#avr Logs

Sep 03 2022

11:03 AM LeoNerd: Well, that was deliciously easy. I have a small bit of code on an ATtiny814 using the pulsewidth measurement mode of TCB0, and one of those really cheap ultrasonic distance sensor modules. I'm getting quite accurate results straight away here
11:12 AM qu1j0t3: :)
11:12 AM qu1j0t3: congrats
11:12 AM LeoNerd: Now I just gotta box it up into something nicely watertight and sit it in the top of my waterbutt, so I can measure rainwater
11:12 AM LeoNerd: It works really well in there. Plastic barrel, dead-flat water surface
11:14 AM exp: well personally i think that probably rules out any dynamic allocation
11:14 AM exp: and it likely means a really very small packet buffer
11:14 AM LeoNerd: Yup indeed
11:14 AM exp: i would not bother with anything more than a simple sequence number and acks
11:14 AM LeoNerd: Yup
11:15 AM LeoNerd: I'm not complaining because it's hard to design
11:15 AM exp: then your choice of PEC etc is up to you, that's another Hard™ problem
11:15 AM LeoNerd: I'm complaining because it's 2022 and I don't know why *I* have to be the first one to design and make it now, when it should have been solved 30 years ago
11:15 AM exp: right what i'm saying is the bits that are feasible have been solved
11:15 AM exp: SMBus/PMBus is used more than I²C
11:16 AM exp: but some parts, like reordering and retransmission over a lossy network, don't have correct solutions
11:16 AM exp: only acceptable tradeoffs