@tulip sleet Yeah I was thinking the same thing but wondering if we should try to have it done for this release or not.
#circuitpython-dev
tulip sleet
do you have more stuff you want to add, like IR?
it might be a q for Scott and Limor
idle owl
IR is going to be a serious learning curve for me, and it's lower priority. The only other thing to consider was the sound meter.
That's it though.
That would be every feature having something in express which I guess is what I pictured.
I thought there was an outstanding issue but it's for lis3dh not express
tulip sleet
I would ask them as well as me. I wasn't home for JP's mem problems; is that resolved enough not to have it frozen?
idle owl
Not sure, he's out for a bit, but he'll be back. And Scott rewrote it to not fail on memory error, so it seems to mostly work even with constant memory errors.
Of course I'd ask them as well, but you deal most with the freezing bit, so I started with you.
tulip sleet
if we can get away without freezing it, then it can be deferred. maybe 2.2 is a little early
idle owl
Yeah I'm still up in the air about it myself. It would be nice to avoid memory issues but there are still changes being made.
tulip sleet
@idle owl I did some copy-editing on the first couple of pages of that guide. Just typos and font styles. And I made the warning about not resetting or unplugging more obvious.
tulip sleet
The screen section needs to say how you exit screen: ctrl-A \ on Linux, ctrl-A ctrl-\ on Mac (right?)
idle owl
Oh I type :exit
But I think you're right there too
not working for me but I think I ran into this last time where it didn't like me then either.
I know crtl+A, :exit works for me.
tulip sleet
ctrl-A :quit on linux. Oy, silly differences.
This is a really great guide. I'm looking forward to pointing people to it.
idle owl
Thank you! I'm excited about it
timber lion
there are two versions of screen actually, a gnu one and another one
BSD i think
each has a different exit sequence
it's why mac is different from linux
and very annoying
umbral dagger
I’ve been using C-a k y on Linux
opaque patrol
Does anyone have a link to the instructions to load CircuitPython on a feather m0 adalogger using bossa?
tidal kiln
@opaque patrol i think it's pretty much just:
bossac -e -w -v -R FIRMWARE.bin
and replace FIRMWARE with appropriate file name
idle owl
Ok I've been going in circles for an hour. I'm trying to get this OLED SPI breakout working with CP and after working through a number of errors, I've come repeatedly to one that I can't seem to get around.
idle owl
It's the SSD1306 1.3"
lofty topaz
Ah, you can hook that up a number of ways I see. You choose I2C. ok. So what kind of errors?
Are you level shifting?
idle owl
SPI. And I'm getting 'SPIDevice' object has no attribute 'write'. The driver for this device uses another library that uses another library. There's a level shifter built in.
lofty topaz
ok, code issue. I see.
so who says there should be a write attribute?
haven't looked at any code yet.
idle owl
This library calls adafruit_bus_device.SPIDevice I believe. I can find write in all three libraries
lofty topaz
try - void fastSPIwrite(uint8_t c);
idle owl
That isn't CircuitPython....
tulip sleet
it should be writeinto, maybe
idle owl
Counting down the days until I upgrade my laptop. Oi... crashed again.
tulip sleet
are you testing the library or just trying to use the display? The I2C impl might work better.
tulip sleet
there's no writeinto never mind
idle owl
Error: Traceback (most recent call last): File "code.py", line 11, in <module> File "adafruit_ssd1306.py", line 183, in __init__ File "adafruit_ssd1306.py", line 46, in __init__ File "adafruit_ssd1306.py", line 72, in init_display File "adafruit_ssd1306.py", line 189, in write_cmd File "adafruit_ssd1306.py", line 189, in write_cmd AttributeError: 'SPIDevice' object has no attribute 'write'
I2C is probably easier, yeah, but this lib is waiting on SPI testing.
lofty topaz
I think I see a "writeto(...)"
idle owl
This is SPIDevice: https://github.com/adafruit/Adafruit_CircuitPython_BusDevice/blob/master/adafruit_bus_device/spi_device.py
lofty topaz
Right, I2C is part of busio which is imported. have you tried 'writeto'?
idle owl
In what? The device library? It's also in the other 2 libraries and I don't have those locally
tulip sleet
I think SPIDevice might be just wrong. It's missing any actual read or write methods of any name.
idle owl
Oi! Vindication!
lofty topaz
i dunno. just lookin at this.
http://circuitpython.readthedocs.io/en/latest/shared-bindings/busio/I2C.html#busio.I2C.writeto
lofty topaz
oh heck, I was in the example. so sorry
tulip sleet
@idle owl did you test lis3dh with spi or just i2c?
idle owl
@tulip sleet I guess just i2c. I didn't think to check both. Or even consider that both were an option. I went with the whatever the example would have been, and then tested shake.
lofty topaz
My bad, found something in bitbangio tho
idle owl
RTD has write for SPI.
But that doesn't mean it's in the lib.
Although Sphinx would have had a fit if it caught it, so I'm not sure this makes sense.
tulip sleet
ok, I see what's going on. __enter__ (invoked by with) returns the busio.SPI object, which you can read and write on.
idle owl
Ok
idle owl
I tried setting it up like that but because some of it's handled in the device lib, it doesn't seem to work right. Maybe it's my code.
import board
import busio
import digitalio
import adafruit_ssd1306
spi = busio.SPI(board.SCK, MOSI=board.MOSI, MISO=board.MISO)
dc = digitalio.DigitalInOut(board.D9)
res = digitalio.DigitalInOut(board.D10)
cs = digitalio.DigitalInOut(board.D11)
oled = adafruit_ssd1306.SSD1306_SPI(128, 64, spi, dc, res, cs)
oled.fill(1)
oled.show()```There's nothing to it though.
Fails "write" on the first oled line
idle owl
It's not getting there though.
It's failing on the oled = adafruit_etc line according to traceback.
I tried that though, no change in error
tulip sleet
I will stare at the code a minute...
lofty topaz
doesn't this SPI need to be locked first?
tulip sleet
def write_cmd(self, cmd):
self.dc.value = 0
with self.spi_device:
self.spi_device.write(bytearray([cmd]))
def write_framebuf(self):
self.dc.value = 1
with self.spi_device:
self.spi.write(self.buffer)
Yeah
tulip sleet
change that code to:
def write_cmd(self, cmd):
self.dc.value = 0
with self.spi_device as spi:
spi.write(bytearray([cmd]))
def write_framebuf(self):
self.dc.value = 1
with self.spi_device as spi:
spi.write(self.buffer)
Oh I need to download the driver then.
tulip sleet
@lofty topaz the locking is done by the SPIDevice when you do with. with calls __enter__ invisibly.
idle owl
Oh I have it from downloading it for the M4
Oh no that's in the one I've been working on. Oi. ok.
tulip sleet
I had never looked at this in detail before. The I2CDevice and SPIDevice work quite differently. I2CDevice returns self; SPIDevice returns the busio.SPI object. Hmmm.... ALso the example code in SPIDevice is wrong, I think.
idle owl
OK.... ok. It's no longer giving me that error
Instead I have AttributeError: 'SSD1306_SPI' object has no attribute 'dc' which I thought it was a parameter?
So I defined it in my code, but if I just put the pin number in the param it gives issues with Pin object, and if I replace dc with digitalio.DigitalInOut(board.D9) I get P07 or something in use
tulip sleet
its a pin, not a pin number. I think it means direction
tulip sleet
oh, i see what you mean
tulip sleet
are you testing this on a feather?
idle owl
Yeah
tulip sleet
so paste your current test code
idle owl
import board
import busio
import digitalio
import adafruit_ssd1306
spi = busio.SPI(board.SCK, MOSI=board.MOSI, MISO=board.MISO)
dc = digitalio.DigitalInOut(board.D9)
res = digitalio.DigitalInOut(board.D10)
cs = digitalio.DigitalInOut(board.D11)
oled = adafruit_ssd1306.SSD1306_SPI(128, 64, spi, dc, res, cs)
with oled as oledspi:
oled.fill(1)
oled.show()```wait... or is it failing in the changes you made?
File "code.py", line 12, in <module>
File "adafruit_ssd1306.py", line 183, in __init__
File "adafruit_ssd1306.py", line 46, in __init__
File "adafruit_ssd1306.py", line 72, in init_display
File "adafruit_ssd1306.py", line 186, in write_cmd
AttributeError: 'SSD1306_SPI' object has no attribute 'dc'```error ^^
tulip sleet
look in the edited version of SPIDevice.py. What does write_cmd look like? paste it here
idle owl
I replaced it with your code
def write_cmd(self, cmd):
"""Send a command to the SPI device"""
self.d_or_c.value = 0
with self.spi_device:
self.spi_device.write(bytearray([cmd]))
def write_framebuf(self):
"""write to the frame buffer via SPI"""
self.d_or_c.value = 1
with self.spi_device:
self.spi_device.write(self.buffer)```that's what it was
tulip sleet
I don't see the self.d_or_c.. I'm not looking in the PR. Hold one.
tulip sleet
replace self.dc.value in my replacements with self.d_or_c.value. I was looking at the old code, sorry.
in both def's
idle owl
Ooh new error!
File "code.py", line 14, in <module>
AttributeError: 'SSD1306_SPI' object has no attribute '__exit__'```I got it!!
I took out the with you had me add to test earlier!
You figured it out!
Aahhhah! Yes!!!
tulip sleet
it displays!?
idle owl
Yes!!!
tulip sleet
OK! Well, that was painful. I didn't really know these libraries. I'm a little weirded out by the non-consistency between the I2C and SPI versions.
idle owl
I was going in circles with that! Thank you so much!!
And that's why we test code, kids.
tulip sleet
I need to file an issue about the bad example. YW!
idle owl
Ok, I need to edit the SPI code in the lib
lofty topaz
...and I need to mind my own business. Didn't mean to get in the way.
idle owl
No worries. You never know when you can help. Thank you for trying. I appreciate it!
lofty topaz
I shouldn't bounce between channels so much and get confused.
tulip sleet
@lofty topaz don't worry - this was really confusing - there was a bug in an example and there was a bug in the library under test. Neither @idle owl nor I understood the corect use of the SPIBusDevice lib up front
lofty topaz
@tulip sleet Yeah, plus I had just returned from celebrating my daughters 22nd birthday. Helpful frame of mind but not so analytical. 🍸
Nice hack tho.... ><>
sacred socket
I just installed CircuitPython on a Huzzah esp8266. I've connecte to it using Putty. I followed the Adafruit tutorial by Tony DiCola. It is not showing up as a mass storage device. Are there additions steps to install the UF2 files/ability to the Huzzah?
idle owl
The ESP chips don't support UF2
sacred socket
Thanks. I was thinking something like that would be the answer.
idle owl
It's something with the hardware not supporting USB like that. Can't be resolved with software.
sacred socket
What would be re recommended method to save code to the Huzzah? FIYI I haven't tried or researched it yet.
FYI...
idle owl
ampy maybe? I'm not familiar enough with working with the ESP chips. There's a few other people who have done it a bunch.
sacred socket
I'll look into that. Thanks again.
raven canopy
@sacred socket This should help you out.
https://learn.adafruit.com/building-and-running-micropython-on-the-esp8266/overview
sacred socket
I'd found that page but have not read through it yet. Reading https://learn.adafruit.com/micropython-basics-load-files-and-run-code/boot-scripts right now.
ruby lake
for some reason the macbook thinks there isn't enough space on FEATHERBOOT to install a new uf2
slender iron
@ruby lake try renaming it to current.uf2 to cause it to replace it
stuck elbow
hmm, d_or_c is a really confusing name
the pin is labeled "dc" on all the modules
(yes, I know technically it's d/c)
tulip sleet
I had to look up that it’s “data/command “
stuck elbow
and sometimes it's also marked as A0
or some equally useless marking
I think that if we want to follow Python's style of using whole words, it should be data_or_command
what I'm trying to say is that d_or_c is not really better than dc in terms of clarity
then it would also make sense to use chip_select for cs too
and perhaps also "master_in_slave_out" for miso
raven canopy
data_cmd seems a happy medium...
stuck elbow
only if you are a native speaker
tulip sleet
miso et al seem like standard terms, cs is kinda that, but I agree data_or_command or data_or_cmd would be a lot clearer, since it's certainly not universal on SPI devices. Happy to entertain a PR.
stuck elbow
dc is quite standard in the niche of serially-controlled displays
there is also the question of being consistent with the drivers for other platforms, like arduino etc.
they all use dc
@idle owl where is that code from? I'm looking at https://github.com/adafruit/micropython-adafruit-ssd1306/blob/master/ssd1306.py but can't see any d_or_c?
nevermind, found it, wrong python
I wonder if there is some way to subscribe to changes from all the circuitpython-related repositories somehow
timber mango
Hello!
Ive just received and plugged in my trinket M0 & Metro express boards to try CircuitPython...
I'm under windows10, I've installed the adafruid drivers 2.0.0, and the trinket M0 is seen as a drive when plugged in.
But when I'm plugging in the Metro Express, It is not seen as a drive, even though it does seem to run the demo python file, since the RGB led is showing some kind of rainbow colors...
I've managed to connect Putty on the newly created COM8, and some numbers are scrolling... but I can't stop the script execution with CTRL-C as for the trinket M0 demo script... :s
I'm looking for any help or suggestion to be able to see the Metro express as a drive and be able to modify the CircuitPython script... thanks! 😃
ok, just found that: "Windows 10
Did you install the Adafruit Windows Drivers package by mistake? You don't need to install this package on Windows 10 for most Adafruit boards. The old version (v1.5) can interfere with recognizing your device. Go to Settings -> Apps and uninstall all the "Adafruit" driver programs."
my mistake... 😦
raven canopy
question: are #elif & #else recognized in the CPy backend? Hoping to use it like:
#ifdef SAMD21
foo = bar;
#elif SAMD51
foo = bar2;
#else
foo = no_bar;
#endif ```@raven canopy you mean in Python code? Nope, though you could run the .py through cpp yourself. I'm not sure I've seen anyone do this in Python code.
@timber mango are you all set now?
raven canopy
@tulip sleet not in Python. In C...working on the unique ID thing.
tulip sleet
In C, yes, we use it ALL the time. But there are also board-specific files where you might want to isolate these routines. Are you doing this on the master branch?
raven canopy
yeah. I thought about using each ports common_hal, but since we're targeting both SAMD & NRF (and ESP uses machine), I'm starting in shared-bindings. if that is the wrong approach, I can go back to the drawing board and figure out how to impl in the ports/boards.
and just thought about recursion to MPy... that's a reason to go the ports/boards route.
tulip sleet
shared-bindings is common code; we don't conditionalize by arch there. So use common-hal, but with identical function names across the ports. See atmel-samd/samd21_peripherals.c and samd51_peripherals.c for an example of the same function name with different impls. These files are compiled as appropriate for each arch. The routines in them are called from a common-hal/ routine.
raven canopy
copy that. thanks for the vector correction! 😄
idle owl
@tulip sleet I just posted a travis-fixing commit to ssd1306, so can you give it a final look and merge it?
tulip sleet
@is it worth testing the SPI changes quickly? I have an SSD1306 but I'd have to unsolder the I2C jumpers. Is yours still set up for SPI?
timber mango
@idle owl @tulip sleet heya Q, for SPI displays, y'all renamed dc to d_or_c?
timber mango
lol ok whats the name now?
idle owl
Or whoever linted it had anyway
tulip sleet
@meager fog mr mcwethy did, but @stuck elbow thought that was obscure, and submitted a PR to make it clearer. See https://github.com/adafruit/Adafruit_CircuitPython_SSD1306/pull/4
idle owl
dc_pin - but this change isn't committed yet. So it's d_or_c right now I believe.
tulip sleet
and @stuck elbow documented what "dc" means in a docstring comment (per my request)
timber mango
ooh ok yah dc_pin is good
thx, just checking - d_or_c is not /technically/ correct
dc_pin is best! 😃
tulip sleet
yeah, it bothered him; great - we want clearer libraries, and he added a lot of doc as well
timber mango
np some of these pin names are unclear but i can help since i label the pins on th eboard - just @ me
i have a few minutes to kill, should i put 2.2 on a CPX and test doubletap?
idle owl
@timber mango Yes please!
@tulip sleet Tested successfully.
I'll merge it then
New release done as well, so that's sorted.
timber mango
i do want to take a look at framebuf and perhaps change it to be more gfx-like at some point
maybe its very close to done :/
idle owl
@timber mango Did you get my emails earlier this week about the guide? It's basically done except for those two questions.
Yay cpx.double_tap works for me with 2.2 and the most recent bundle!
manic glacierBOT
lucid arch
Hi all, I've literally started exploring and reading about circuit playground express just now and was going through the adafruit learning page. There's an example to try but I'm getting this error:
No I2C device at address: 40
Anybody have any idea what's the problem? Ta
idle owl
Yes, in fact, I believe I do. Can you link the code you're looking at?
lucid arch
It's here: https://learn.adafruit.com/adafruit-circuit-playground-express/about-circuitpython-libraries
This is the code:
import adafruit_si7021
import busio
import board
i2c = busio.I2C(board.SCL, board.SDA)
sensor = adafruit_si7021.SI7021(i2c)
print("Temperature:", sensor.temperature)
print("Humidity:", sensor.relative_humidity)
idle owl
I think the issue is in the line where you setup the i2c instance of whatever you're using. So instead of just (i2c), for the Circuit Playground Express you need (i2c, address=0x19)
lucid arch
It was just demonstrating how to and why to install latest libraries and etc
lucid arch
ah. I see
idle owl
For that to work overall, you need that sensor attached, so while it's walking you through how to install libraries, it won't fully work on your board unless you also have that sensor.
opaque patrol
@lucid arch I did the exact same thing when I started with that example, maybe it should be changed to indicate it is demonstrating an error more clearly
idle owl
@opaque patrol You are correct, and there is currently a guide that's going to replace a lot of those generalised getting started pages.
It's not quite finished but that's the plan.
lucid arch
Ok, that makes more sense now. Is there an onboard address i can use, since 19 is not found either
idle owl
That should have worked. I'm wondering if the issue is that the example won't work without the sensor breakout board.
opaque patrol
@idle owl speaking of guides....I talked to Scott earlier in the week and he said to get with you about helping with guides....
idle owl
So it's looking for something on that address and it's not finding anything because there's nothing attached.
@opaque patrol Oh nice! Yeah, I can get that sorted.
lucid arch
yeah. I'm assuming you dont use the i2c to read the temperature sensor?
idle owl
Not the onboard temp sensor, no
I don't think anyway.
The examples further on in that guide will cover a lot of it.
That example is simply showing how to create an error by not having a module installed, and then installing the library you need to resolve the error.
lucid arch
ok
got it
makes sense.
cool, i'm moving on with the next pages then
just one last question. I should just override the code.py everytime with my code and save and that should be it, for uploading to the board?
idle owl
Thank you for the feedback! I'm adding a couple of sentences to the upcoming guide to make it clearer what's going on. The example is different in the new guide, but still we want it to be as clear as possible.
Yeah that is a way to do it, or you can edit it every time. Whichever way you want to do it. Just be sure to save it somewhere if you want to keep it.
lucid arch
cool .Thanks, you've been really helpful. this forum is 😃
idle owl
That's great to hear! That's one of the biggest reasons we're here!
lucid arch
i've been exploring the world of flora lately before i came across the CPX and so i'm more used to the arduino ide. I quickly went through a few examples there and was having a hard time getting the sensors working.
i tried the colour example, where the colour sensor senses colour and flashes the neopixels of that colour but that did not work as it should
similarly same result for the light sensor
are those examples massively out of date? or..
lucid arch
circuitpython examples via the arduino ide
idle owl
I have no idea then. I've never worked with Arduino, so I haven't gone through them, nor would I be able to tell by looking.
lucid arch
ok. no worries. thanks, i guess i'll keep going through the circuit python examples for now and see where they lead.
idle owl
Ok, sounds like a plan!
Someone else would know better than me, I'm sure. So hopefully someone catches your question.
lucid arch
cool. ta
opaque patrol
I am looking at the color_sense example now....It basically uses the buttons to flash colors and detect the amount of light reflected...I might try to convert this to CP
lucid arch
thanks BravoDelta. I basically kept getting medium values, even though i put the board next to stark green felt paper or red paper.
idle owl
@opaque patrol Yeah it's not really a color sensor, the examples in MakeCode use the LED next to it to flash and then you use that data. Sounds similar. I know there was a plan to add that in CircuitPython, but as far as I know, no one has done anything with it.
opaque patrol
@idle owl Do you know off the top of your head what the range (or resolution) of the light sensor on the cpx?
idle owl
Nope.
opaque patrol
here is what I have so far to sense color....
from adafruit_circuitplayground.express import cpx
import time
from simpleio import map_range
LIGHT_SETTLE_MS = 0.1 # 100 ms
RED = (255, 0, 0)
GREEN = (0, 255, 0)
BLUE = (0, 0, 255)
CLEAR = (0, 0, 0)
def sense_color(verbose_out = False):
cpx.pixels.fill(CLEAR) # clear pixels so they don't interfere
time.sleep(.5)
# Save the current pixel brightness so it can later be restored. Then bump
# the brightness to max to make sure the LED is as bright as possible for
# the color readings.
old_brightness = cpx.pixels.brightness
cpx.pixels.brightness = 1.0
# Set pixel 1 (next to the light sensor) to full red, green, blue
# color and grab a light sensor reading. Make sure to wait a bit
# after changing pixel colors to let the light sensor change
# resistance!
cpx.pixels[1] = RED
time.sleep(LIGHT_SETTLE_MS)
raw_red = cpx.light
cpx.pixels[1] = CLEAR
time.sleep(.01)
if verbose_out:
print("raw red = {}".format(raw_red))
cpx.pixels[1] = GREEN
time.sleep(LIGHT_SETTLE_MS)
raw_green = cpx.light
cpx.pixels[1] = CLEAR
time.sleep(.01)
if verbose_out:
print("raw green = {}".format(raw_green))
cpx.pixels[1] = BLUE
time.sleep(LIGHT_SETTLE_MS)
raw_blue = cpx.light
cpx.pixels[1] = CLEAR
time.sleep(.01)
if verbose_out:
print("raw blue = {}".format(raw_blue))
# Turn off the pixel and restore brightness, we're done with readings.
cpx.pixels.brightness = old_brightness
# Now scale down each of the raw readings to be within 0 to 255.
red = map_range(raw_red, 0, 255, 0, 330)
green = map_range(raw_green, 0, 255, 0, 330)
blue = map_range(raw_blue, 0, 255, 0, 330)
return int(red), int(green), int(blue)
try:
cpx.pixels.brightness = .2
verbose = False
while True:
if cpx.button_a or cpx.button_b:
verbose = cpx.button_b
time.sleep(.05) # debounce
if verbose:
print("button pressed")
while cpx.button_a or cpx.button_b:
time.sleep(.01)
r, g, b = sense_color(verbose)
print("({}, {}, {})".format(r, g, b))
cpx.pixels.fill((r, g, b))
except:
cpx.pixels.fill(CLEAR)
actually works pretty good
idle owl
Nice!!
lucid arch
i tried your code and i get a memory error:
Traceback (most recent call last):
File "code.py", line 1, in <module>
MemoryError:
idle owl
So we're super tight on memory with these boards. If you have too much on your board already, you can run into that with beefy code.
Do you have the most updated version of CircuitPython on your board?
lucid arch
yep
idle owl
And the updated libraries?
idle owl
Yep
idle owl
Ok
Not sure then. It's early working code, so we'll have to test it more to see what's up
lucid arch
ah i tried the keyboard exmaple just now and i get another memory error:
Traceback (most recent call last):
File "code.py", line 7, in <module>
File "/lib/adafruit_hid/keyboard.py", line 34, in <module>
MemoryError: memory allocation failed, allocating 896 bytes
idle owl
Hmm. Did you save a bunch of files to the board or anything? Or you're still working with just one code.py
Also if you put three backticks (the one in the top left corner of your keyboard) on both sides of your code, it will make it a codeblockOne backtick on either side and you get inline code
idle owl
So when you're posting code or serial errors, it's good to use the backticks for readability.
idle owl
Hmm.
lucid arch
backticks : got it.
half sedge
@idle owl When you say the lattest, are you talking about 2.2.rc1 or is there now a definitive 2.2.0?
opaque patrol
cpx had an issue for a time with lis3dh causing memory errors
lucid arch
i got this: adafruit-circuitpython-bundle-py-20171216.zip
idle owl
I actually meant 2.1.0 since that's the latest stable. But, to answer the question, no it's still rc1.
opaque patrol
rc1
idle owl
@opaque patrol Not any more.
It prefers frozen modules over local copies now to eliminate that.
opaque patrol
I can change it to not use cpx
idle owl
@lucid arch That's good, the most updated libraries are best to have. There's a couple of new features that need a newer version of CircuitPython, but they're not in any examples yet.
@opaque patrol That shouldn't be it. But you can give it a try.
@lucid arch is getting memory errors on other things too.
lucid arch
yeah, keyboard module.
opaque patrol
@lucid arch Were you using the bundle from 20171216 when you got the memory error?
lucid arch
am defo still working on code.py and with the same lib folder. so it cant be a space issue.
i've had the same bundle since an hour, which is when first installed my cpx latest.
idle owl
@opaque patrol The cpx class is designed to use the frozen modules now instead of the local copies which was what was causing the memory errors with lis3dh.
lucid arch
havnt changed the lib.
idle owl
Have you physically reset the board?
I'm not sure it will help but it's an easy step to try. Friendly eject it and then press the reset button.
Then reconnect to serial and see what you get
lucid arch
pressed the reset button? yeah, a bunch of times.
idle owl
Oh ok
lucid arch
ok, let me try that.
idle owl
If you've already reset, then I doubt it will help.
opaque patrol
I am actually using the lib bundle from the 12th. cpx is version 0.9.1 according to versions.txt
idle owl
We might need to fully erase the board and start again. Something got unhappy.
That's the newest, @opaque patrol
iirc.
Let me check.
opaque patrol
Could try flashing 2.2.0 rc1, it is super easy to flash the circuit playground express
lucid arch
still the same issue. keyboard mem error 😦
idle owl
Oh there's a 1.0 release, but it only adds another thing to the API, no bug fixes.
lucid arch
@opaque patrol : flash the bundle? how.
idle owl
Ok, you seem pretty comfortable with all of this, I want you to follow the steps here: https://learn.adafruit.com/adafruit-circuit-playground-express/troubleshooting#for-the-circuit-playground-express-feather-m0-express-and-metro-m0-express
lucid arch
coool
idle owl
We're going to erase the board completely, so back up everything you want
idle owl
Then I'm going to link you the release candidate of CircuitPython
Download this and flash it instead of 2.1 (this shouldn't matter, but we would love to have some more testing of this version)
opaque patrol
Testing it now 😃
idle owl
Thanks!
lucid arch
they have different sizes
idle owl
Should say 2.x
but we created the lib bundle stuff before we released 2.2rc1
Here:
That. If you didn't already get that same one.
lucid arch
works
😃
the keyboard example works
@opaque patrol your code about the colour sensor works as well 😃
idle owl
Yay!
lucid arch
so the board did get unhappy ... hmm, a bit temperamental, isnt it?
idle owl
It can happen. They're not always tempermental, but when they are, sometimes it takes a full flash to fix it.
Could have been a memory leak in something and then whatever it was couldn't clear, something like that.
The other thing is you don't want to reset without ejecting
and if you do, files can become corrupted if anything was trying to write at the time
often this corrupts the entire board
but it can affect one file, and then that file can just cause memory issues.
So that also could have been it. Never know. For what it's worth, though, it seems really difficult to actually break anything. Even if it's not showing up at all, as long as you can get to the bootloader, you can reflash it with the steps I sent, and get everything happy again.
lucid arch
cool
pretty cool
ok. its been joy talking to you guys
am off . will continue exploring the cpx more tomorrow.
is the cpx meant to be wearable, like the flora is?
or is it just a tiny board for exploring lights and sensors?
i love that theres so much in the cp as compared to the flora. it means you can start doing small stuff and seeing things really quick, like a rad sort of thing.
idle owl
It can work as a wearable! There's a bunch of projects where it is.
But it's specifically meant to be a perfect intro to programming and electronics.
So many options to learn different concepts.
idle owl
😃
lucid arch
well, thanks and good night.
idle owl
You're welcome and you have a lovely night as well!
idle owl
Blergh, travis failed on that PR I just did, except I didn't change anything but the one line in travis.yaml so I'm a bit confused.
opaque patrol
travis is just upset that you took them out of it
idle owl
Apparently.
But it means it wasn't linted completely to begin with or they literally just added this check. Which on research says they did that in September.
This:python def _read_register(self, reg, count=1): self.buf[0] = _COMMAND_BIT | reg if count == 2: self.buf[0] |= _WORD_BIT with self.i2c_device as i2c: i2c.write(self.buf, end=1, stop=False) i2c.readinto(self.buf, start=1) if count == 1: return self.buf[1] elif count == 2: return (self.buf[1], self.buf[2])Says:Either all return statements in a function should return an expression, or none of them should. (inconsistent-return-statements)
I don't understand what it means.
Found this:
```A new Python checker was added to warn about inconsistent-return-statements. A function or a method
- has inconsistent return statements if it returns both explicit and implicit values :
- .. code-block:: python
- def mix_implicit_explicit_returns(arg):
-
if arg < 10: -
return True -
elif arg < 20: -
return - According to PEP8_, if any return statement returns an expression,
- any return statements where no value is returned should explicitly state this as return None,
- and an explicit return statement should be present at the end of the function (if reachable).
- Thus, the previous function should be written:
- .. code-block:: python
- def mix_implicit_explicit_returns(arg):
-
if arg < 10: -
return True -
elif arg < 20: -
return None```
But I don't see how that's what's happening, so...
I'm lost to fix it.
opaque patrol
It thinks you are returning an expression in the second because of the parens
opaque patrol
What does the calling class expect? The second is returning a tuple (right?)
idle owl
I think so... and I have no idea. I didn't do anything else with this.
opaque patrol
This is beyond my knowledge of python...I think you could put parens on the first return but I am not sure what any side effects might be by doing that
tulip sleet
Suppose count == 3, then there's no return statement that's operative
idle owl
I don't understand
Like I said, all I did was update that same travis.yaml issue
And then the build failed on the actual driver file.
tulip sleet
if count == 1:
return self.buf[1]
elif count == 2:
return (self.buf[1], self.buf[2])
suppose count is not 1 or 2, if there's no further statements, then the code just falls off the end. I don't know why pylint didn't catch this before with travis; that's a different q
but i think that's what's wrong with the code
opaque patrol
Good catch,
tulip sleet
bye - sorry - on our way out
idle owl
Later!
Hmm ok... I'm going to close the PR and delete the branch, then clone it locally and try to deal with it because I don't know how to do multiple changes on a GitHub created branch with an active PR in place.
opaque patrol
I do believe you could change the first to return (self.buf[1],) but I might be wrong
idle owl
PyCharm disagrees, whatever linter it's using says those are redundant.
Wants me to remove the parens from the last one instead. But I don't think that's the problem.
Need to clone it, brb
idle owl
Any idea how that's supposed to look?
What?.... pylint locally doesn't catch it.
Ok now I'm super confused.
opaque patrol
It has if count == 1 elif count == 2, what if count < 1 or > 2? Probably unlikely but the linter just sees that you have an if, elif and no else.
can't type tonight....what *if count.... keep forgetting there is an edit mode
opaque patrol
the example has no else either so I am not sure
idle owl
yeah. hmm.
I don't think this is something I can do, heh.
So it gets to wait.
And that's weird that travis ci fails, but pylint passes. They're usually in unison.
raven canopy
you could try setting a variable in the if..., then return the variable at the end.
def _read_register(self, reg, count=1):
self.buf[0] = _COMMAND_BIT | reg
if count == 2:
self.buf[0] |= _WORD_BIT
with self.i2c_device as i2c:
i2c.write(self.buf, end=1, stop=False)
i2c.readinto(self.buf, start=1)
if count == 1:
buff_read = self.buf[1]
elif count == 2:
buff_read = (self.buf[1], self.buf[2])
return buff_read
might have to eval buff_read before the return...
opaque patrol
or just use return None
raven canopy
yeah
if not buff_read:
return None
else:
return buff_read
note: structure illustration...i'm losing all my syntax from all the lang switching I'm doing lately. 😄
idle owl
I'm not even sure where that would go.
raven canopy
at the end, when returning the buffer read.
opaque patrol
def _read_register(self, reg, count=1):
self.buf[0] = _COMMAND_BIT | reg
if count == 2:
self.buf[0] |= _WORD_BIT
with self.i2c_device as i2c:
i2c.write(self.buf, end=1, stop=False)
i2c.readinto(self.buf, start=1)
if count == 1:
return self.buf[1]
elif count == 2:
return (self.buf[1], self.buf[2])
return None # this could actually be un-indented one as well
Won't let me edit twice
def _read_register(self, reg, count=1):
self.buf[0] = _COMMAND_BIT | reg
if count == 2:
self.buf[0] |= _WORD_BIT
with self.i2c_device as i2c:
i2c.write(self.buf, end=1, stop=False)
i2c.readinto(self.buf, start=1)
if count == 1:
return self.buf[1]
elif count == 2:
return (self.buf[1], self.buf[2])
else:
return None
You get the idea....
idle owl
R:183, 8: Unnecessary "else" after "return" (no-else-return)
Is the new error.
Ooh it works the first way
Let's commit and push and see how travis ci likes it! pylint is happy.
I forked it so I can break it all I want 😄
Hmm should test it first. I'll need to get that sensor going first.
opaque patrol
the beauty of source control, easy rollback of changes
idle owl
Truth, but I can't get travis ci to check it specifically without creating a PR, so I'll test it first and then do the PR.
raven canopy
gee-wiz question: is it intentional to return an expression if [1], and a tuple if [2]? From the pydocs:
A single expression without a trailing comma doesn’t create a tuple, but rather yields the value of that expression
idle owl
I have zero clue on that one.
I removed parens on the second one
I think it might still be a tuple though
raven canopy
yeah. from what I'm reading, the parens won't make a difference.
raven canopy
so if you put a comma after a single item, the return will be a singleton (1 deep tuple)
idle owl
I remember that from testing code a while back, yeah
split ocean
@slender iron thanks, this code is working well!
idle owl
@raven canopy @opaque patrol The update works!
raven canopy
sweet! how did it end up?
idle owl
Need to do the PR to find out if travis passes, but I tested it with the sensor and that works
raven canopy
they should rename travis to zoul. "Are you the gatekeeper?" 😄
idle owl
I don't even know how you'd change that count anyway. or why. So I'm not sure how to even see if that part is working or failing. Does the example code even use it? No idea.
Now we wait for a bit.
For travis ci.
Hey! It passed!
Thanks for the help!
slender iron
@lucid arch I believe you are having memory troubles because you are using the py bundle instead of a mpy bundle. Try this: https://github.com/adafruit/Adafruit_CircuitPython_Bundle/releases/download/20171216/adafruit-circuitpython-bundle-2.1.0-20171216.zip
@split ocean 👍
pastel panther
My fun for the evening; with any luck they'll run 3.0 with some tweaks
idle owl
@pastel panther Is this the right reset button for the metrowings? https://www.adafruit.com/product/1489
Nm, evidently it should work 😃
idle owl
@pastel panther @solar whale did and is sending me one. Turns out he already has those so it wasn't an issue.
pastel panther
@idle owl Here's the BOM for future reference: https://octopart.com/bom-tool/s72yPqrf
idle owl
Oh nice! Thanks
pastel panther
The only thing of note is the switch since the footprint is made to fit that one, but any switch that can bridge the appropriate pads will work
It needs to short one of the pairs of pads on the 'short' side of the square
err rectangle
brb
solar whale
I’ll try to test the tactile switch tomorrow and let you know if it works.
idle owl
Ok
Oh @solar whale Do you have mk2? I was looking at the wrong one I think.
Would have messed that up even more looking at the wrong one. Oops. Good thing you're on it.
pastel panther
I just un-shared the mki from OSHpark
idle owl
Ok, a lot of things make a lot more sense. I was looking at the wrong one. Needs a reet button with 4 feet instead of 2 now, it seems.
pastel panther
It only really needs two
idle owl
Oh nice!
solar whale
Yes , I am sending mkii. Reset pin is different. So @pastel panther the tactile switch wold go between two pins on”short” side?
pastel panther
There are two pairs of feet that are essentially shorted together
idle owl
I see
pastel panther
Yes; If you follow the trace from the last pin in the nearby row, it goes to the two closest pads
idle owl
Ah yes it does
pastel panther
meaning they're already connected so you need to short two pads on either of the short sides
If you're really itching for the 'right' button you can always steal one from a metro; it's the same switch
idle owl
6mm x 3mm though it looks like, so any that size would work as long as the feet matched the pads?
idle owl
Ok yeah
solar whale
Ok, I’ll see if it works and if not re-evaluate. I did not want to place a order for just the switches.
pastel panther
I think even one of those normal breadboard compatible tactile switches would work
You'd just need to bend the leads flat
idle owl
Right
solar whale
Good night all!
pastel panther
night @solar whale
robust coral
Funny brainless CPX CircuitPython toy. Accelerometer direction tells the Servo to swing hard the other way. Direction indicated by NeoPixel color. I didn’t expect it to look so freaky.
pastel panther
lol
If you shorten the cable it might be a little bit more predictable, then again maybe that's not the point
robust coral
Haha or lengthen the arm.
pastel panther
Having a point is overrated
robust coral
Really whip it around!
The point was just to entertain/inspire a 6yr old for an hour. Mission accomplished!
idle owl
Nice!
robust coral
It feels like the kind of quickie thing Circuit Python is perfect for, so I thought I’d share.
idle owl
Thank you!
idle owl
@tulip sleet https://github.com/adafruit/Adafruit_CircuitPython_TSL2561/pull/9 is the PR for the code from earlier
Thanks!
raven canopy
ahhh. my old friend, debugging. this is the fun part, right? 😐
idle owl
Yeah already have it. Do I just need to test it on 2.1?
slender iron
I tested it on both 2.1 and 2.2
idle owl
Ok, I tested it on 2.1, no errors. Was it failing using the class in general? Or did it fail specifically using lis3dh for something
slender iron
the init fails
idle owl
Excellent, tested on both.
slender iron
wanna approve?
slender iron
thanks @idle owl! I'll release it
idle owl
Great! Thanks for catching that and getting it resolved!
timber mango
https://github.com/wa1tnr/cpx-basic-studies
Got a bunny sketch going on CPX using lis3dh accel. 😉
slender iron
yup, np
idle owl
@slender iron I went through and updated the lint all the libraries issue with what's pending, what's done and what's remaining. I'm not sure the difference between "PRs needing review" and "Pending PRs" but I didn't make that change so I didn't alter it. Everything that has an outstanding PR that I found went into "Pending PRs" and I tested some more today too.
And I did releases on a few that had been finished and merged, but no release was done.
slender iron
thanks @idle owl !
timber mango
screenshot!
timber mango
Anyone up for a robotic remote haircut? 😉 It's weird to have something in your hand that makes the screen image do the same thing your hand is doing, when it's entirely rotational, rather than mouse movements on an x-y grid.
timber mango
@tulip sleet : yes, thanks!
raven canopy
<@&356864093652516868> anyone dealt with this error when compiling (background: adding a parameter to microcontroller.cpu):
In function '_sbrk_r': sbrkr.c:(.text._sbrk_r+0xc): undefined reference to '_sbrk'
umbral dagger
That's a linker error... looks like it's missing a library
raven canopy
well, I found a syscall that handles '_sbrk', but it is in a dir different that the location given for the compile error. hmm.
umbral dagger
And it built fine before your change?
raven canopy
yeah. even built during changes.... back to regressing to maybe narrow this down.
narrowed it down to my property getter. now to try and understand why...😵
fading solstice
@idle owl There is no difference between the Pending Review and Pending RPs. You can move RGB_Display into the Pending PRs section and delete Pending Review. Sorry.
indigo hazel
I am running into an issue with my new Circuit Playground Express where when I add the latest bundle, I get an error in the serial terminal. Can someone help me out?
tidal kiln
@indigo hazel what is the error?
tidal kiln
how large is your code? if it's not too large, paste it here.
indigo hazel
import adafruit_si7021
import busio
import board
i2c = busio.I2C(board.SCL, board.SDA)
sensor = adafruit_si7021.SI7021(i2c)
print("Temperature:", sensor.temperature)
print("Humidity:", sensor.relative_humidity)
tidal kiln
how do you have the SI7021 attached to the CPX?
indigo hazel
Oops, I am following the online guide, and ran into this error. Now I see that I do not have that hardware, hence the error.😶
tidal kiln
yep. that's it. which guide are you following?
indigo hazel
tidal kiln
hmmm. i think that could be updated a little to make it more explicit that the example is for an additional hardware item .
indigo hazel
yes, a little confusing when starting out for the first time.
tidal kiln
and even more so when you have a CPX that does have a lot of sensors already on it
raven canopy
i was just going to suggest that. NOTE: this example will not work properly without an external SI7021 sensor connected to the CPX. This is just an illustration of how to import libraries.
that really is the hard part about writing guides focusing on early users. you wan't to keep it fun, simple, and light-hearted (which ladyada does SO well), without getting too technical and littered with if...s, notes, and such.
thanks @indigo hazel for bringing the issue up. as has been echoed time and again, this is what makes open source work.
indigo hazel
No problem.
tidal kiln
i sent in a suggestion to have that looked at
tulip sleet
@tidal kiln a user in the forums was similarly confused. I think we might use Adafruit_CircuitPyton_HID as the example instead, since it requires no external hardware. I'll revise that section.
tidal kiln
thanks @tulip sleet
might be good to have both, a generic nothing additional required example, but then also one that does show how one would get and install the library to go with whatever new gadget they bought (more or less like the existing one)
tulip sleet
@tidal kiln @idle owl realized HID is not a good example, because it's one of the libraries included in lib for Trinket and Gemma M0. I like your phrasing and will add it for now.
idle owl
@tulip sleet @tidal kiln The example that's replacing that one uses simpleio and blinky, so the example for installing libraries will work once it's done. It's much more clearly explained.
tidal kiln
@indigo hazel ladyada just updated the guide pages, take a look and let us know how it goes. hopefully it helps out better now.
meager fog
@idle owl i think i did a fair job re-orging that page, dunno if there's anything you want to grab!
idle owl
@meager fog All of the libraries page has been integrated into the Welcome guide already, do I need to go through it and compare it again?
idle owl
Ah I could update the images.
meager fog
nah, ill get to revewing the guide maybe later today, still getting thru emails
idle owl
The libs page is split between the libs page in the Welcome guide and the Troubleshooting page in all the other guides
meager fog
im doing cpx text now
idle owl
Nice! np
meager fog
so instead of set_click...whats the current cpx lis3dh api? 😃
idle owl
That was updated, it skips it if you're running 2.1.0
meager fog
im on 2.2-rc`
i was using
Set up the accelerometer to detect tapping ('click')
TAP_THRESHOLD = 20 # Accelerometer tap threshold.
cpx._lis3dh.set_click(1, TAP_THRESHOLD) # detect single tap only
meager fog
not single? 😃
idle owl
Yeah that was what I was saying I Could explain if needed
meager fog
yah
idle owl
Turns out you can't have a double click event without two single click events, it's a hardware/software thing, it's not possible. A double click event is two single click events, so if you wanted to have double click do one thing and single do another, you would have conflicting responses, so if one was red LEDs and double was blue, you'd get red or purple.
So after a huge discussion about it, we decided to include only double_click
meager fog
yah
idle owl
You can do both with LIS3DH, it's based on setting single or double, but you can't do both
meager fog
yah i remember this from the arduino driver
idle owl
Double is harder to have accidental clicks as well
meager fog
so how do i detect a single click then?
idle owl
so we decided double made the most sense.
You'd have to set it up using lis3dh, not cpx class...
meager fog
ok
idle owl
In this line: lis3dh.set_tap(2, 18, time_limit=4, time_latency=17, time_window=110)
You would set 2 to 1
and it detects single instead
Having both in cpx wouldn't have worked and would have had that huge caveat of both being present with double
idle owl
so Scott redesigned LIS3DH to only detect single OR double based on what that line is. The line that reads the tap returns "true" instead of single/double, and what that event is translated to is based on what you set in set_tap
meager fog
kk
idle owl
Yeah because the idea was that double_tap would have worked instead of needing _lis3dh being public.
We didn't know you'd want to stick for sure with single tapping instead of double_tap working.
meager fog
i think, its a little confusing still 😃
idle owl
Basically, express only has double_tap capabilities built in.
meager fog
why not have it either single or double?
cpx.detect_taps = 1 or 2
then cpx.tapped
?
idle owl
I'm not sure, it was a design decision, but I think it could be done.
Issue is if you try to do both I guess
That was what we were trying to avoid was the huge caveat, but I guess since the driver was redesigned it might be different.
I can look into it tonight if you'd like
meager fog
k i can also try both, double-tapping is just really hard to time
so i usually just have single-tap detection! i can also try changing i tup
see if i can get both working. do i need to mpy the library to fit, still?
idle owl
It tested pretty well, I had other people do it too. The idea being that if random people could pick it up and get near 100% response from the board, it was the right idea.
Yeah
It's an easy change to get it to single.
meager fog
k let me try it, see what i can get!
idle owl
Getting it to double is a variable.
Ok!
I'll take a look at it later if needed, and I can test it for you later as well if you'd like
You can check the data sheet for what the rest of those numbers are for, it took tweaking them solidly to get it to where it worked consistently.
meager fog
looks like you cannot use with mpy?
idle owl
How so?
meager fog
huh! thats odd, on 2.2?
idle owl
oi yes ok.
meager fog
k np ill do that!
idle owl
I'm back on the right page. Yes it needs mpy
meager fog
ha! ok - yah this is what i wrote for arduino https://github.com/adafruit/Adafruit_LIS3DH/blob/master/examples/tapdemo/tapdemo.ino
idle owl
Yeah that's the idea, but it always returns a single around the doubles on circuitpython
interrupts aren't implemented
the way we have it now, it returns true when a click event happens, and then tells you whether it was single or double based on what you set it to be in the set_tap
meager fog
yah
idle owl
It's nearly dinner time here so I'm going to head off for a bit. I'll be back later. Ping me if you have any other questions though, I'll have my phone.
solar whale
@idle owl @pastel panther I don't htink the tactile switch will work very well on the metrowing. At least, not with my soldering skills. I think I'll just tack on two wires and use a pushbutton until I can get some of the correct switches.
meager fog
@idle owl pr is sent, small change but i think will be good. its really annoying to time the double tap 😄
half sedge
What is supposed to be the library bundle release matching the 2.2.0rc1 firmware release?
By the way, I changed my nick name from Instinctive to DavidGlaude. That is best because it is my name. 😃
And another question, is there a plan to release the big demo code from Tony for the Circuit Playground Express.
solar whale
@half sedge the latest 2.x Bundle should be what you want - in this case it is silt lableed as 2.1.0-20171217.
half sedge
Doc seems to say lattest... https://github.com/adafruit/Adafruit_CircuitPython_Bundle/releases/tag/20171217
Thanks @solar whale
meager fog
2x matches all 2x
solar whale
@half sedge New Bundles are now released daily if one of the libraries has been updated, so check back often!
The only difference (I think) betwen the 2.x and 3.x bundles is the version of mpy-cross used to build them.
half sedge
Just checking, but is this https://github.com/mrmcwethy/Adafruit_CircuitPython_APDS9960 some kind of official to be driver?
Like to be in the bundle one day?
If it work (and I was planning to test), that would be great an there are some issue to be updated.
I am cleaning my useless and now empty place holder (with a typo): https://github.com/dglaude/adps9960-circuit-python
meager fog
it looks like its in-progresss, @fading solstice has been working really hard with us to clean up our existing drivers, so there is no ETA 😃
half sedge
in-progress is better than the nothing I had on my side...
pastel panther
success!
errant grail
@pastel panther Is that a Roman numeral pin numbering system? 
pastel panther
@meager fog Thanks, it is indeed! It's my bitsy-ish m4 express board. It's got QSPI on the top and a spare 'normal' SPI underneith for ease of porting; Once I get CP working on it I'm going to try getting QSPI going
@errant grail Yes, on one side. I had to use a X for a 10 for space issues and then ran with it. I moved things around so the 10 would fit but I'm digging the roman numerals
pastel panther
@errant grail I'm working on my '11 livewell board now also. I should have the board layed out by the end of the week
I saw someone using safety pins to program a ESP8266 via the castillated pins so I might try that with this rev of the motor board as to not squander valueable mm^2
errant grail
@pastel panther I'm looking forward to seeing that design. Haven't played much with castillated pins yet. Have you done that with other projects?
pastel panther
@errant grail nope! I'll be following OSHPark's guidelines and hoping for the best. For the first version I'll probably have some pads for redundancy
errant grail
... that would be a good way to make a smaller StringCar PCB and still use a stock Trinket M0 and 8833 motor controller. Thanks for the idea!
@pastel panther I'm working on PCBs for StringCarGen2CPy and one for a secret retro-inspired project. Should have them done tonight if the prototypes work as designed.
pastel panther
@errant grail cool, I'd love to check them out when you're done
errant grail
They're pretty simple, but could lead the way to a custom M0-based board similar to the SMD aTtiny85 controller I made a year ago.
@pastel panther Okay, mebbee an M4-based board..
raven canopy
@pastel panther nice work on the bitsy-m4 board.
pastel panther
The SAMD21E can be had in a 32 pin format which is pretty small
@errant grail I'm happy to share my m4 board as a starting point once I've squashed any bugs
Thanks @raven canopy!
errant grail
@pastel panther I'll take you up on that offer! Especially since it has Roman numerals. M-IV.
pastel panther
@errant grail Almost right...
slender iron
<@&356864093652516868> We're on for our normal meeting tomorrow at 11am Pacific / 2pm Eastern here on Discord. The following few weeks will be weird due to the holidays and I'll post later what the plan is. (We'll discuss it tomorrow.)
slender iron
Rather not 😃
cunning crypt
Man, no fun at all!
slender iron
I was thinking we should do one the week in between
cunning crypt
There is no Monday between Christmas and New Years? Or do you mean sometime Tuesday the 26th to Sunday the 31st?
slender iron
Yeah, some other day is what I'm thinking
cunning crypt
Friday might be a good one. Might not, though.
slender iron
my girlfriend has friday off so I should probably avoid it
cunning crypt
I expect Tuesday wouldn't be a grand idea - people traveling and all
slender iron
yeah, I figure we can discuss it during the meeting
cunning crypt
I'll be working, which is why I'm putting my thoughts down.
A weekend day might be able to grab a bunch of people, but then again - it's a weekend day.
Odds are I wouldn't be able to come anyway though.
Stupid work and holidays and stuff.
slender iron
yeah, which day would be best for you?
(anyone else who can't make the meeting please post as well)
cunning crypt
Monday, Tuesday, and Friday are the days I'd be able to make it, unless it was done later.
pastel panther
I'll be off work after the 22nd so I can likely attend for once whenever it is, unless there is an impromptu fishing trip
cunning crypt
After 5pm Eastern I'd be able to do Wednesday/Thursday. Saturday's right out for me.
slender iron
kk
cunning crypt
Unless you want to hold it absurdly late (>10pm Eastern) or stupid early (<9am Eastern)
slender iron
kk, good to know since I'm 3 hours behind
cunning crypt
Yeah, I'm keeping that in mind. I don't think a 7pm would be too bad for you, but 6am would be... not the best.
slender iron
yeah, 6am won't happen with me 😃
idle owl
@meager fog I'm looking at the PR now
umbral dagger
@slender iron I see that I'm a CP helper. I'll start showing up for the meetings, that time will typically work well. Alas, tomorrow I will be watching Star Wars then.
lofty topaz
@umbral dagger Congrats. Enjoy SW.
keen urchin
Just received my circuit playground express, but I'm having trouble getting circuit python on it. I'm on a Macbook Pro running Mac OS High Sierra version 10.13.2 . When I first plug it in it shows up as CPLAYBOOT. I followed the guide and downloaded the latest version of circuitpython. adafruit-circuitpython-circuitplayground_express-2.1.0.uf2 I drag it over and drop it in CPLAYBOOT. It ejects and then remounts as a camera I own that is not plugged in "Nikon D3400". I went ahead and tried to load the adafruit library as suggested in the guide, but then I get an error message "The operation can’t be completed because one or more required items can’t be found. (Error code -43)". I've tried other files just to see if something would work, but the same error message pops up. I have even used the erase file to completely erase everything flash_erase_express.ino.circuitplay.uf2 still no luck. The process repeats itself. I have also downloaded all files again, but I run into the same problem. Does anyone have any ideas? #circuitpython-dev #help-with-projects
tulip sleet
@keen urchin so you don't see CIRCUITPY in a finder window, but only "Nikon D3400"?
@keen urchin if you see this later, see https://forums.adafruit.com/viewtopic.php?f=57&t=128112 aboout checking the drivers on your Mac and removing any that may be interfering (certainly sounds like the camera driver might be).
pastel panther
hey @slender iron; I'm I correct in thinking that the dotstar for the trinket m0 is not supported yet in 3.0?
slender iron
it should be since we have SPI support now
pastel panther
so it should work? It looks like it's still commented out in the trinket_m0 mpconfigboard.h
pastel panther
kk; I put a dotstart on my board so I'll throw that into the file for my board
If it works I'll re-enable the one for the trinket m0
slender iron
thanks!
pastel panther
I got uf2 loaded and working so I know the hardware is good
I was honestly surprised
slender iron
👍 nice work!
pastel panther
Thanks! I have no idea if the flash works but I guess we'll see in a few
limber tulip
@atletika5#4704 FWIW I just went through getting my first Circuit Playground Express running circuit python on my MacBook Pro on Saturday, with High Sierra, and it worked fine. Hopefully @tulip sleet advice will help.
fading solstice
@slender iron @idle owl i updated the library list as a result of merging bus_device and register https://github.com/adafruit/circuitpython/issues/475
stuck elbow
the meeting is in 45 minutes, right?
idle owl
Yes
stuck elbow
thanks
slender iron
@umbral dagger you don't have to join the meeting regularly though you are welcome. Enjoy Star Wars!
stuck elbow
tap tap tap tappity tap
idle owl
@cunning crypt Weekly if you're around
cunning crypt
At work!
idle owl
Ok!
manic glacierBOT
tidal kiln
NEXT = ??
NEXT NEXT = Jan. 8th
and normal after that
slender iron
<@&356864093652516868> Next meeting is: Thursday December 28th, 6pm Eastern / 3pm Pacific. If you can't make it thats OK! Enjoy the holidays.
timber mango
hiya - couldn't make the weekly, was busy with some other meetings, thanks everyone 😃
slender iron
@meager fog no worries!
errant grail
Thanks! Happy Holidays!
raven canopy
Happy Holidays errbody!
idle owl
while True:
cpx.detect_taps = 2
if cpx.tapped:
cpx.pixels.fill((0, 0, 30))
cpx.play_tone(292, 0.1)
print("doubletap")
else:
cpx.pixels.fill((0, 0, 0))
cpx.detect_taps = 1
if cpx.tapped:
cpx.pixels.fill((30, 0, 0))
cpx.play_tone(262, 0.1)
print("singletap")
else:
cpx.pixels.fill((0, 0, 0))```while True:
lis3dh.set_tap(2, 80, time_limit=4, time_latency=17, time_window=110)
last_tap = False
tap = lis3dh.tapped
if tap and not last_tap:
print("tapped!")
pixels.fill((30, 0, 0))
else:
pixels.fill((0, 0, 0))
last_tap = tap
lis3dh.set_tap(1, 80, time_limit=4, time_latency=17, time_window=110)
last_tap = False
tap = lis3dh.tapped
if tap and not last_tap:
print("tapped!")
pixels.fill((0, 0, 30))
else:
pixels.fill((0, 0, 0))
last_tap = tap``` @detect_taps.setter
def detect_taps(self, value):
self._detect_taps = value
try:
self._lis3dh.set_tap(value, 80, time_limit=4, time_latency=17, time_window=110)
except AttributeError:
pass
self._last_tap = False```Gotta run - talk to everyone later.
idle owl
tidal kiln
i2c = busio.I2C(board.ACCELEROMETER_SCL, board.ACCELEROMETER_SDA)
raven canopy
RE: Google SoC - definitely an OS who's who list of participants for last year. Python participated, as well as BeagleBone, so there is some relevance for CPy. Thought for discussion: if Adafruit/CPy were to apply to participate, would a discussion be needed/desired with the MPy guys? I may be closer related to the student side in GSoC, so I'm not advocating either way.
stuck elbow
@raven canopy I asked Damien last year about gsoc, but by the time we got to it it was already too late to apply - but the conclusion was that he doesn't mind doing it if someone will do all the paperwork for it
tidal kiln
stuck elbow
I have to run
tulip sleet
bye
idle owl
later!
slender iron
@tulip sleet and @idle owl I'm back
idle owl
Have a lovely lunch?
slender iron
haha
slender iron
@tidal kiln what are you up to?
tidal kiln
playing with accelo on cpx
slender iron
what are you going for? sounds like I missed some conversation about it when I was at lunch
opaque patrol
Has anyone worked on converting the Circuit Playground Accelerometer Mouse Example to CircuitPython yet?
tidal kiln
at this point i'm just testing it's behavior
tidal kiln
but yah, dan, kattni, and i had a big follow on discussion w.r.t tap detect
opaque patrol
Its in the Arduino Example list when you install Circuit Playground, give me a minute and I will see if I can find it on github
tidal kiln
opaque patrol
@slender iron Its in this repo https://github.com/adafruit/Adafruit_CircuitPlayground/tree/master/examples
slender iron
thanks!
opaque patrol
@slender iron I already created a color_sense version in circuit python, thought i would tackle some more
slender iron
great!
slender iron
<@&356864093652516868> Here is the recording from the meeting today: https://youtu.be/vp5_u3oV9Ro We'll do our next meeting for those who are available on Thursday December 28th at 3pm Pacific / 6pm Eastern here on Discord. Have a great holidays and new year!
slender iron
@timber lion why won't this work with 3.0? https://github.com/adafruit/Adafruit_CircuitPython_MAX31865/releases/tag/2.0
tidal kiln
@idle owl been any further discussion on tap?
tidal kiln
don't need to if you think you're getting it done. don't need to get too many cooks.
tidal kiln
cool.
opaque patrol
finished accel_mouse.py example https://github.com/BravoDelta151/CircuitPython_Misc/blob/master/accel_mouse.py
idle owl
@opaque patrol Have a few minutes for a chat?
opaque patrol
@idle owl sure
idle owl
@opaque patrol I'm looking at your color sense code. This looks great! I changed the last bit to this: ```python
cpx.pixels.brightness = .2
verbose = False
while True:
if cpx.button_a or cpx.button_b:
verbose = cpx.button_b
time.sleep(.05) # debounce
if verbose:
print("button pressed")
while cpx.button_a or cpx.button_b:
time.sleep(.01)
r, g, b = sense_color(verbose)
print("({}, {}, {})".format(r, g, b))
cpx.pixels.fill((r, g, b))
else:
cpx.pixels.fill(CLEAR)
@idle owl I think that will clear the pixels immediately after setting the color. What I can do is store the color variables and just update them on the button press
idle owl
Oh it does do that! I couldn't get it to work the other way though. I wonder what I was doing wrong?
opaque patrol
not sure, I am playing with it now and trying to see if I can add gamma correction to get a little closer to the color
idle owl
Oh! It's working for me now the way you had it. I see!
opaque patrol
I have the clear in the except block to clear if you hit ctrl-c, I plan to change that to clear on KeyboardInterrupt and another except block for other exceptions
idle owl
Ah ok
opaque patrol
If an exception is thrown with the current code, it will hide the error message
idle owl
Got ya. Yeah that could be frustrating 😃
opaque patrol
except KeyboardInterrupt:
cpx.pixels.fill(CLEAR)
except Exception as e:
print("Exception", str(e))
like this
idle owl
oh ok yeah
opaque patrol
you may need to adjust the brightness, with it set at .2, colors with all values below roughly 55 will not appear to work
idle owl
Right, I've run into that multiple times before I remember I can't set it to (0, 0, 10) with brightness set super low.
idle owl
@tulip sleet I'm testing AMG88XX. I'm getting an error - TypeError: 'float' object not iterable on the first print line:```python
import time
import busio
import board
import adafruit_amg88xx
i2c_bus = busio.I2C(board.SCL, board.SDA)
amg = adafruit_amg88xx.AMG88XX(i2c_bus)
while True:
for row in amg.pixels:
#pad to 1 decimal place
print(['{0:.1f}'.format(temp) for temp in row])
print("")
print("\n")
time.sleep(1)
@idle owl what library is that?
tidal kiln
this one?
https://github.com/adafruit/Adafruit_CircuitPython_AMG88xx/blob/master/adafruit_amg88xx.py
tidal kiln
try this in repl, what does it show?
import busio
import board
import adafruit_amg88xx
i2c_bus = busio.I2C(board.SCL, board.SDA)
amg = adafruit_amg88xx.AMG88XX(i2c_bus)
amg.pixels
[[0, 0, 0, 0, 0, 0, 0, 0], [0, 0, 0, 0, 0, 0, 0, 0], [0, 0, 0, 0, 0, 0, 0, 0], [0, 0, 0, 0, 0, 0, 0, 0], [0, 0, 0, 0, 0, 0, 0, 0], [0, 0, 0, 0, 0, 0, 0, 0], [0, 0, 0, 0, 0, 0, 0, 0], [0, 0, 0, 0, 0, 0, 0, 0], 24.5, 23.0, 22.0, 22.75, 22.75, 22.5, 22.75, 23.0, 22.25, 22.5, 22.5, 22.25, 22.75, 22.75, 23.5, 23.25, 22.25, 22.25, 22.25, 22.25, 23.25, 23.5, 24.5, 23.25, 22.5, 21.75, 22.5, 22.25, 22.75, 24.0, 25.0, 23.75, 21.5, 22.25, 22.25, 22.5, 23.25, 23.75, 24.25, 24.25, 21.75, 21.25, 22.25, 22.25, 22.5, 22.75, 23.25, 22.75, 22.0, 22.0, 21.75, 21.75, 22.75, 22.75, 23.0, 23.0, 22.25, 21.75, 22.25, 21.75, 22.75, 23.0, 23.5, 24.75]
tidal kiln
yeah, the buffer is initialized to 0's and then appended to
idle owl
Ok.. so something is wrong with the code then? It's not appending properly?
tidal kiln
yah. think so.
try changing this line:
https://github.com/tannewt/Adafruit_CircuitPython_AMG88xx/blob/3a0b88ed989c3f7fa01da46f86366ebc89a0d535/adafruit_amg88xx.py#L152
to:
retbuf = []
and run that above example again
tidal kiln
yep. it's just a list, not a list of lists.
based on the code comment, looks like that was the intent. just need to add a little more code....
try this:
@property
def pixels(self):
"""Temperature of each pixel across the sensor in Celsius.
Temperatures are stored in a two dimensional list where the first index is the row and
the second is the column. The first row is on the side closest to the writing on the
sensor."""
retbuf = []
buf = bytearray(3)
with self.i2c_device as i2c:
for row in range(0, _PIXEL_ARRAY_HEIGHT):
rowbuf = []
for col in range(0, _PIXEL_ARRAY_WIDTH):
i = row * _PIXEL_ARRAY_HEIGHT + col
buf[0] = _PIXEL_OFFSET + (i << 1)
i2c.write(buf, end=1, stop=False)
i2c.readinto(buf, start=1)
raw = (buf[2] << 8) | buf[1]
converted = _signed_12bit_to_float(raw) * _PIXEL_TEMP_CONVERSION
rowbuf.append(converted)
retbuf.append(rowbuf)
return retbuf
Ah! Yah!
tidal kiln
works?
idle owl
Yep!
My hands are apparently so cold, I had to grab someone else to make sure it was actually working 😄
tidal kiln
see the issue?
idle owl
I... don't
tidal kiln
who wrote the original?
idle owl
Looks like Dean Miller
tidal kiln
i'm mean that PR version
idle owl
Scott
tidal kiln
need to talk to him about it, cause he's doing some changes and not sure what his intent was
idle owl
Ok
tidal kiln
looks like the original version was just a single list:
https://github.com/adafruit/Adafruit_CircuitPython_AMG88xx/blob/master/adafruit_amg88xx.py#L155
like what my fix to the code did:
[value value value etc...]
but looks like Scott wants to make that more like a 2D array (which makes sense):
[ [value value value ... ], [value value value ...], etc...]
like row,col
idle owl
Yeah
tidal kiln
but not sure why he needed to init it to all zeros, and he can probably come up with a more clever way of creating the retbuf list
mine was just a simple fix / hack
and then you were getting that error because this:
for row in amg.pixels:
was returning a float, instead of a list of floats
idle owl
Oh ok
tidal kiln
and then you were trying to iterate on that float:
print(['{0:.1f}'.format(temp) for temp in row])
Do you want to review as well? Otherwise I can do line by line and relay what we talked about.
tidal kiln
you can do it
idle owl
Ok
tidal kiln
guess it's kind of a question back to scott, cause i'm not sure what he's was trying to do.
idle owl
Yeah I'm going to request changes and let him respond
tidal kiln
since he pre-allocated the buffer, maybe he intended to directly index things instead of using .append? dunno.
like retbuf[row][col] = converted or something
idle owl
I'm going to tag you in fyi since you helped.
tidal kiln
k
interesting...
>>> buf = [[0]*4]*4
>>> buf[1][3] = 'foo'
>>> buf
[[0, 0, 0, 'foo'], [0, 0, 0, 'foo'], [0, 0, 0, 'foo'], [0, 0, 0, 'foo']]
>>>
updates all of them.
idle owl
hmm
stuck elbow
yes, the * operator doesn't make copies
it repeats a reference to the same object
you want:
buf = [[0] * 4 for _ in range(4)]
what you did is the same as:
x = [0] * 4
buf = [x, x, x, x]
thanks @stuck elbow , now if i can just remember that!
timber mango
hi all, has anyone had a problem where code.py has become corrupted and is unreadable? I've tried deleting it, but that doesn't work. Even tried reflashing the circuitplayground, the file is still there after, and still can't be deleted or edited
idle owl
Yep! You'll need to use the flash eraser on it. This will erase the board entirely and you'll need to reflash CircuitPython and add your libraries once completed. Follow the steps here: https://learn.adafruit.com/adafruit-circuit-playground-express/troubleshooting#circuitpy-drive-issues
idle owl
Yeah for sure!
umbral dagger
@timber mango You can also end up in a situation where the CICUITPY drive is read only. The flash erare & reinstall circuit python works to fix that as well.
@timber mango Loading an arduino sketch and then reinstalling circuit python seems to work in many cases as well.
timber mango
Keep in mind the SPI flashROM is a separate chip and is often ignored by the SAMD21G18A microcontroller. That's why its contents persist even after loading an Arduino sketch, 'on top of' a CircuitPython .UF2 boot file -- the Arduino sketch completely ignores the SPI flashROM chip, unless you specifically tell it not to ignore it, in some way.
manic glacierBOT
umbral dagger
Right... i was thinking of non-external flash boards.
timber mango
I'm still pleased and surprised every single time I find that my files have remained present on the SPI flashROM of an Express target board, after a major operation (such as loading a different .UF2).
It hasn't sunk in yet, that my files are going to be there, unless something unplanned has happened.
stuck elbow
@timber mango I think it's a good practice to be prepared for them to be gone, though
pastel panther
@slender iron I was finally able to compile 3.0 for my m4 board, however when I load it my dotstar turns purple and I have no MSC or CDC; since it's not flashing after the purple, I'm assuming it's the bootloader and not CP that is setting it?
slender iron
@pastel panther yeah, that sounds suspect
I'm not sure why it'd be purple. I think that might be an arduino thing
pastel panther
Someone might call it pink?
slender iron
¯_(ツ)_/¯
pastel panther
I should look at uf2-samd for a decoder ring?
slender iron
I'd expect it to be red or green if actually in the bootloader
pastel panther
calling gdb with firmware.elf doesn't seem to do what I want. Is there something I can do to get source lines showing? gdb is saying No symbol table is loaded
pastel panther
right
ah, this is more informative: HardFault_Handler () at supervisor/port.c:264
backtrace is only so useful; @slender iron The '51 datasheet mentions a Embedded Trace Buffer which I assume is different from the Micro Trace Buffer so the MTB setup in your GDB tutorial won't apply here?
slender iron
nah, I haven't been able to get it going
try breaking at main() and see if it gets that far
idle owl
@slender iron With AMG88XX, I'm now getting IndexError: list index out of range with the example. Traceback (most recent call last): File "code.py", line 10, in <module> File "adafruit_amg88xx.py", line 164, in pixels File "adafruit_amg88xx.py", line 164, in pixels IndexError: list index out of range
idle owl
what do I need to change to test that
slender iron
I amended the commit twice
idle owl
I can try again
pastel panther
@slender iron yea, it's getting to main; I'll have to play around with it after work. On a positive note, it would seem that maybe (maybe?) the dotstar changes I wrote worked?
idle owl
But I am pretty sure I did
slender iron
@pastel panther maybe, it might be whats crashing 😃
@idle owl whats the pre-allocation line for you?
pastel panther
@slender iron that's where the smart money is ;P
pastel panther
thanks for the help 👋
slender iron
should retbuf = [[0]*_PIXEL_ARRAY_WIDTH for _ in range(_PIXEL_ARRAY_HEIGHT)]
np @pastel panther good work!
idle owl
retbuf = [[[0]*_PIXEL_ARRAY_WIDTH] for _ in range (_PIXEL_ARRAY_HEIGHT)]
extra brackets in mine
slender iron
yeah, I fixed that 😃
slender iron
yeah, I was sneaky and amended. you just were super fast
slender iron
👍
idle owl
I went through and found all the circuitpython-travis-build typos left and fixed them. There's a PR in for one of them that I didn't catch until after I merged it, so I went and found all the rest, fixed and committed them.
slender iron
thank you!
idle owl
https://github.com/adafruit/Adafruit_CircuitPython_CCS811/pull/8 if you want to nab it and I'll do the release when you get to it
slender iron
merged
slender iron
😃
manic glacierBOT
idle owl
@slender iron adafruit_irremote isn't autogenerating the zip/mpy files, and isn't autobundling, is that intentional?
idle owl
Ok
idle owl
Oi, see? Same exact thing that it caught the other day. That must have been added in the update.
slender iron
it must skip the deploy when the build fails
idle owl
Ah ok
JP's code stopped working. I just redid it to not use CPX at all, and it's working again.
slender iron
define stopped working
idle owl
Every time you press a button on the remote: Memory allocation failure: 512 bytes and it would pick up 3 pulses only, fail to decode, and repeat because you wrote it into the code to continue after a memory allocation failure.
I had it happen before he did but assumed it was me.
slender iron
right
slender iron
did he send you the updated version?
idle owl
Unless you posted another version, yes
You had "here is the latest version that I got working" or something like that?
That's what I was using. From a gist.
slender iron
right, I imagine he changed it though
idle owl
He didn't imply that he did
slender iron
or did you just use it with 2.2?
slender iron
yeah, perhaps
idle owl
I mean it's back and works great without using cpx
Which is what I altered before you posted a better option with the memory allocation exception
slender iron
ya, with less code loaded
idle owl
He's ok with running it without cpx afaik
slender iron
kk, thanks for helping him out!
idle owl
I specifically loaded it all fresh - that's how I noticed the lib wasn't in the bundle
so double bonus discovery there.
slender iron
it not being in the bundle could be a different problem
idle owl
Ah ok
I'll have to mention that to him, since at the moment the only way to get the .mpy version of it is to make it yourself.
slender iron
are you gonna try fixing the irremote build or should I?
idle owl
oh, I can try, but I'll need help. I fixed that issue in another lib the other day but I doubt it's exactly the same issue.
slender iron
I'll take a look
idle owl
ok
slender iron
I haven't seen that issue so it'll be a good thought process
idle owl
I had a lot of help on the lib I fixed before. If you search for inconsistent-return-statements in discord you'll find the convo if you want to see it
slender iron
looks like it was just confused by the infinite loop in r ead_pulses
idle owl
ah
idle owl
Merged. Want me to try another release?
slender iron
yes please
idle owl
(for the record, I tested it first)