adafruit/circuitpython] Issue reopened: #1330 call for the creation of the first "LEARN SESSION"

DakotahRedstone

[adafruit/circuitpython] New comment on issue #1330: call for the creation of the first "LEARN SESSION"

oops wrong button

[exotic pumice]

I know a guy at ARM so I'm trying to get as much SWD info as possible from him.

I'll report back if it's not NDA'ed

[gusty topaz]

Guys, any guidance on using the Adafruit Featherwing with CircuitPython? https://learn.adafruit.com/adafruit-mini-tft-featherwing

[solar whale]

@gusty topaz best to wait for displayio to become available. Current support is minimal.

Or check out https://forums.adafruit.com/viewtopic.php?f=60&t=142003 for a work in progress.

[manic glacier]

caternuson

[adafruit/circuitpython] Issue opened: #1355 create driver for TMP006

For this:
https://www.adafruit.com/product/1296

A driver for the TMP007, which replaces the TMP006, already exists. However, might help to go ahead and make one for the TMP006 also, for people still using it, ex:
https://forums.adafruit.com/viewtopic.php?f=50&t=143712

[gusty topaz]

@solar whale excellent, thanks!

[pastel panther]

@main meteor I would love to see an SWD implementation on an itsy m4. Enough that I might be inclined to throw some money at the person who does it (a token amount, not a fortune)

[main meteor]

Hmm, I have the Itsy and a Metro M4 with an SWD connector. All I'd need would be the SWD adapter to start playing with it. I may take a crack at that.

[pastel panther]

Ironically I think you would probably want a working SWD debugger for this project if you don't already have one

[main meteor]

True. I'm likely to brick my Metro while debugging it, and it would be nice to have a way to bring it back.

[exotic pumice]

it would also be interesting to see if you could reconfigure the swd port on the metro m4 to an output rather than an input

it's bidirectional but I mean a programmer rather than a programming port

[exotic pumice]

that's what they do on the blue pill

then you wouldn't even need a SWD breakout

for metro to metro anyway

[main meteor]

I'll admit, that thought occurred to me too (the connector is right there, and cables are available)

[exotic pumice]

I got cables and the breakout coming in the mail for an unrelated project

no itsy tho

[main meteor]

Heh, I have an Itsy but no cables. I made an order yesterday (I think a bunch of people did), but didn't happen to include the cable.

[stuck elbow]

I always keep a stock of male and female dupont cables

[pastel panther]

@stuck elbow the pitch on the standard swd cable is different from the normal ones

[stuck elbow]

ah, I have an adapter for that

https://1bitsquared.com/collections/accessories/products/jtag-swd-100mil-pitch-breakout

but yeah, it requires the ribbon cable as well

[pastel panther]

that reminds me that I need to restock on those shroudless swd headers

they're nice because the footprint is smaller and the way the cable sits when plugged in it will clear many common SMD components

[main meteor]

For years, I'd just have a big 0.1" through-hole connector for JTAG and the like, but as boards get smaller, I'm starting to look at those smaller connectors, and (like you) the non-shrouded varieties. I'm also eyeing things like TagConnect.

I can't help wondering if the ST-Link is cross-compatible.

[exotic pumice]

not for samd apparently https://electronics.stackexchange.com/questions/250664/can-i-use-st-link-programmer-for-non-st-chips

but you can reprogram an st-link with bmp firmware and maybe then it would work

[sacred edge]

Mu has stopped working...

More specifically, Mu no longer goes to the "Save" menu anymore... It just sits there saying it is busy.. ( Busy Icon)..

[stuck elbow]

what did you do just before that happened?

[slender iron]

@main meteor @pastel panther @stuck elbow ladyada is looking into header with a keying shroud which are the best of both worlds

@stuck elbow I dig the all one one side breadboard style so the board can be vertical

I made to simple boards like that a few days ago. one for passives and one for sot23

https://oshpark.com/shared_projects/y61Ammy3

[stuck elbow]

for two-row connectors like that I usually put the pcb in between the pins, and solder them flat on it

[slender iron]

https://oshpark.com/shared_projects/PuI3Z5LH

[pastel panther]

@slender iron I've seen one that might be talking about (on the black magic probe). If I could find them at a comparable price I might get some but they were more expensive last I checked.

[slender iron]

ya I think esden had them a while

the nordic 52840 DK has one as well

this is the style of connector: https://1bitsquared.com/collections/accessories/products/jtag-swd-smd-connector

[pastel panther]

ya, that's it

also: FIVE DOLLARS?!

[slender iron]

😃

[pastel panther]

In all fairness that's probably not much more than his cost.

[manic glacier]

TG-Techie

[adafruit/circuitpython] New star added

[sacred edge]

OK, when I start Mu, it opens and I get Mu 1.0.1 - Untitled (Not Responding)

[manic glacier]

TG-Techie

[adafruit/circuitpython] New comment on issue #1346: Scratch Ram or Something

I'd love to give it a shot, however, should this go in storage? if so could someone please remind me where it is?

yeyeto2788

[adafruit/circuitpython] New comment on issue #1094: String internationalization for Spanish\.

@c47d is there anything I can help with the translation into Spanish? As far as I saw in the file there are some string pending to translate.

Let me know and we can work together on it.

Regards.

[sacred edge]

Mu takes a very long time (2 Minutes) to open the "Save File" dialogue... Any ideas?

[teal thorn]

This is a nice cost effective SWD debugger: https://www.maximintegrated.com/en/products/microcontrollers/MAX32625PICO.html/tb_tab3

It supports 1.8V to 3.3V I/O and it comes with the 10pin fine pitch ribbon cable.

The drag-n-drop feature is target specific, but the the CMSIS-DAP interface is universal.

[exotic pumice]

you can get knockoff st-links for like $2-5 on aliexpress. I still think it would be good to have samd swd though so those that have 2 adafruit boards need no other hardware to program

[teal thorn]

The MAX32625PICO board also embeds UART signals in the same 10pin header.

[pastel panther]

@teal thorn looks nice. Does it support GDB?

[exotic pumice]

except maybe the breakout

[teal thorn]

It will support GDB with openocd, or any CMSIS-DAP compliant debugger

[pastel panther]

cool

(I know precious little about CMSIS-DAP so perhaps that should have been obvious)

[teal thorn]

I like to use it with this cable if you want to skip the connector on the target: http://www.tag-connect.com/TC2050-IDC-NL-050-ALL

[pastel panther]

I've been using my JLink edu-mini

I've seen that connector but I'm not super stoked about it. It doesn't save much space and you have to have their cable

should someone produce stls for an OS version, I might be more interested

[teal thorn]

I just used the MAX32625PICO to load the UF2 bootloader into the blank SAMD21 parts on my FPGAhelper boards using openocd, so I know it work with the SAMD devices.

[exotic pumice]

yeah, that's not what I was saying

[pastel panther]

neat; it does seem like a pretty cool little chip

[exotic pumice]

I want a samd-based programmer

[teal thorn]

You could program the tinket-m0 to do that.

[exotic pumice]

we were thinking itsy m4

you sure m0 is fast enough?

[teal thorn]

ARM implements CMSIS-DAP in the M0 all the time (LPC11U35)

[exotic pumice]

cool, good to know

how is CMSIS-DAP related to SWD?

[teal thorn]

It is the standard USB to SWD convert protocol developed by ARM to eliminate all the different debugger drivers

[exotic pumice]

oh cool

[teal thorn]

You can get the source code here if you want to port it to SAMD: https://github.com/ARMmbed/DAPLink

[exotic pumice]

tyvm

[teal thorn]

It also runs on the st-link

[exotic pumice]

when they refer to "board" and "target" are they talking about the programmer or the device being programmed?

[pastel panther]

I believe the target is the thing being programmed

[exotic pumice]

so it seems I would have to add support both for samd targets and samd programmers

is the HIC the programmer?

[teal thorn]

If you are using pyocd or openocd, they need to be aware of the target. You do not need to modify CMSIS-DAP to add new targets.

This is a difference between CMSIS-DAP and the black magic probe. Because the black magic probe runs the GDB server, it has to understand the internals of the target. CMSIS-DAP does not becuase the target specific portion is handled by the openocd or pyocd.

DAPLink also offers drag-n-drop programming. If you want to use the drag-n-drop programming, you need to add target specific code to do that, but you do not need to add target specific code to use the debugger.

DAPLink implements a composite USB device with three endpoints (MSD, HID and CDC)

MSD (Mass Storage Device) is the drag-n-drop programmer that looks like a USB thumb drive. This is the only interface that needs to be customized for the target.

HID is the generic CMSIS-DAP SWD debug adapter interface.

CDC is the virtual UART for serial communication

[exotic pumice]

when I say HIC I am referring to this folder, it seems this is where the programmer code lives? https://github.com/ARMmbed/DAPLink/tree/master/source/hic_hal

[teal thorn]

That does not look like the target specific programmer code. The target specific code is in the source/target folder.

[exotic pumice]

not target, but programmer

[teal thorn]

Part of the daplink project also includes a drag-n-drop bootloader for firware updates.

[exotic pumice]

the device doing the swd communication

to the target

I think that's the part I need to dive into

which I'll try to do at a later date

[teal thorn]

The low level SWD code is in this directory: https://github.com/ARMmbed/DAPLink/tree/master/source/daplink/cmsis-dap
It is fairly portable if you want to use it separately.

How difficult would it be to use the SD card slot on the Feather M0 Adalogger as the circuitpython drive?

[exotic pumice]

fairly difficult I think, but you can have circuitpython load things from the sd card

including code and data I think

[solar whale]

@teal thorn if you just want to execute code from it, that can be done but it can’t be accessed as a USB drive like CIRCUITPY.

https://learn.adafruit.com/micropython-hardware-sd-cards/code-storage-on-sd-card

[pastel panther]

@teal thorn It's probably possible to make it show up as the CIRCUITPY drive, but it would be a substantial change to the core code as you'd have to replace the SPI/QSPI flash driver that the MSC code uses with a driver for the SD card. Probably other stuff too that I'm not aware of. @slender iron or @tulip sleet could point you in the right direction

would be kinda nifty though

[manic glacier]

C47D

[adafruit/circuitpython] New comment on issue #1094: String internationalization for Spanish\.

Hi @yeyeto2788, i had been busy with work and as you already noticed there are some not translated or not updated strings, so feel free to send a pull request and me or @sabas1080 can review it.

Gracias por la ayuda!

[manic glacier]

jerryneedell

[adafruit/circuitpython] New comment on issue #1341: SPI not "deinited" on soft reboot for nrf boards

appears to be fixed on pca10059

>> 
soft reboot

Auto-reload is on. Simply save files over USB to run them or enter REPL to disable.



Press any key to enter the REPL. Use CTRL-D to reload.
Adafruit CircuitPython 4.0.0-alpha.3-13-g292737fa2 on 2018-11-24; PCA10059 nRF52840 Dongle with nRF52840
>>> 
>>> import sdmount_lib
>>> import hello
Hello World!
>>> 
>>> 
>>> 
>>> 
soft reboot

Auto-reload is on. Simply save files over USB to run them or enter REPL to disable.
...

jerryneedell

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

REPL appeared immediately on PCA10059 with current master

Press any key to enter the REPL. Use CTRL-D to reload.
Adafruit CircuitPython 4.0.0-alpha.3-13-g292737fa2 on 2018-11-24; PCA10059 nRF52840 Dongle with nRF52840
>>> 

[teal thorn]

Is there a place for feature requests? Enabling an SD card for the circuitpy drive would be very useful. All the pieces should be there. I might even be able to figure it out, but someone else could probably do it quicker. I ordered the feather m0 adalogger with their black feather sale, but it will probably be a while before I have time to try a project like that.

[pastel panther]

You can create an issue on the main CP repo to request a feature

[manic glacier]

klardotsh

[adafruit/circuitpython] New comment on pull request #1274: Implement CPython\-compatible gzip\.decompress, restore uzlib\. Enable both for atmel\-samd

Sorry for the over a month of radio silence on this - burned out for a bit there on OSS work (and the project this PR was ultimately to be in support of). I'm back and plan to fix this branch up some time in the next week or two - looks like the only conflicts to merge in are localization-related, thankfully.

[manic glacier]

tannewt

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

Ok, thanks @jerryneedell! Closing this.

tannewt

[adafruit/circuitpython] Issue closed: #1340 nRF boards slow to enter REPL

[slender iron]

@gentle bronze did you get the metro m0 example going? Its not working for me with the makefile

[gentle bronze]

@slender iron work as well as m4. I used the segger studio. It could be the clock set up, I got into that issue previously as well.

[slender iron]

ya, I think it is. the 1 second blink is more like 1.3

I

I'm surprised though because that should be shared code

will look again tomorrow

[gentle bronze]

Samd21 does not implement the systemcoreclock

[slender iron]

ya, I fixed that and its still wrong

[gentle bronze]

Properly, I got it blink super fast as well

Can you try with the segger studio, just to make sure it work there. Just install ide then samd21 package, prettty simple

[slender iron]

ya, maybe tomorrow

I know it works with circuitpython but it uses different clock init code

my work in progress branch is here: https://github.com/tannewt/tinyusb/tree/makefile

[manic glacier]

ladyada

[adafruit/circuitpython] New comment on pull request #1274: Implement CPython\-compatible gzip\.decompress, restore uzlib\. Enable both for atmel\-samd

@klardotsh no rush :) whenever you're ready - take breaks whenever ya need!

[gentle bronze]

@slender iron yeah, you should sleep now.

[slender iron]

yup, winding down. thanks! I hear usb is working better with your latest code. good job! will release a new alpha release with it next week

[gentle bronze]

Thanks, I am glad that it helps :)

[slender iron]

💤

[gentle bronze]

Good night :)

[manic glacier]

jerryneedell

[adafruit/circuitpython] New comment on issue #1341: SPI not "deinited" on soft reboot for nrf boards

This issue has also been resolved on the feather_nrf52832. OK to close?

[manic glacier]

tannewt

[adafruit/circuitpython] New comment on issue #1341: SPI not "deinited" on soft reboot for nrf boards

Yes, thank you!

tannewt

[adafruit/circuitpython] Issue closed: #1341 SPI not "deinited" on soft reboot for nrf boards

[manic glacier]

ladyada

[adafruit/circuitpython] New comment on issue #1355: create driver for TMP006

@caternuson ok we still have some in stock, and we've already got some python code , so between the original py library and the tmp007 it should not be hard

[manic glacier]

notro

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

Running many unittest tests through the repl is working again.
Uploading files is working.

The script I use to upload, use os.stat() to check if files need uploading. When I rerun the script after it has uploaded it all files, I get into trouble.

It consistently fails when stat'ing these files:
/lib/unittest/test/__init__.mpy
/lib/unittest/test/dummy.mpy
/lib/test/imghdrdata/python.pbm

(I've had 2-3 times out of over a hundred times that one of these succeeded)
...

[haughty bobcat]

https://github.com/dhalbert/CircuitPython_LCD

can someone help me make this an mpy and use it?

[tidal kiln]

you can get mpy-cross builds from the CP release page:
https://github.com/adafruit/circuitpython/releases/latest
and then you just run that on the .py file to create a .mpy

[haughty bobcat]

Instructions on how to use MPY cross on windows?

[tidal kiln]

i think you just get the exe:
https://github.com/adafruit/circuitpython/releases/download/3.1.1/mpy-cross-3.x-windows.exe
and then run it in a command window

[haughty bobcat]

I don't know how to do that

I got the exe

[tidal kiln]

have you ever done anything in the command window before?

[fierce girder]

@slender iron I will be flying to Berlin and miss the meeting, but thanks to @meager fog for helping out with BeagleBone in Blinka. GPIO, I2C and SPI are now working in Blinka for BeagleBone Black

I tested with BME280 in i2c and SPI mode. Additional tasks: test other sensors to make sure the CircuitPython drivers work too on BeagleBone

[haughty bobcat]

yes....but few very specific things

most of the time copy and paste

[tidal kiln]

that library looks old - why are you wanting to use it?

[haughty bobcat]

PCF8574 support

[tidal kiln]

can you open the command window and navigate to where to saved mpy-cross?

[haughty bobcat]

with CD?

or just with a file path

[tidal kiln]

yep. with cd

[haughty bobcat]

ok

I can do that

[tidal kiln]

and then just try running it, no arguments, and see what it outputs

[haughty bobcat]

no input file

[tidal kiln]

cool. it runs.

have you downloaded that repo yet?

[haughty bobcat]

yes

[tidal kiln]

did you clone it, or download zip?

[haughty bobcat]

download zip

I just copied the py files

thats what I need right?

[tidal kiln]

i think so

if you unzip those .py files into the same folder where you have mpy-cross, it will be easier to run

[haughty bobcat]

yup

[tidal kiln]

then you can just run it again, but this time specify one of the filenames

[haughty bobcat]

how do I specify the file names?

just tack it on the end?

[tidal kiln]

yep

mpy-cross foo.py

in general

your actual executable may have a different name

[haughty bobcat]

ok it worked

[tidal kiln]

cool you should see .mpy files now?

[haughty bobcat]

yes

[tidal kiln]

that's it then. to use, just copy the .mpy files to your board

[haughty bobcat]

``from lcd.lcd import LCD
from lcd.i2c_pcf8574_interface import I2CPCF8574Interface

from lcd.lcd import CursorMode

Talk to the LCD at I2C address 0x27.

The number of rows and columns defaults to 4x20, so those

arguments could be omitted in this case.

lcd = LCD(I2CPCF8574Interface(0x27), num_rows=4, num_cols=20)

lcd.print("abc ")
lcd.print("This is quite long and will wrap onto the next line automatically.")

lcd.clear()

Start at the second line, fifth column (numbering from zero).

lcd.set_cursor_pos(1, 4)
lcd.print("Here I am")

Make the cursor visible as a line.

lcd.set_cursor_mode(CursorMode.LINE)``

this is the example code

I get: ImportError: no module named 'lcd.lcd'

Im probably missing something simple

[tidal kiln]

where did you put the files on your board? what folder?

[haughty bobcat]

I tried the files just in the lib folder, didnt work. Tried to put them in a folder called lcd in lib, didn't work. Tried to put them on the root of circuitpy too.

[tidal kiln]

where are they now?

[haughty bobcat]

root, and lib>lcd

[tidal kiln]

try without the lcd. part in front

ex: from lcd import LCD

or...put them all in a folder called CIRCUITPY/lib/lcd

then the example syntax should work

[haughty bobcat]

hmm

I seem to have satisfied it

[tidal kiln]

maybe it doesn't like a .mpy version of __init__.py

[haughty bobcat]

hmm nevermind

ValueError: cannot perform relative import```

what the heck

[tidal kiln]

https://github.com/dhalbert/CircuitPython_LCD/blob/master/lcd/i2c_pcf8574_interface.py#L27

you have 3 .mpy files, right?

[haughty bobcat]

yes

[tidal kiln]

one of them is called __init__.mpy?

[haughty bobcat]

yes

[tidal kiln]

move the other two into a folder called CIRCUITPY/lib/lcd

[haughty bobcat]

oh

it's satisfied again

no error

but also no display on my lcd

[tidal kiln]

the modules are importing at least. so must be something else.

[haughty bobcat]

yeah could be a million things

[tidal kiln]

yep. as a start. 😃

good luck

[exotic pumice]

I'm looking at this https://github.com/adafruit/Adafruit_ZeroDMA/blob/f7a3b774400988d0ad2ae0654c21281d6063b692/Adafruit_ZeroDMA.cpp#L13
First I had trouble understanding it at all, then I found out it's a #define of __attribute__ ((section(".hsram")))
So it's just telling it the variable is in RAM?
Aren't all non-static variables in RAM?

I guess it's arduino not cpy, but this is where all the knowledgable hardware ppl hang out

[exotic pumice]

and besides there's no #arduino

[main meteor]

You can define any section you want in the memory map. I could go look at it, but I'm guessing .hsram might be "high-speed RAM" or somesuch?

[exotic pumice]

yeah it is

but why would you have to specify something is RAM?

isn't it the default?

[main meteor]

In this case, it's specifying a particular section of RAM.

[exotic pumice]

hsram is just defined as 0x20000000

[main meteor]

Hmm, that doesn't sound right.

[exotic pumice]

#define HSRAM_ADDR (0x20000000u) /**< HSRAM base address */

https://github.com/robbie-cao/samd-touch/blob/34a9c30a51608215d9d6811895c577cca4f6b421/sam0/utils/cmsis/saml21/include_b/saml21j18b.h#L624

[main meteor]

Ah, that's where HSRAM begins, but it's not the same as the .hsram section of memory.

[exotic pumice]

would listing the sections with radare2 tell me?

[main meteor]

I'm not familiar with radare2. I think there's a file named something like memory.x the linker uses to get the memory map. I'm trying to find it.

Hmm, I found some arm scripts in Library/Arduino15/packages/arduino/tools/arm-none-eabi-gcc/4.8.3-2014q1/arm-none-eabi/lib/ldscripts but they don't look like the right ones.

[exotic pumice]

gr8

$ arm-none-eabi-readelf -S blinky_basic
There are 20 section headers, starting at offset 0x82f7c:
readelf: Error: Reading 800 bytes extends past end of file for section headers

radare 2 sez

[00:00] * pa=0x00001000 r-x va=0x00002000 sz=0x0a7c vsz=0x0a7c LOAD0
[00:01] . pa=0x00002000 rw- va=0x20000000 sz=0x0000 vsz=0x0004 LOAD1

[main meteor]

Looks like the loader scripts are expressed differently than the AVR ones.

Which board are you using? The boards.txt file uses that to look up the matching loader script.

[exotic pumice]

the ones I showed are cpx

[main meteor]

Ah, boards.txt has adafruit_circuitplayground_m0.build.ldscript=linker_scripts/gcc/flash_with_bootloader.ld

[exotic pumice]

no hsram

[main meteor]

Nope. I wonder where it's getting it.

There's a comment: ```/* Linker script to place sections and symbol values. Should be used together

• with other linker script that defines memory regions FLASH and RAM.```

I wonder where we find that other linker script

[exotic pumice]

yeah it's not defined in asf4 either

I did a recursive grep

should I try to find what it's really defined as in gdb?

or would it compile out

[main meteor]

It's a linker variable, so there wouldn't be anything in the symbol table for it.

[exotic pumice]

😦

also arduino produces bins not elfs so I can't even try

[main meteor]

I believe it produces elfs as well

[exotic pumice]

yeah it puts them in /tmp

you're right though, I told gdb to break on line 13 of Adafruit_ZeroDMA.cpp, and it goes to line 26

here we go

(gdb) info files
Symbols from "/tmp/arduino_build_395518/sketch_nov02b.ino.elf".
Local exec file:
    `/tmp/arduino_build_395518/sketch_nov02b.ino.elf', file type elf32-littlearm.
    Entry point: 0x6a7c
    0x00004000 - 0x00008d6c is .text
    0x20000000 - 0x20000110 is .data
    0x20000110 - 0x20000510 is .hsram
    0x20000510 - 0x20000da4 is .bss

[main meteor]

Wow, I didn't know the memory sections were placed in the elf file. Sure enough, there's hsram, although we still don't know where it comes from.

[exotic pumice]

yeah I knew elfs specify sections, that's kind of the point

[main meteor]

I learned something today, because I didn't know that.

[exotic pumice]

https://en.wikipedia.org/wiki/Executable_and_Linkable_Format

would it be a bad idea to just define it in a memory.x with those values?

I guess it depends how big the data segment is

because that's what it comes after

[main meteor]

It's already getting it somewhere, isn't it? Or am I missing what you're asking?

[exotic pumice]

oh, I forgot to mention I'm trying to rewrite the dma code in rust

[main meteor]

Aha!

I don't know if rust uses the same linker as gcc. If it does, you're probably golden as soon as you figure out the rust pragma to specify a memory section. If not, you can probably use the value from the header file you found earlier, but I don't know if the rust linker uses memory.x or something else.

[exotic pumice]

you can specify the linker and I'm already using a memory.x

[main meteor]

In that case, it's worth a try. I've modified memory.x myself for an equivalent sort of purpose.

[exotic pumice]

let me try and find out if it's dependent on the .data segment first

[main meteor]

I think there's a way to specify "begin this section after that one".

In case you're curious, I added a section to the MSP430 memory map to use some of the FRAM as non-volatile RAM.

[exotic pumice]

ok, I'd be interested to know how to do "this section after that one"

currently my memory.x looks like this

MEMORY
{
  /* Leave 8k for the default bootloader on the Circuit Playground Express */
  FLASH (rx) : ORIGIN = 0x00000000 + 8K, LENGTH = 256K - 8K
  RAM (xrw)  : ORIGIN = 0x20000000, LENGTH = 32K
}
_stack_start = ORIGIN(RAM) + LENGTH(RAM);

after I get the memory.x figured out I still need to know how to tell rust to place something in a particular section

[main meteor]

Right. And for that trick, I have no clue.

[exotic pumice]

yep

I've asked in rust discord and irc

ok, got that covered

[manic glacier]

karhohs

[adafruit/circuitpython] New star added

[exotic pumice]

maybe if I do that it will create the section too

let me try

[exotic pumice]

solution for the curious
memory.x:

MEMORY
{
  /* Leave 8k for the default bootloader on the Circuit Playground Express */
  FLASH (rx) : ORIGIN = 0x00000000 + 8K, LENGTH = 256K - 8K
  RAM (xrw)  : ORIGIN = 0x20000000, LENGTH = 32K
}
_stack_start = ORIGIN(RAM) + LENGTH(RAM);

SECTIONS {
     .hsram (NOLOAD): ALIGN(16) {
       *(.hsram);
       . = ALIGN(16);
        KEEP(*(.hsram));
     } > RAM
} INSERT AFTER .data;

blinky_basic.rs:

#[link_section = ".hsram"]
#[used]
pub static foo: u32 = 1;

gdb:

(gdb) info files
Symbols from "/home/paul/rust/atsamd/boards/circuit_playground_express/target/thumbv6m-none-eabi/release/examples/blinky_basic".
Local exec file:
    `/home/paul/rust/atsamd/boards/circuit_playground_express/target/thumbv6m-none-eabi/release/examples/blinky_basic', file type elf32-littlearm.
    Entry point: 0x29dc
    0x00002000 - 0x000020b0 is .vector_table
    0x000020b0 - 0x00002a2a is .text
    0x00002a2c - 0x00002a90 is .rodata
    0x20000000 - 0x20000000 is .data
    0x20000000 - 0x20000010 is .hsram
    0x20000010 - 0x20000014 is .bss

still not the best for a library to need changes to memory.x, but good enough for now

oh hang on, that's not right

darn

well maybe it is, but why is .data 0 length

I guess I can #[link_section] something else to .data and see if .hsram comes after or at the same spot

[manic glacier]

tannewt

[adafruit/circuitpython] New comment on issue #1330: call for the creation of the first "LEARN SESSION"

How about Thursday at 2pm Pacific? I'll do it as a twitch stream.

tannewt

[adafruit/circuitpython] New comment on issue #1346: Scratch Ram or Something

Great! I'd create a separate memory module for it and copy the nvm api. So, start with duplicating the nvm folder in shared bindings and create a corresponding folder in shared-module. Get as far as you can and then create a PR. That way we can comment directly on your code with suggestions.

[slender iron]

<@&356864093652516868> meeting notes doc is here: https://docs.google.com/document/d/1942qPdQf1ZW9HB3PwgsB_7nKN0VCM3cNQdEpI1Ep6LM/edit?usp=sharing

[manic glacier]

DakotahRedstone

[adafruit/circuitpython] New comment on issue #1330: call for the creation of the first "LEARN SESSION"

Thur. 1400PST/1700EST sounds excellent! I’ll bring the questions :)

( _What is this Twitch of which you speak? ) :O LOL. ;)

[marble hornet]

hey, i'm trying to clone 3.x so i can stably compile my code into binaries but i'm getting this error: ```$ git clone -b 3.x [e-mail]:adafruit/circuitpython.git
Cloning into 'circuitpython'...
[e-mail]: Permission denied (publickey).
fatal: Could not read from remote repository.

Please make sure you have the correct access rights
and the repository exists.```

is it not open to the public?

[timber mango]

@marble hornet use the https url.
with the '[e-mail]' you try to 'login with ssh' - and as the message tells: Permission denied (publickey) you have the wrong public key...
i think this is the problem. try with https://github.com/adafruit/circuitpython.git

[marble hornet]

@timber mango THANK YOU!!!!!!!!!!!

do you remember the line to run to recursively cone the sub repos?

[solar whale]

git submodule update --init --recursive also it may help to do git submodule sync

[marble hornet]

thanks Jerryn, can i ask the difference?

[solar whale]

I think the second just updates the links to the proper tags -- the first actually downloads the submodules.

[manic glacier]

dhalbert

[adafruit/circuitpython] New comment on pull request #1323: shared\-module/os: Fix os\.mkdir\('a/b'\)

@notro Do you think this is fine as is now? Removing the backslash support is OK with me.

[river quest]

Yay! We are 9,000+ humans! Thank you Adafruit community!

[manic glacier]

notro

[adafruit/circuitpython] New comment on pull request #1323: shared\-module/os: Fix os\.mkdir\('a/b'\)

@notro Do you think this is fine as is now?

Yes, it's not perfect but it fixes a corner case while retaining the functionality that the original issue asked for.
Some CPython stdlib tests use a/a paths, that's how I discovered this.

[faint ibex]

Hello, I'm trying to port CircuitPython over to my own custom board and I'm a little confused on what actually needs to be ported. I have been looking at this repo https://github.com/adafruit/Adafruit_Blinka and I see for the beaglebone that it imports Adafruit_BBIO and I'm wondering if that would need to be ported to my board first.

[idle owl]

@faint ibex Blinka is a wrapper library we wrote to make CircuitPython libraries work on single-board Linux computers. Are you trying to port CircuitPython to a custom microcontroller? Or a custom single-board linux computer?

[faint ibex]

Its a custom single board computer

[idle owl]

Ah ok. That's really neat! I'm not sure where you'd start with that. I haven't done much on that end. Someone else will probably have some better info though.

[faint ibex]

Ok, thanks. Yeah, I'm just a little confused about what libs are needed/used

[idle owl]

I think you need to add support to Blinka to expose the GPIO on your computer. But like I said I haven't done much on that end, so I'm unsure.

[faint ibex]

Yeah, I just saw some mention of libgpiod and the BBIO, so was also unsure

[timber mango]

@faint ibex thats correct

[marble hornet]

@faint ibex is that online? or have a website somewhere?

[timber mango]

as far as i have understand it adafruit has wrapped a already existing gpio library -

[faint ibex]

Its still early days for this board, but I do have a website groboards.com

Yeah, I figured it was a wrapper, just not sure which gpio lib I should go for as the base

[marble hornet]

thanks ☺ 😁

[faint ibex]

I figured I would copy more or less what the beaglebone does as thats the only linux computer in there to go off of

[timber mango]

find one that is allready integrated in blinka - hopefully already supports your processor - and use this as a astarting point 😉

i think with this you are on a good way

[faint ibex]

I don't think there is much support for this MPU yet on the maker side od things

I'm more or less starting from scratch on a lot of stuff

[gusty kiln]

looks like i'm out for the meeting today. power's out at my house and i don't think i'm going to get a stream working on the available coffeeshop bandwidth.

[main meteor]

Nice looking board. What are all the inductors for? Is that a SAMA CPU?

[faint ibex]

Its a SAMA5D27C-D1G and uses a PMIC for all the power rails and includes lipo charging/support.

[main meteor]

How many power rails are there?

[faint ibex]

3

1.8, 1.2 and 3.3

[main meteor]

Ah, that explains the inductors!

[faint ibex]

Yeah 😄

[tidal kiln]

lurking

[raven canopy]

<@&356864093652516868> and any who wish to join, weekly meeting in a couple minutes. Meeting notes are here: https://docs.google.com/document/d/1942qPdQf1ZW9HB3PwgsB_7nKN0VCM3cNQdEpI1Ep6LM/edit?usp=drivesdk&ouid=108087914205417918047

[errant grail]

No audio for me today.

[timber mango]

@hard heart

[gusty kiln]

missing the meeting today; my notes are in the doc.

[errant grail]

A quick question for In The Weeds : Is there a way to force reset of an OLED display (SSD1306, PID#931) upon instantiation rather than using the display’s dedicated reset pin? Would allow use of the STEMMA I2C interface without running an additional wire from a GPIO pin.

[raven canopy]

Stay warm and dry @brennen#127!

[gusty kiln]

thx. :)

[raven canopy]

Dear Awesome Budget Person, can I haz more monitor?

[errant grail]

Hmmm. We'll try this mic out here in the workshop...

[raven canopy]

Haha. I was on 15" until like a year ago when I got hand-me-down.

[hard heart]

Got a memory usage question. I am using the Adafruit Motor library to drive my Servo. I looked in the docs and there does not appear to be a Deinit() function for the Servo class. I've noticed that I can Deniit() the PWM pin associated with a servo, but I'm afraid that is leaving the Servo object behind in memory. Any suggestions on how to delete the allocation for the Servo instance itself? I care because I am recreating the Servo instance each time I make a function call to move the arm to a different location, and I'm not sure the old one is being GC'd.

[tulip sleet]

@hard heart deinit() gives up the pin and some internal stuff -- it does not delete the object. We really on python's memory mgmt for that. If you reassign something else to that servo object's variable, there's no issue - it will just go away during a garbage collection.

you might consider not re-creating the servo object each time -- make it an instance variable in some class, etc.

[hard heart]

So ServoName = null will set it up for GC? Thanks

[tulip sleet]

yes

[tawny creek]

can't make it today but extra hug reports to @slender iron @exotic pumice for helping me my feet started on debugging and figuring out what's going on with a SPI flash I'm using on a custom board last week! and group hug to the rest of the growing community

[manic glacier]

vhugo

[adafruit/circuitpython] New star added

[slender iron]

got it @tawny creek, thanks!

[onyx hinge]

hi guys! just lurking and listening, hug reports to everyone who is working on bug triage and code reviews! (I won't have any other reports during this meeting) @idle owl

[hard heart]

@tulip sleet I initially had the object initialized once and just drove the arm through changes to servo.angle. But I found that after it ran for a while, resetting the servo arm approximately once every 5 minutes, the servo would start to "jitter" even though it was not being commanded to change. That was at least wasting battery, so I decided to try deleting the servo between movements. That did seem to eliminate the jitter, but the board (Trinket M0) would eventually go bonkers. I figured it was likely a memory exhaustion issue, hence my question.

[raven canopy]

Just a gyronormous Group Hug this week!!

[tulip sleet]

@hard heart we are going to add an "off" mode for servos, so that you can turn off the power rather than have it actively hold a position. That sounds like it might help here.

[hard timber]

Just my first time listening in to learn how to help.

[tulip sleet]

https://github.com/adafruit/Adafruit_CircuitPython_Motor/issues/17

[hard heart]

Yep. That would help. In the meantime, my dodge should work if it just reassign the Servo instance to null. THanks for the tip!

[tidal kiln]

group

hug

[onyx hinge]

oops

[disconnects from the audio so he can listen to the insight landing livestream]

[raven canopy]

Last Week:

• Travelled much, ate more, and worked less
• Worked a little on Adabot-on-Travis:Bundle updates are failing due to credentials (in the weeds)
• Added Validators a la carte to Adabot

This Week:

• Finish aforementioned Adabot work
• Research/work Lady Ada’s “Adabot for Arduino” request
• Catch up on outstanding PRs (SPI FRAM comes to mind)

[hard timber]

Still listening in, just learning

😉

[neat folio]

Dakotah upvotes eat more work less

[pastel panther]

gotta run 👋

[raven canopy]

Yep. saw that too! Also, sudo: required is going away... @slender iron

[tidal kiln]

yah. just misc stuff.

[marble hornet]

"pin-plementation"

[errant grail]

https://youtu.be/ktjgc43j7EQ

[neat folio]

SMD = so many distractions 😦

[slender iron]

I think you could bitbang it instead @inland tusk

[raven canopy]

@solar whale yay for Particle boards!!!

[slender iron]

mine are slated to arrive on thursday!

[tulip sleet]

@inland tusk guessing that this is what you found: https://www.raspberrypi.org/forums/viewtopic.php?t=19908&start=25

[manic glacier]

insolor

[adafruit/circuitpython] New star added

[tidal kiln]

@errant grail very nice!

[errant grail]

@tidal kiln Thanks!

[neat folio]

I have an after-the-weeds issue regarding the luck attached to peeling the wrap off in one piece on the acrylic parts in the trellis kit... 😉

[errant grail]

"First, you have the satisfying task of peeling off the protective paper from the acrylic pieces. If you peel off all of the paper without tearing them you will have seven weeks of good luck. Do not squander them."

[inland tusk]

@tulip sleet Thats right. But it confuses the busio module if you want to use other serial devices at the same time (ie. serial lcds).

[tulip sleet]

@inland tusk yah, just seems to be a real deficiency of the RPi in this case

[inland tusk]

@tulip sleet Yes.

[tulip sleet]

https://github.com/adafruit/Adafruit_CircuitPython_SSD1306/blob/master/adafruit_ssd1306.py#L140

[slender iron]

👍

[errant grail]

Thanks!

[solar whale]

@idle owl well done!!

[raven canopy]

Thanks everyone!! Have an excellent week!

[hard timber]

thank you!

[inland tusk]

Thanks to @tulip sleet

[raven canopy]

Agree with all; well done @idle owl!

[fluid helm]

Should this work with Crikit?

Sorry I missed the meeting, had homework today 😦

[idle owl]

No worries!

[tidal kiln]

tree trunks is yellow!

[solar whale]

@fluid helm depends on what is providing the power. It will power the crickit fine but if it only providing .5 Amps you may have issues with some devices.

[tulip sleet]

also make sure the polarity of the barrel jack is correct

[neat folio]

cats! can't live with 'em... can't .... 😉

[raven canopy]

👋

[fluid helm]

I've got a charger that can give at least 2A

[manic glacier]

tannewt

[adafruit/circuitpython] Pull request opened: #1356 Update to Xenial on Travis

[fluid helm]

with the USB cable in that it should be fine i'd think

Don't know much about it all so if someone can prove me wrong, go for it 🙂

[inland tusk]

@slender iron you went to 16.04 ubuntu, I just moved from 16.04 to 18.04

[slender iron]

ya, for travis

we were on trusty which is 14.10

or is it 16.04?

¯_(ツ)_/¯

[inland tusk]

@slender iron II stay with LTS releases

[raven canopy]

14: Trusty
16: Xenial
18: Bionic

[inland tusk]

Correct

[slender iron]

it is LTS I think

just a newer one

[raven canopy]

Yeah, it is.

[tidal kiln]

LTS = xx.04

[inland tusk]

That is what gave me a hard time when I was setting my volumnes

[onyx hinge]

ooh I'll have to take a look at whether having Xenial lets me do something on travis I wanted to do for another project...

[inland tusk]

It might very well be the case. I know that they changed the WM to Gnome version 3 in Bionic

[manic glacier]

dhalbert

[adafruit/circuitpython] New comment on pull request #1356: Update to Xenial on Travis

well, not faster :frowning_face:

trusty:

xenial:

[errant grail]

@tulip sleet Thanks for the ideas for the SSD1306 reset. The code didn't make a difference, but putting a 1uF cap from the breakout board's reset pin to ground seemed to work. Given the internal 10K pullup, that provided the ~10ms reset signal it needed. I'll need to run a few more tests to verify. Also want to see if it impacts normal use of the reset pin or if it's a bodge just for the Stemma implementation.

[manic glacier]

sabas1080

[adafruit/circuitpython] Pull request opened: #1357 Added board Electronic Cats CatWAN USB Stick

Hi

We've ported our CatWAN USB Stick board to Circuit Python, support LoRa and LoRaWAN, working with library RFM9x.

https://github.com/ElectronicCats/CatWAN_USB_Stick

Thanks

jepler

[adafruit/circuitpython] New comment on pull request #1356: Update to Xenial on Travis

Ooops! thanks for spotting this. Could be its own PR, if the update to enial is held back for any other reason.

[tidal kiln]

hey @idle owl - got some TMP006 driver code worked up. needs a home. can you make an empty repo please. i'll even try and delete README.md this time. 😃

[manic glacier]

DuckRogers

[adafruit/circuitpython] New comment on issue #1118: Frequency related pulseio PWMOut\.duty\_cycle does not retain set values accurately

Unfortunately, it still happens. I finally got a chance to test this again (on 3.1.1)
Caveat: I am running it on a M4 Feather now instead of a Metro, doubt that would change it, but...

[idle owl]

@tidal kiln Please send me an email with that request though I may get to it before you send me the email.

[sacred edge]

Can I run Neopixel's (Sticks) with the CPX?

[idle owl]

@sacred edge Yes.

[sacred edge]

is the process of using the individual neopixels a big soldering job? and can I drive them from the CPX?

[idle owl]

Individual NeoPixels are a bit of work, yes. But you can drive them from the CPX as well.

[tidal kiln]

the software aspects are simple - it's the same as what you'd do for any other board

[sacred edge]

Thanks Kattni, The Neopixel "Jewel" looks like it might be exactly what I am needing.. CPX fine for that too??

[tidal kiln]

but for the connections - you might able to use alligator clips

[idle owl]

I mean the individual NeoPixels are a bit of soldering work. Software is the same between the individual ones and any other formfactor.

Yep Jewel will work as well.

You can either solder alligator clips to the jewel and connect them to the CPX or you can solder directly to the CPX if you want

or alligator clips to breadboard... lots of options

[sacred edge]

wonderful!... I was looking for a PIR, but you folks are out of stock at this time..

[tidal kiln]

another idea is to use screws that pass thru the holes of the CPX

at the jewel end, you'd want to solder though

[sacred edge]

thanks @tidal kiln

basically I want to use the PIR for motion, to trip the jewel into a subdued light setting for a safety light for my wife to get from bed to bathroom easily..

just figuring it out at this point...

[tidal kiln]

do you think you need more neopixels than what's already on the CPX?

also, check this out:
https://learn.adafruit.com/circuitpython-powered-gemma-nightlight/nightlight-pro-deluxe-cpx-edition

[sacred edge]

Thanks @tidal kiln I'll check it out...

question, if you have multiple versions of Python on the computer, would that confuse "Mu"??

[sacred edge]

I cannot get Mu 1.01 to launch - It just sits there (Not Responding) HELP!!

[tidal kiln]

@sacred edge how did you install Mu?

[sacred edge]

just according to the directions on the Mu Installation page... using the 64 bit installer..

[tidal kiln]

what OS?

[sacred edge]

Win 10

[tidal kiln]

and this is with trying to run it right after the install?

[sacred edge]

yes, it would not even get to the "Choose ADafruit" screen..

It has worked before, and then stopped, so I uninstalled it , re-installed it...

same result,

just says Mu 1.02 - Untitled - (Not Responding)

rather 1.01

[tidal kiln]

maybe related?
https://github.com/mu-editor/mu/issues/665

[tidal kiln]

do you have a board currently plugged in?

[sacred edge]

yes, green light, and the drive shows up..

[tidal kiln]

try unplugging it and see if this happens:
https://github.com/mu-editor/mu/issues/665#issuecomment-426683975

[sacred edge]

drive disappears..

wow!! whats up with that, it is there now...

[tidal kiln]

don't know. but looks like a known issue.

[sacred edge]

yeah, so I will try to plug it in now and see what happens..

When you use Mu, what is the extension you use to save the code in?

uf2?

[tidal kiln]

typically you are saving directly to a circuitpython file on the board itself, so a .py extension

[sacred edge]

Thanks!!... You are the Man!! @tidal kiln!!

[tidal kiln]

np

[steady pivot]

Are we doing MicroPython here too?

[solar whale]

@steady pivot ask away. There are some folks here with Micropython experience.

[steady pivot]

Not looking for help, just a community.

[solar whale]

Ah. Welcome. The focus here is Circuitpython but we respect our heritage 😉

[pastel panther]

Also plenty of people here that go both ways

[steady pivot]

Great! I'll have to buy a board now, lol

Thanks, I'll stick around 😃

[manic glacier]

FreakHandy

[adafruit/circuitpython] Fork created: FreakHandy/circuitpython

[pastel skiff]

Hi folks, a few weeks ago it was mentioned on the Adafruit video channel that Edublocks was in final phases of testing a circuit python version of their block programming tool, I have poked around and really cant find much about it, does anyone have any insight on how/when/where this will be available?

[pastel panther]

I believe @fluid helm is the person responsible for Edublocks and should be able to let you know what the status is

[manic glacier]

dhalbert

[adafruit/circuitpython] Pull request closed: #1356 Update to Xenial on Travis

dhalbert

[circuitpython:master] 2 new commits

0c55ddf Update to Xenial on Travis - tannewt
027ed82 Merge pull request #1356 from tannewt/xenial_tr... - dhalbert

[cunning trail]

Question: Is there an equivalent function in CP for arduino map() for scaling number ranges? Closest thing I found was interp function in python numpy library which isn't one of the CP libraries. Thanks.

[tidal kiln]

yep. map_range in simpleio:
https://circuitpython.readthedocs.io/projects/simpleio/en/latest/api.html#simpleio.map_range

[timber mango]

Just picked up Feather LoRa for Cyber Monday. Super stoked! 📻

[cunning trail]

@tidal kiln thanks, I'll give it a try tomorrow.

[manic glacier]

tannewt

[adafruit/circuitpython] Pull request opened: #1358 Fix QSPI on Feather nRF52840

We were writing with quad page program including the address (0x38)
which is unsupported by the GD25Q16C but it is supported by the
flash on the DK. So, we use the single address, quad data command
(0x32).

[manic glacier]

turbinenreiter

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

On 027ed82 it's somewhat different.

If I copy-paste above example line-per-line to the REPL, nothing crashes, but I get this error:

>>> ss.send("GET / HTTP/1.0\nHost: %s\n\n" % host)
Traceback (most recent call last):
  File "<stdin>", line 1, in <module>
OSError: [Errno 5] Input/output error

... which might not be a problem in the code, but could also be our network. I do get an IP and True on connected and dhcp.

But when I run it from code.py it breaks the s...

[manic glacier]

dhalbert

[adafruit/circuitpython] Pull request closed: #1358 Fix QSPI on Feather nRF52840

dhalbert

[circuitpython:master] 2 new commits

d446d32 Fix QSPI on Feather nRF52840 - tannewt
4e8531f Merge pull request #1358 from tannewt/fix_840_qspi - dhalbert

[pastel panther]

@tulip sleet or @slender iron this sounds maybe dumb but is there a limit of some sort on the number of methods in a module? I'm running into a strange scenario on my feather m0 express with 4.0.0-alpha.3-9-g3ca74d869 where I'm working on a driver and when I import it I get a memory allocation error if I have more than 11 methods. Uncommenting the 12th causes the import errors and so far it doesn't seem to matter which method

[slender iron]

No limit but I’m not sure of the allocation policies for the dictionary underneath. It could be grabbing more than it needs

[pastel panther]

It's strange because it's trying to allocate < 300 bytes (depends on which functions are uncommented) but mem_free is showing like 13k when everything is done

[slender iron]

Don’t forget about fragmentation

That’s 300 continuous

[pastel panther]

right

[slender iron]

And loading costs temp memory too

That’s when mpy can be helpful

[pastel panther]

is CTRL-D-ing the REPL going to be getting things back to normal for valid tests or should I be going through a full reset?

In the end it doesn't matter for this driver because I only hit that apparent limit because I was in the middle of replacing some methods and had the old ones still lying around. That said I had planned to do some refactoring which might not be possible now

[slender iron]

Reload does reset the heap entirely

[pastel panther]

kk

is this worth creating an issue for? I'm leaning towards yes since I don't think I'm doing anything terribly abnormal

Now I'm just getting blank MemoryErrors with no message:

>>> import adafruit_adxl34x
Traceback (most recent call last):
  File "<stdin>", line 1, in <module>
MemoryError:
>>>

hmm, resetting fixed that...

Maybe that was a false positive; I'll keep digging

gotta run for now though

@slender iron what did you use to make the memory visualization that you did for 3.x?

[slender iron]

@pastel panther without a message means it didn't have enough space for the error string.

@pastel panther https://github.com/adafruit/circuitpython/blob/master/tools/analyze_heap_dump.py

thats what draws the heap

https://github.com/adafruit/circuitpython/blob/master/tools/output_gc_until_repl.txt can be used with gdb to capture the heap after every change

[meager fog]

@ruby lake hey CS i have a RAMPs q

and you're the only person i know with a RAMPS board

[ruby lake]

@meager fog hm?

[meager fog]

we wanna make sure that the grand central can work with a RAMPS shield

but we're not sure of some usage cases

all the stepper pins are just plain GPIO, thermistors we have connected to ADCs. there's servo ports but we dont know if they're actually used for anything?

do heaters/fans need to be on PWM or are they straight on-off

[ruby lake]

If I recall the ramps firmware uses pwm for heaters and 2 fans

[meager fog]

ok cool, they'll be 8 bit PWM but thats plenty

[ruby lake]

I am more used to the Marlin fw, which also WMs

~PWMs

8bit is fine

[meager fog]

kk

wow this software is...somethin

https://github.com/MarlinFirmware/Marlin/tree/1.1.x/Marlin

thats a lot of files! o_O

[ruby lake]

the servo outputs are for options like auto bed-leveling, I've never used that myself

yeah the marlin source tree is kind of crazy

[meager fog]

k perhaps someone will be excited to do the port

[manic glacier]

oshadmon

[adafruit/circuitpython] Issue opened: #1359 use circuitpython on generic linux platform with FT232H

Machine Specs: Ubuntu 18.04 Desktop on an i7 based workstation.

Background: I'm using an Adafruit HTU21D-F Humidity/Temp sensor connected to a USB pluggable FT232H breakout board (C232HM-DDHSL‌-0 USB2.0/I2C cable). Adafruit GPIO has support for FT232H (https://github.com/adafruit/Adafruit_Python_GPIO/blob/master/Adafruit_GPIO/FT232H.py)

Question: I want to use circuitpython, in particular, the HTU21D-F code (https://github.com/adafruit/Adafruit_CircuitPython_HTU21D/blob/m...

[prime flower]

marlin is huuuuge

[manic glacier]

tannewt

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

There is a ring buffer on the CDC output that could be filling up. Is that all allocated as one string and printed?

[idle owl]

@tidal kiln https://github.com/adafruit/Adafruit_CircuitPython_TMP006

[tidal kiln]

woot. thanks @idle owl

[manic glacier]

tannewt

[adafruit/circuitpython] Pull request closed: #1351 json: enable on nrf boards and "large flash" M4 samd boards

tannewt

[circuitpython:master] 3 new commits

78972cc ports/atmel-samd: enable json module on M4 boar... - jepler
52fd151 ports/nrf: enable json module on nrf boards gen... - jepler
60480ff Merge pull request #1351 from jepler/issue1307-... - tannewt

tannewt

[adafruit/circuitpython] New comment on issue #1330: call for the creation of the first "LEARN SESSION"

Great! I'll stop doing reviews until then. :-)

https://www.twitch.tv/tannewt

tannewt

[adafruit/circuitpython] New comment on issue #1359: use circuitpython on generic linux platform with FT232H

Here or the forum is fine. @ladyada is probably the person who knows best.

tannewt

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

@nickzoic Any ideas?

oshadmon

[adafruit/circuitpython] New comment on issue #1359: use circuitpython on generic linux platform with FT232H

Thank you. I have also posted the question on the forum

notro

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

That string was a combination of read and write (ie. the session):

read:  b'>'  # previous raw repl prompt
write: b'\x01'  # Ctrl-A
read:  b'raw REPL; CTRL-B to exit\r\n>'
write: b'print(__import__(\'_os\').stat(\'/lib/unittest/test/test_assertions.mpy\'), end="")'
write: b'\x04'  # Ctrl-D
read:  b'OK'
read:  b'(32768, 0, 0, 0, 0, 0, 10790, 1543174556, 1543174556, 1543174556)\x04'
read:  b'\x04'

tannewt

[adafruit/circuitpython] New comment on issue #1118: Frequency related pulseio PWMOut\.duty\_cycle does not retain set values accurately

Now that I read this I realize its by design. The number you read back is the actual duty cycle being output. As you've found, the accuracy varies somewhat with frequency. This is due to the top value of the underlying counter varying to achieve the target frequency. I think this is how it should work, if we store the input value and then return it we'd be giving false info back.

If you want to ensure even steps from your end I recommend storing the last set value. Will that work for you?
...

tannewt

[adafruit/circuitpython] Issue closed: #1118 Frequency related pulseio PWMOut\.duty\_cycle does not retain set values accurately

tannewt

[adafruit/circuitpython] New comment on pull request #1323: shared\-module/os: Fix os\.mkdir\('a/b'\)

@dhalbert Can this be merged?

tannewt

[adafruit/circuitpython] Pull request closed: #1357 Added board Electronic Cats CatWAN USB Stick

tannewt

[circuitpython:master] 3 new commits

1da2425 Add Electronic Cats CatWAN USB Stick - sabas1080
a053eb2 auto-built and documentation - sabas1080
913b513 Merge pull request #1357 from ElectronicCats/ad... - tannewt

[timber mango]

does anybody know what is the fastest PWM frequency i can generate with an ItsyBitsy M4?

(with CircuitPython)
in my first test i got 6MHz output.

i only need it as clock output (50% fixed duty-cycle )

i had a look at the datasheet of the SAMD51 - but that is really a big thing to read and navigate...
( i managed to get timers setup in AVRs.. but my feeling is that on this chips its another level of complexity..)

[manic glacier]

ladyada

[adafruit/circuitpython] New comment on issue #1359: use circuitpython on generic linux platform with FT232H

hiya this is something we do want to do soon, it requires adding to Adafruit-Blinka and its not easy :( you can subscribe to https://github.com/adafruit/Adafruit_Blinka to get notified when we'll add it. if you have any links to good FT232H examples for python3, that would be rad. or, if you can contribute code, that would be really great :)

[sacred edge]

I fear my CPX is DOA.. 😦

[idle owl]

@sacred edge Why?

[sacred edge]

I write a py file over to it, and get no response... I do have a green light showing on the board.

[idle owl]

Do you have all of your code backed up?

[sacred edge]

I save a simple (Blink) .py file and no response from the board..

yes..

[idle owl]

It's really hard to brick these boards. But you can corrupt the file system from time to time.

Ok, can you still get to the REPL through Mu?

Is the CIRCUITPY drive showing up?

[sacred edge]

yes the drive shows..

[idle owl]

And you're not getting any errors when you save the file? It's simply not responding?

Errors like "file system is read only"

[sacred edge]

Mu has not worked since a couple of days ago.. 😦

[idle owl]

Hmm ok. What OS are you on?

[sacred edge]

Win10

[idle owl]

What are you using to edit code?

[sacred edge]

when I first got the board, everything worked great... then now it doesnt

[idle owl]

We'll get it figured out. What program are you using to edit code?

[sacred edge]

right now since Idont have Mu just "Notepad"

[idle owl]

I seem to remember seeing from an earlier discussion that the Mu problem is a known issue, and they're likely working on it.

Ok, let's do a couple of things here.

[sacred edge]

ok

I do know that I tried to "delete" my Py code on the board, but it doesn't seem to be deleted..

[idle owl]

First of all, download Atom and use that as your code editor. It has a lot of extras with it, but it provides a simple interface that shouldn't confuse or intimidate you. I started with Atom when I first got started programming. Notepad can cause issues in the code.

Second, we're going to wipe the board completely and start fresh.

Let me know when you're ready and we'll go through the steps to wipe the board. This will erase everything on the board, and is why I asked if you had your code backed up.

(Atom doesn't have the REPL built in like Mu, but it won't cause the issues that Notepad causes.)

[sacred edge]

ok, atom is installing now

[idle owl]

Ok. We can start on the board if you're ready

[sacred edge]

ok, lets do it

[idle owl]

Download this file: https://cdn-learn.adafruit.com/assets/assets/000/048/745/original/flash_erase_express.ino.circuitplay.uf2?1512152080

Double-tap the reset button on your Circuit Playground Express, so that the CIRCUITPY drive disappears and the CPLAYBOOT drive appears.

[sacred edge]

atom is on board and up... the Cplayboot is up

[idle owl]

Drag the .uf2 file you just downloaded to CPLAYBOOT. A NeoPixel should turn blue for ~15 seconds, and then it should flash green.

[sacred edge]

ok, so far so good

[idle owl]

Download a new copy of CircuitPython for Circuit Playground Express: https://github.com/adafruit/circuitpython/releases/download/3.1.1/adafruit-circuitpython-circuitplayground_express-3.1.1.uf2

[sacred edge]

blinking green

[idle owl]

Excellent!

Now double-tap the reset button again to bring up the CPLAYBOOT drive.

[sacred edge]

ok

[idle owl]

drag the new CircuitPython .uf2 to the CPLAYBOOT drive.

CPLAYBOOT should disappear and CIRCUITPY should return

[sacred edge]

CircuitPY drive up

[idle owl]

Excellent. Now, go into Atom, and create a new file containing the following: ```python
import time
from adafruit_circuitplayground.express import cpx

while True:
cpx.red_led = True
time.sleep(0.5)
cpx.red_led = False
time.sleep(0.5)```

Save it as code.py to the CIRCUITPY drive.

Sorry, had extra spaces in there initially.

[sacred edge]

ok, saved...

[idle owl]

Blinking red LED?

[sacred edge]

no

[idle owl]

hmm.

[sacred edge]

green light only

[idle owl]

the green "on" led?

the tiny one, I mean

not a NeoPixel, right?

[sacred edge]

right just beside the connector

[idle owl]

Let's make sure it's not a library issue. Download this: https://github.com/adafruit/Adafruit_CircuitPython_Bundle/releases/download/20181126/adafruit-circuitpython-bundle-3.x-mpy-20181126.zip

then unzip it, and copy the lib folder inside to CIRCUITPY

[sacred edge]

that is my drive contents

[idle owl]

ok did you paste what I had directly? because I had to fix it, it had too many spaces in it

[sacred edge]

I fixed the spaces and re-saved

[idle owl]

That's what the drive should look like right now, so that's correct

ok

download the library bundle and copy it to CIRCUITPY

We may have to walk through connecting to the REPL because that's where errors go and there's no indicator LED on the CPX.

[sacred edge]

just the lib folder?

[idle owl]

correct

that shouldn't change anything because the cpx library is built in, but it's worth trying.

[sacred edge]

copying now

[idle owl]

ok

[sacred edge]

I super appreciate your taking care of me!!

[idle owl]

You're welcome! Hopefully we'll get this figured out.

Ok, let's get you connected to the REPL because without Mu working, you'll want it anyway.

[sacred edge]

ok

[idle owl]

Open up "Windows Device Manager"

and click on "Ports (COM P& LPT)" to expand it. What's below it?

[slender iron]

@timber mango it should be able to go faster than 6. The core is 120

[sacred edge]

USB aSerial Device( Com 6)

[idle owl]

And that's it?

[sacred edge]

yes under ports.. and a Communications Port (Com 1)

[idle owl]

Perfect.

Go to this part of the Welcome guide: https://learn.adafruit.com/welcome-to-circuitpython/advanced-serial-console-on-windows#install-putty-14-8 and click on "latest version of PuTTY" to install PuTTY. Then open PuTTY, and fill in the information like seen in the screenshot right after it. Except for "Serial line" you'll put "COM6"

[sacred edge]

ok, now putty is open to a configuration page... next?

[idle owl]

fill it in like in the screenshot, except put COM6 for "Serial line"

[sacred edge]

ok serial line com6

[idle owl]

and click "serial" for connection type

if it isn't already

and put 115200 for speed

then click "Open"

and it should open a terminal window

[sacred edge]

click and I get a "dinging bell" from my system..

[idle owl]

Um... you filled in Serial Line, Speed, clicked "serial" under "Connection type"?

oh that's not the right window

oh click "Session"

at the top of the list

[sacred edge]

found it!!

[idle owl]

that's where you fill in the info, is in that first window.

[sacred edge]

Puty is open

[idle owl]

ok, click in the terminal window so it's active, and then type "ctrl+D" to reload

[sacred edge]

soft reboot

Auto-reload is on. Simply save files over USB to run them or enter REPL to disable.
code.py output:
Traceback (most recent call last):
File "code.py", line 3
SyntaxError: invalid syntax

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

[idle owl]

Hooray!

Ok there's a syntax error in your code.py

on line 3

[sacred edge]

import cpx

[idle owl]

Also, you can format code in Discord by putting three backticks on either side of it, that's the one found near the 1 key on US keyboards. code block

[sacred edge]

`

[idle owl]

paste the entire contents of your code.py into discord surrounded by three backticks on each side

[sacred edge]

from adafruit_circuitplayground.express
import cpx

   while True:
       cpx.red_led = True
       time.sleep(0.5)
       cpx.red_led = False
       time.sleep(0.5)```

[idle owl]

there it is! Ok, the import cpx should be on the same line as the from adafruit_etc

it's all one line.

[sacred edge]

from adafruit_circuitplayground.express import cpx

   while True:
       cpx.red_led = True
       time.sleep(0.5)
       cpx.red_led = False
       time.sleep(0.5)```

[idle owl]

The spacing is still wrong on the rest of it

from adafruit_circuitplayground.express import cpx

while True:
    cpx.red_led = True
    time.sleep(0.5)
    cpx.red_led = False
    time.sleep(0.5)```

Python is a stickler for indentation 😄

But now that you have the serial connection through PuTTY, you'll know when it's an error in your code that's causing the problem!

And you can troubleshoot it from that point instead of guessing.

[sacred edge]

thats good to have I am thinking.. So now just save code.py, and see what happens?

[idle owl]

yep

[sacred edge]

ok, stand by... Do I need to bump the button once or anything?

[idle owl]

nope

when you save it should reload

make a change and then look at the serial output as you save it

either way check the serial output if it's still not working to see if you get an error

[sacred edge]

I had to do Ctl-D, and now it is blinking!!! Yay!!

[idle owl]

yay!

Ok, so for now, use Atom and connect using PuTTY so you have the serial console available to see what's happening with your code!

but keep an eye on Mu and hopefully they'll get that issue resolved and you can go back to using it because it's easier

[sacred edge]

Is putty where the "Serial" prints will show up now?

[idle owl]

Yes!

[sacred edge]

I liked the graphing built into Mu....

[idle owl]

Put above the while True: a print("Hello, world!")

(single back tick on either side of something makes it inline code)

Yah, the graphing is nice. That one is unique to Mu.

[sacred edge]

yeah, that print statement worked!! You Rock Kattni!!! You got a more dedicated customer!!!!

[idle owl]

I'm glad we got it sorted! Great job!

[sacred edge]

Adafruit!! #1!!!

Thanks again!!

[idle owl]

You're very welcome!

[hard heart]

Anyone familiar with the meaning of the various colors of the Dotstar LED on the Trinket M0? I ask because I have a CP program running just fine on it while connected to an active Serial Monitor, but an hour or two after disconnecting from the SM and reverting to a 6V battery pack (with very fresh batteries) the program seems to become unresponsive and the Dotstar starts flashing colors in a jittery/random sort of way. Frustrating that the error condition doesn't seem to happen while connected to a Serial Monitor.

Per the Trinket M0 specs: RGB DotStar LED - Instead of an always-on green LED we provide a full RGB LED. You can set it to any color in the rainbow. It will also help you know when the bootloader is running (it will turn green) or if it failed to initialize USB when connected to a computer (it will turn red). By default after you boot up the Trinket M0 it will turn a lovely violet color.

[idle owl]

@hard heart https://learn.adafruit.com/adafruit-trinket-m0-circuitpython-arduino/troubleshooting#circuitpython-rgb-status-light-18-18 Explains what the status blinking means. I'm not sure whether this is what is happening for you though.

[hard heart]

@idle owl Thanks! I knew that had to be in there somewhere. Thought I had seen that but couldn't remember where. Troubleshooting, of course! <bangs head> Thanks again!

[idle owl]

You're welcome!

[tidal kiln]

@tulip sleet this thread just got some updated info if you can take a look:
https://forums.adafruit.com/viewtopic.php?f=8&t=143708&start=15#p712022
summary:

• Win10 + Trellis M4 -> boardnameBOOT locks up
• firmware was updated on a Win 7 machine
• none of the known problematic software is installed
• a CPX and Trinket work as expected on same Win10 machine

[manic glacier]

DuckRogers

[adafruit/circuitpython] New comment on issue #1118: Frequency related pulseio PWMOut\.duty\_cycle does not retain set values accurately

Your statement seems to imply that the read back value of the actual duty
cycle would fluctuate slightly because the underlying target counter is
fluctuating? So if I repeatedly give it the same value at the same
frequency, and read back the result, then, because the underlying target
counter is fluctuating, the read back value would fluctuate?

I did a series of tests involving setting the duty cycle to a constant and
printing out the value. I did these at multiple frequencies and with a
r...

dhalbert

[adafruit/circuitpython] Pull request closed: #1323 shared\-module/os: Fix os\.mkdir\('a/b'\)

dhalbert

[circuitpython:master] 2 new commits

28383af shared-module/os: Fix os.mkdir('a/b') - notro
f6c73b3 Merge pull request #1323 from notro/os_mkdir_a_b - dhalbert

dhalbert

[adafruit/circuitpython] New comment on pull request #1329: Support OSError subclasses and attributes

@tannewt This is OK by me in the current form. OK by you?

[timber mango]

@slender iron yeah that was my thought also...

Adafruit CircuitPython 3.1.1 on 2018-11-02; Adafruit ItsyBitsy M4 Express with samd51g19
>>> import board
>>> import pulseio
>>> pwm = pulseio.PWMOut(board.D9, duty_cycle=(2 ** 15), frequency=(6000 * 1000))
>>> pwm.frequency
6000000
>>> 

works :-)

Adafruit CircuitPython 3.1.1 on 2018-11-02; Adafruit ItsyBitsy M4 Express with samd51g19
>>> import board
>>> import pulseio
>>> pwm = pulseio.PWMOut(board.D9, duty_cycle=(2 ** 15), frequency=(6001 * 1000))
Traceback (most recent call last):
  File "<stdin>", line 1, in <module>
ValueError: Invalid PWM frequency

does not - so i think the limit is the 6MHz - where it comes from - i don't know 😉

[vapid juniper]

Hi guys! I just want to share our great success with the CPX

[tulip sleet]

@tidal kiln I'm confused that it "works as an Arduino". Do they mean on Win10? But TRELM4BOOT locks up Would be useful to know if the Win10 machine is USB-C + an adapter. Could try a USB hub on the Win10 machine and see if it works better. It would be good to disable Norton temporarily and see if that helps. It could be some kind of USB timing problem but that's weird. Can a .uf2 be dragged to TRELM4BOOT? Try this latest build: https://adafruit-circuit-python.s3.amazonaws.com/bin/trellis_m4_express/adafruit-circuitpython-trellis_m4_express-en_US-20181127-4e8531f.uf2

[vapid juniper]

With a program we did called Invent to Learn https://codeparkhouston.org/blog/2018/10/5/invent-to-learn

and also lesson plans we made

http://bit.ly/invent2learn

We started with simple minecraft pi and python, and ended up showing students how to collect store and visualize data with their CPX

and tied everythign together with python.

[tidal kiln]

@tulip sleet I think they mean they can program it on the Win10 machine with Arduino. But any attempt to convert to CP fails as the TRELM4BOOT folder locks up when they try to drag over firmware.

[vapid juniper]

sorry if this is the wrong channel to share.

[tulip sleet]

@tidal kiln All I saw was that trying to read INFO_UF2.TXT causes a lockup, but not about dragging to it.

[prime flower]

@vapid juniper nice!!

[tulip sleet]

@vapid juniper This is great! Could you post this in https://forums.adafruit.com/viewforum.php?f=48

[tidal kiln]

hmmm. good point....i think i was assuming that.....

[tulip sleet]

it might be due to some weird editor they were using to try to read that file, or something like that

[tidal kiln]

@vapid juniper nice! esp. like the use of data plotting!

[vapid juniper]

@tulip sleet will do!

@tidal kiln thank you! 😃 our students really enjoyed it too. it was awesome hearing them talk about charts and understanding trends, beyond pie chart that were taught to them in school.

they also got to keep their CPX at the end htanks to our site partner Fab Lab Houston.

hopefully they'll do more stuff with it.

[tulip sleet]

@timber mango @slender iron 6MHz limit is hardwired in circuitpython/ports/atmel-samd/common-hal/pulseio/PWMOut.c. Not sure where that number comes from. It could be computed rather than hardwired, based on the minimum divisor and the input clock to the TC/TCC.

[timber mango]

ok for now i go to bed.

tomorrow i will have a look and try to understand the code and see if i get someting modified to work 😉

thanks @tulip sleet for the information and pointer!

[tulip sleet]

@timber mango (I corrected a typo in the path above)

[timber mango]

@tulip sleet thanks :-)
at a first look i think the problem is that the code currently tries to get the highest resolution out of the timer - this limits the frequency. so eventually it is needed to specify this somehow another way... (make a subclass that does the same but with highspeed as priority...)
https://github.com/adafruit/circuitpython/tree/master/ports/atmel-samd circuitpython/ports/atmel-samd/common-hal/pulseio/

[tulip sleet]

are you just trying to get a square wave at max frequency out?

[timber mango]

yes

[tulip sleet]

running continuously?

[timber mango]

best case would be a square wave with up to 33MHz.
yes

[tulip sleet]

So rather than PWMOut, it's really just ClockOut, because the duty cycle is 50% always, right?

[timber mango]

correct - is there a ClockOut thing?

[tulip sleet]

no, I was thinking about SPI

but the clock doesn't run all the time

[timber mango]

ah yea makes sens-

iam using spi at the same time..

(as normal spi ...)

[tulip sleet]

this is that TLC chip still, right?

[timber mango]

😉 correct

its working here... so fare - but i would like to test the higher grayscale clocks...

https://raw.githubusercontent.com/s-light/magic_amulet_pcbs/master/LEDBoard_4x4_HD/photos/P1700258_small__TLC5957_controlled_by_ItsyBitsyM4CircuitPython.jpg

[tulip sleet]

well, for now, you can just remove that limit check and see if it helps. If so, we can think about a simple ClockOut or FrequencyOut or something.

[timber mango]

i try tomorrow 😃

thanks for your help!

[manic glacier]

winneymj

[adafruit/circuitpython] Fork created: winneymj/circuitpython

[manic glacier]

nickzoic

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

I'll have a look ...

OK so I've built Adafruit CircuitPython 4.0.0-alpha.3-25-gf6c73b35f-dirty (current master: it's only dirty due to some problem with the nrfx submodule, I'll worry about that later)

The code on the PR works:

import board
import busio
import wiznet

spi = busio.SPI(clock=board.SCK, MOSI=board.MOSI, MISO=board.MISO)
eth = wiznet.WIZNET5K(spi, board.D10, board.D11)
eth.connected

host = 'example.com'
import socket
fam, typ, pro, nam, sockaddr = socket....

[manic glacier]

winneymj

[adafruit/circuitpython] Fork created: winneymj/circuitpython

[hard heart]

Trinket M0, board locks up after a while when operating on 6v battery pack. The Green LED stays constantly lit, the Red LED flashes on and off slowly, and the DotStar RGB LED flashes alternately Yellow and Purple. The Purple color occurs in sync with the Red LED coming on. This does not seem to match any of the info on the RGB LED in the Debugging Trinket M0 section of the docs.

Any suggestions?

It all seems to work while hooked up to an active Serial Monitor.

But stops after about an hour of operating on battery alone.

Was hoping the flashing LEDs would give me a clue, but if it does, I don't understand it.

[timber mango]

Why are you describing something in detail while at the same time providing no code at all?
RESCINDED (without rewriting history.) ;)

[tulip sleet]

@hard heart meaning of RGB colors: https://learn.adafruit.com/adafruit-trinket-m0-circuitpython-arduino/troubleshooting#circuitpython-rgb-status-light-18-18

yellow/purple might be value error on line 10

@timber mango please ask q's in a more positive way

[timber mango]

Understood!

[tulip sleet]

thanks!

[timber mango]

@hard heart if you would please provide your code, I may have some further insight to offer.
sorry about the 'tone of voice' in my last. ;)

[exotic pumice]

Once I get my clocks sorted we need to add rust to this article https://learn.adafruit.com/alternative-languages-for-programming-SAMD-boards

still waiting for my swd breakout to ship

[manic glacier]

tannewt

[adafruit/circuitpython] Pull request closed: #1329 Support OSError subclasses and attributes

tannewt

[circuitpython:master] 3 new commits

c5aa2e9 Support OSError attributes - notro
704d0c6 samd51: Support more uerrno errno values - notro
c47c495 Merge pull request #1329 from notro/oserror_errno - tannewt

tannewt

[adafruit/circuitpython] New comment on issue #1118: Frequency related pulseio PWMOut\.duty\_cycle does not retain set values accurately

Sorry, I meant to say the read back value may be different from the set value. The returned value should always be the same for a given frequency, set value and underlying timer. Its due to resolution loss in the underlying capture/compare value that drives the duty cycle.

[manic glacier]

tannewt

[adafruit/circuitpython] New branch created: tannewt\-patch\-5

tannewt

[circuitpython:tannewt\-patch\-5] 1 new commit

272be10 Fetch back to 4.0.0-alpha.2 so Travis has the l... - tannewt

[manic glacier]

tannewt

[circuitpython:tannewt\-patch\-5] 1 new commit

f4e7d7f Try to get the last tag. - tannewt

s-light

[adafruit/circuitpython] New comment on issue #673: DMX \(lighting protocol\)

some more references i found on my search that could be helpfull:

• Generating DMX512 - simple demo pointing to clacktronics/pyb_dmx
• claudeheintz/LXESP32DMX
• chrisstaite/TeensyDmx (active development)

tannewt

[circuitpython:tannewt\-patch\-5] 1 new commit

f13fac0 try 3 - tannewt

tannewt

[adafruit/circuitpython] Pull request opened: #1360 Make sure the commit builds include version info

Some libraries change behavior so lets make sure we have the latest tags when Travis builds.

[manic glacier]

turbinenreiter

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

• on current master c47c495
• Featherwing on top of the Feather, RST is not connected
• powered from my laptop USB - has not been a problem with the same device when using Arduino

On the REPL:

Adafruit CircuitPython 4.0.0-alpha.3-28-gc47c495ac on 2018-11-28; Adafruit Feather M4 Express with samd51j19
>>> import board
>>> import busio
>>> import wiznet
>>> import socket
>>> spi = busio.SPI(clock=board.SCK, MOSI=board.MOSI, MISO=board.MISO)
>>> eth = wiznet.WIZNET5K(spi, bo...

[manic glacier]

s-light

[adafruit/circuitpython] Fork created: s-light/circuitpython

anwar-hegazy

[adafruit/circuitpython] New star added

[manic glacier]

jerryneedell

[adafruit/circuitpython] Issue opened: #1361 byte overflow \-\- bug or feature?

I ran into this difference in behavior between CircuitPython and Python 3.6 on my Linux box.
Is one more correct than the other?

When a variable stored as a byte overlflows, on CircuiPython, it just "wraps" but on Python 3.6 it throws an error.

The CircuitPython behavior is more like "C" if the variable is declared as a uint8_t.

CitcuitPython

Adafruit CircuitPython 4.0.0-alpha.3-25-gf6c73b35f on 2018-11-27; Adafruit ItsyBitsy M4 Express with samd51g19
= bytearray({0x80,0x70...

[manic glacier]

dhalbert

[adafruit/circuitpython] Pull request closed: #1360 Make sure the commit builds include version info

dhalbert

[circuitpython:master] 4 new commits

272be10 Fetch back to 4.0.0-alpha.2 so Travis has the l... - tannewt
f4e7d7f Try to get the last tag. - tannewt
f13fac0 try 3 - tannewt
ab94344 Merge pull request #1360 from adafruit/tannewt-... - dhalbert

7wonders

[adafruit/circuitpython] New star added

[manic glacier]

jerryneedell

[adafruit/circuitpython] New comment on issue #1361: byte overflow \-\- bug or feature?

FYI - there is a simple workaround under Linux:

jerryneedell@Ubuntu-Macmini:~/circuitpython_master$ python3
Python 3.6.7 (default, Oct 22 2018, 11:32:17) 
[GCC 8.2.0] on linux
Type "help", "copyright", "credits" or "license" for more information.
>>> x = bytearray({0x80,0x70,0x60,0x50})
>>> x[0] = (x[0] <<1)&0xff
>>> x[0]
0
>>> 

[marble hornet]

I'm having trouble fining the storage, nvm, and like the main 'sterf' in cp.
Anyone mind showing me where? (attempting work on the scratch ram module)

please

[tulip sleet]

@marble hornet look in shared-bindings/<module-name> first. These define the Python-visible bindings to built-in modules. Then there are port-specific implementation routines in ports/{atmel-samd,nrf, etc.}/common-hal/<module-name>.

If a built-in module has implementation code that is port-independent, it's in shared-module/<module-name>. There may be some common code there and some port-specific code in common-hal/.

[marble hornet]

thank you @tulip sleet, i have another question 😅 . is there a naming convention for cp branches?

[tulip sleet]

you mean for a feature addition in a PR? Nope, just choose something meaningful. It doesn't matter, since once it's merged the branch will go away anyway

[marble hornet]

thnx

[tulip sleet]

if it were, say, 3.x-specific, I might prefix it with 3.x, but only to distinguish it from a similar branch that's for master

[marble hornet]

and it should go in 4?

[solar whale]

@tulip sleet just found another strange Cpython vs CircutPyhton difference -- shoudl I open an Issue or just post it here for discussion?

[tulip sleet]

I am looking up the bytearray assignment thing. It's some MicroPython decision, e.g. https://github.com/micropython/micropython/issues/3204

ga and mention it here

[solar whale]

here is a Linux python 3 example ```jerryneedell@Ubuntu-Macmini:~/circuitpython_master$ python3
Python 3.6.7 (default, Oct 22 2018, 11:32:17)
[GCC 8.2.0] on linux
Type "help", "copyright", "credits" or "license" for more information.

x = bytearray({0,1,2,3})
y = bytearray([0,1,2,3])
z = bytearray((0,1,2,3))
x
bytearray(b'\x00\x01\x02\x03')
y
bytearray(b'\x00\x01\x02\x03')
z
bytearray(b'\x00\x01\x02\x03')
x==y
True
x==
File "<stdin>", line 1
x==z
True
x = bytearray({0x00,0x01,0x02,0x03})
y = bytearray([0x00,0x01,0x02,0x03])
z = bytearray((0x00,0x01,0x02,0x03))
x
bytearray(b'\x00\x01\x02\x03')
y
bytearray(b'\x00\x01\x02\x03')
z
bytearray(b'\x00\x01\x02\x03')
x = bytearray({0x80,0x70,0x60,0x50})
y = bytearray([0x80,0x70,0x60,0x50])
y = bytearray((0x80,0x70,0x60,0x50))
z = bytearray((0x80,0x70,0x60,0x50))
x
bytearray(b'\x80Pp') y bytearray(b'\x80pP')
z
bytearray(b'\x80p`P')
x==y
False
x==z
False
y==z
True

here it is unde CircuitPython ```Press any key to enter the REPL. Use CTRL-D to reload.
Adafruit CircuitPython 4.0.0-alpha.3-25-gf6c73b35f on 2018-11-27; Adafruit ItsyBitsy M4 Express with samd51g19

x = bytearray({0,1,2,3})
y = bytearray([0,1,2,3])
z = bytearray((0,1,2,3))
x
bytearray(b'\x00\x01\x02\x03')
y
bytearray(b'\x00\x01\x02\x03')
z
bytearray(b'\x00\x01\x02\x03')
x==y
True
x==z
True
x = bytearray({0x80,0x70,0x60,0x50})
y = bytearray([0x80,0x70,0x60,0x50])
z = bytearray((0x80,0x70,0x60,0x50))
x
bytearray(b'\x80pP') y bytearray(b'\x80pP')
z
bytearray(b'\x80p`P')
x==y
True
x==z
True

note how they agree for 0,1,2,3 but not for 0x80,0x70,0x60,0x50

[marble hornet]

which comes first during intit? __init__.c or .h?

[tulip sleet]

@solar whale bytearray({x,y,z}) is iterating over a set to get the values, and the order of that may not be defined. There's a whole issue about whether insertion order is preserved for dictionaries (and probably sets).

@marble hornet not sure I understand the question. __init__.h is not executed: it's just declarations

[solar whale]

@tulip sleet thanks -- clearly the use of "{}" is not what I wanted !

[marble hornet]

okay... i think i;m just under-educated as far as internal cp. when one makes cp are .h files used? are .c files the libraries themselves ?

[tulip sleet]

the .h files are used by the compiler to provide declarations, etc. This is a C thing, not CPy.

the .c files are the executable code (though sometimes people put inline decls in .h files (we usually don't)

[lusty bough]

Is there any way I could run standard Python and circuitpython in Juypter?

[marble hornet]

okay. so ___init___.c makes the nvm module.(correct?) but the other files facilitate it's function.

[tulip sleet]

@solar whale regular dicts were officially guaranteed to be ordered dicts (insertion order, not sorting) in Python 3.7, though as a feature it was added in 3.6. Huge thread here: Guarantee ordered dict literals in v3.7?: https://mail.python.org/pipermail/python-dev/2017-November/150142.html. MicroPython drama here: https://mail.python.org/pipermail/python-dev/2017-November/150383.html

[solar whale]

@tulip sleet as to the bytearray issue -- that is not quite the same but probably would be responded to the same -- the case disussed is for a negative value but the case I ran into is an overflow. On can argue if 0xff iss == -1 but clealy 0x100 cannot be stored in a byte so I can see the reason for the ValueError.

Thanks for the link re: dicts -- interesting.

[tulip sleet]

@marble hornet the other files provide the classes in the module.

[marble hornet]

👌 thx

[tulip sleet]

the __init__.c is the module itself, yeah.

@solar whale the MICROPY_FULL_CHECKS thing mentioned in the micropython issue is only used one place, when a bytes() is initialized. MPy is not very consistent.

[marble hornet]

and cp names the module nvm b/c of the folder name? or is that declared and i just missed it?

[solar whale]

@tulip sleet At least I know how to fix the overflow issue and I should not have been using "{}" anyway -- learning ...

[tulip sleet]

@marble hornet the filenames don't name the module - it's in a declaration. The filenames could be anything: it's just a conventino. Look for the QSTR_blah_blah stuff. The .c and .h files are preprocessed to look for QSTR_something, and those mentions are converted into special strings that are kept in a dictionary ("interned" strings that are used as names): e.g. https://github.com/adafruit/circuitpython/blob/master/shared-bindings/nvm/__init__.c#L52

@solar whale yes {} is def not the right choice there 😃

[solar whale]

@tulip sleet shpuld I go ahead and clse the bytearry issue or do you want to add a note to it

[tulip sleet]

just point to the micropython issue and then close it. You could add a comment to the micropython issue as well if you want. We tend to leave core language issues to MicroPython unless they're egregious. Jeff Epler did a lot of work on finding errors in the core language stuff and we encouraged him to submit that upstream.

[manic glacier]

jerryneedell

[adafruit/circuitpython] New comment on issue #1361: byte overflow \-\- bug or feature?

Thanks @dhalbert for point out this Micropython discussion on a related issue
https://github.com/micropython/micropython/issues/3204
Closing this as it appears to be a Micropython implementation decision and is easily fixed.

jerryneedell

[adafruit/circuitpython] Issue closed: #1361 byte overflow \-\- bug or feature?

[tulip sleet]

@lusty bough you mean simultaneously?

[solar whale]

@tulip sleet Thanks -- I closed it in CP and posted it to the MP thread.

[lusty bough]

@tulip sleet yes

I was hoping to stream some data to Jupyter and process it

[tulip sleet]

@lusty bough @prime flower knows a lot about Jupyter. repeat q again. If he doesn't respond, would be good to post in the forums: https://forums.adafruit.com/viewforum.php?f=58

[lusty bough]

@prime flower Is it possible to run both Circuitpython code and normal python code in Jupyter? I was hoping to get some data from Circuit python and then process it in Jupyter

[craggy harbor]

hey guys, I need some help with busio.spi read

How can I send an address (where I want to read from) and then get the response?

Essentially, I need to make this code work with circuitpython: return self._spi.xfer([self.READ_SINGLE_BYTE | address, 0x00])[1]

[prime flower]

@lusty bough hi! kinda

@lusty bough do you want to run CP and Python in the same notebook?

[lusty bough]

Yes, ideally

If that's not possible, finding some way to stream data live would be great as well.

[prime flower]

short answer - not really, since each language is a kernel and the notebook is specific to that kernel. CP is really different since it provides more than language support (board/repl/serial)

https://vatlab.github.io/sos-docs/

SoS is a workflow system I was looking into earlier this summer when I was working with the Jupypter integration. It can provide a multi-language notebook, which is what you're looking for.

[lusty bough]

I kind of suspected that was the case. I'll look into SoS

[prime flower]

It can also facilitate data exchange between kernels, which is exactly what you want. I didn't dig as far into it since I was focused on the cpy kernel at the time, but it seemed to be what I wanted at the time.

[craggy harbor]

@fluid helm How can I send an address (where I want to read from) and then get the response?
Essentially, I need to make this code work with circuitpython: return self._spi.xfer([self.READ_SINGLE_BYTE | address, 0x00])[1]

[prime flower]

there's an active proposal for a Data Bus, which will render SoS extinct (https://github.com/jupyterlab/jupyterlab/issues/5548) but for now, SoS is exactly what you want. I haven't tested it with the CPY kernel, but if you hit a bump file an issue and I'll work with it more

@craggy harbor something like this is possible with the busdevice-https://github.com/brentru/Adafruit_CircuitPython_TinyLoRa/blob/master/adafruit_tinylora/adafruit_tinylora.py#L369

[tulip sleet]

@craggy harbor write_readinto is write and read simultaneously: https://circuitpython.readthedocs.io/en/3.x/shared-bindings/busio/SPI.html#busio.SPI.write_readinto

the "address" is just data in an SPI request, nothing special about it from the transfer point of view

what code are you trying to translate?

[hard heart]

@timber mango Happy to provide code. Not near my rig now, so it will be later today. Also, will attach a video of the lights blinking. Didn't realize that would be helpful.

[craggy harbor]

@tulip sleet I tried that already, but got "None" as an answer :/ Probably an error somewhere else..

[manic glacier]

caternuson

[adafruit/circuitpython] New comment on issue #1355: create driver for TMP006

Done.
https://github.com/adafruit/Adafruit_CircuitPython_TMP006

caternuson

[adafruit/circuitpython] Issue closed: #1355 create driver for TMP006

[craggy harbor]

I am "porting" this: https://github.com/nahueldsanchez/pycc1101

[tulip sleet]

you mean you got None in buffer_out? the function doesn't return the data: it writes it into buffer_out

@prime flower's example is good

[craggy harbor]

I'm looking at the example now 😃

[prime flower]

@tulip sleet I'm running against a pylint error with methods which get flaggedno-self-use, so I've removed the self in the functions but pylint tosses more errors

i.e: https://github.com/brentru/Adafruit_CircuitPython_TinyLoRa/blob/master/adafruit_tinylora/adafruit_tinylora_encryption.py#L204. I understand why Pylint throws it, I think. It's because it never accesses self.X

[craggy harbor]

Nice, it's working now!

I'm not using the presented code directly, but it gave me the hint why it did not work. I now use this:
with self._device:
self._spi.write(databuffer, end=1)
self._spi.readinto(databuffer, end=2)

[prime flower]

whoo!

[tulip sleet]

@prime flower you could make them be @staticmethod and take away the self arg

[craggy harbor]

😄

[main meteor]

@pastel panther Mentioned wanting to see a SWD implementation on an Itsy M4. Looks like you can do SWD with a plain Arduino: https://learn.adafruit.com/programming-an-m0-using-an-arduino

[manic glacier]

DuckRogers

[adafruit/circuitpython] New comment on issue #1118: Frequency related pulseio PWMOut\.duty\_cycle does not retain set values accurately

Okie Dokie

On Tue, Nov 27, 2018 at 11:49 PM Scott Shawcroft [e-mail]
wrote:

Sorry, I meant to say the read back value may be different from the set
value. The returned value should always be the same for a given frequency,
set value and underlying timer. Its due to resolution loss in the
underlying capture/compare value that drives the duty cycle.

—
You are receiving this because you authored the thread.
Reply to this email directly, view it on GitHub
<ht...

ladyada

[adafruit/circuitpython] New comment on issue #1355: create driver for TMP006

niiiiice

[prime flower]

@tulip sleet @staticmethod worked, didn't know about those, thanks!!

[slender iron]

@main meteor I'd love to see it from CircuitPython too. I have some samd09 boards I could use it for

[manic glacier]

dhalbert

[adafruit/circuitpython] branch deleted: tannewt\-patch\-5

[craggy harbor]

is zfill not included in the python standard library?

circuitpython doesn't seem to like it

🤔

[pastel panther]

@main meteor looks like I owe Alex Taradov yet another beer

I'll probably look into porting that when I have some spare cycles

[idle owl]

@tidal kiln TMP006 is the same thing as 007 right?

hardware-concept-wise

[tidal kiln]

same in what they do, but different products - the TMP007 is the newer TMP006

[idle owl]

the what they do was what I was wondering about. Thanks

[tidal kiln]

yep - this is why i was asking you about renaming a repo, whenever days ago

[idle owl]

was able to copy the setup.py from 007 and change it to 006. easy peasy.

Right.

[tidal kiln]

the TMP006 and TMP007 drivers could both live in the same place

[idle owl]

I thought I knew, but wanted to make sure.

Right but that would involve breaking the repo and a lot of other things.

So it is what it is.

[tidal kiln]

yep. fine by me. or if you can think of a better solution, go for it.

[idle owl]

Only option is a new repo and then update everything from the new repo.

which we talked about but decided not to do.

eh. it's fine how it is. if it becomes a nightmare to support them separately, we'll combine them.

[tidal kiln]

agree

[idle owl]

Pro-tip: sometimes something you think is in your bash history isn't, and you add the wrong remote to a repo by running the last remote add you had in your history.... 🤦

Didn't notice until I pushed my branch to the entirely wrong repo.

[craggy harbor]

AttributeError: 'str' object has no attribute 'zfill'

🤔

[tidal kiln]

probably not supported in CP

[craggy harbor]

hmm ok weird. it's a standard library function.