<_whitenotifier-e>
[GlasgowEmbedded/glasgow] i-infra 66833cd - applet.measure.curve_trace: implement two quadrant (1.8v-5v) and four quadrant (-1.7v to +1.5v) curve tracing for quantifying the behavior of LEDs, other diodes, and resistors
<_whitenotifier-e>
[GlasgowEmbedded/glasgow] i-infra 66833cd - applet.measure.curve_trace: implement two quadrant (1.8v-5v) and four quadrant (-1.7v to +1.5v) curve tracing for quantifying the behavior of LEDs, other diodes, and resistors
<_whitenotifier-e>
[GlasgowEmbedded/glasgow] github-merge-queue[bot] 6b7c602 - Deploying to pages from @ GlasgowEmbedded/glasgow@a627726e8999d07bbf07f03757f3a1e4ec0c2354 🚀
ldcd has quit [Ping timeout: 276 seconds]
<alephbias[m]>
I have an old corpo mini-PC that I’ll use for some Glasgow work, so that’s my remoting solution
<alephbias[m]>
Could you add something like a cortex-m connector with a small footprint to expose the GPIOs of the STM32 as a third bank, but with reduced functionality? Even if it’s not reprogrammable, it could be used for simple things like EN or pulling a pin low for entering ESP32
ldcd has joined #glasgow
ldcd has quit [Ping timeout: 245 seconds]
ldcd0[m] has joined #glasgow
<ldcd0[m]>
whitequark: the main impediment for me with glasgow remoting is just dealing with python
<ldcd0[m]>
So a little statically linked binary that just forwards the protocol layer would definitely be handy
<ldcd0[m]>
Basically in my (admittedly very abnormal) use case, my Glasgow is often stuffed halfway inside something very weird in an environment that is deeply unpleasant to work in
<ldcd0[m]>
Like in that case it's a machine floor space with loud gas handling equipment and slated floors that are not nice to sit on
<abrasive>
I am +1 for glasgow remoting. not for such exciting reasons but due to Health™ I sometimes can't work at lab bench or whatever (and other times I'd just prefer to be sitting somewhere nicer)
<abrasive>
I do have an SBC there for this sort of thing and I greatly appreciate it when things can bridge relatively cleanly to the network rather than my having to juggle files back and forth as part of my debug loop
<whitequark>
alephbias[m]: i do not want to be on the hook to provide the stm32 in future revisions
<whitequark>
ldcd0[m]: yeah, the idea for remoting is to have a small rust binary that uses nusb and bridges that to a tcp socket
<whitequark>
josHua[m]1: oh, pyusbip is making usbip ... more bearable
<whitequark>
it's still a bad idea, but it's not as much of a bad idea. thanks!
<whitequark>
abrasive: yep, i also have Health™
<Darius>
I've used VirtualHere for remote USB to program microcontrollers using a raspberry pi in the lab
<whitequark>
a nice thing about having a native TCP connection multiplexer is that (a) it works over high latency links too, (b) unlike usbip it's not crash prone, and (c) at some point we're probably going to staple an ethernet phy to the device & that will use the same protocol
GNUmoon2 has quit [Remote host closed the connection]
<whitequark>
$ glasgow voltage
<whitequark>
Port Vsupply Vlimit Isupply Ialert Vsense Vsense(alert)
<whitequark>
A 0.000 4.913 0.000 0.000 0.000 0.000-0.000
<whitequark>
B 0.000 3.264 0.000 0.000 0.000 0.000-0.000
<whitequark>
new features!
GNUmoon2 has joined #glasgow
<whitequark>
$ glasgow voltage A
<whitequark>
Port Vsupply Vlimit Isupply Ialert Vsense Vsense(alert)
<whitequark>
A 3.322 4.913 (SHORT) 0.000 0.000 0.000-0.000
<alephbias[m]>
That makes sense to not overly be bound to the STm32! The impetus for it did ring a bell that I finally tracked down. I’m sure you’re already aware of the IC, but on the off chance you’re not: this board from Adafruit has an AW9523 gpio expander over I2C and may me helpful https://www.adafruit.com/product/4886
whitequark[cis] has joined #glasgow
<whitequark[cis]>
we have two PCA6416As GPIO expanders already
<whitequark[cis]>
the STM32 does SMBus to SPI bridging too
<whitequark[cis]>
in fact that is why we picked an MCU
<josHua[m]1>
pyusbip is really slow, and does have the same problems as all usbip (a badly behaved remote network agent can lock up the client’s usb stack forever). but if you’re in a bind, it works.
<josHua[m]1>
for this user, I think the actual answer is to superglue a Raspberry Pi Zero to their glasgow.