06:48 <re_irc> <@djc:mozilla.org> : thanks for the suggestions around the clock stuff!

06:51 <re_irc> <@djc:mozilla.org> so my next issue is that I'm trying to flash a minimal blinky LED thingy onto the dongle, but I can't figure out the right memory layout thing? I tried the "memory.x" for the dk and the dongle from the embedded-trainings-2020 repo (https://github.com/ferrous-systems/embedded-trainings-2020/blob/main/boards/dk/memory.x) (and the memory layout...

06:51 <re_irc> ... (https://infocenter.nordicsemi.com/index.jsp?topic=%2Fsdk_nrf5_v17.0.2%2Flib_bootloader.html&cp=7_1_3_5_0_7&anchor=lib_bootloader_memory) documentation), but when I try to build with the dongle layout I get errrors like this: https://gist.github.com/djc/b4d59299e7cc803697eb8a9c3378455c and when I try the dk layout nrfdfu says "error: firmware starts at address 0x0, expected an address equal or higher than 0x1000 to avoid a collision with the...

06:51 <re_irc> ... bootloader" (which makes sense given the layout stuff)

06:52 <re_irc> <@djc:mozilla.org> (I did also copy the profile stuff from https://github.com/ferrous-systems/embedded-trainings-2020/blob/main/advanced/firmware/Cargo.toml#L30)

06:53 <re_irc> <@djc:mozilla.org> (is this an okay place to ask these kinds of questions, or is there some other, more appropriate venue?)

08:02 <re_irc> <@thejpster:matrix.org> You might try the nrf-rs room

08:20 <re_irc> <@burrbull:matrix.org> : Maybe add https://github.com/rust-embedded/awesome-embedded-rust#community-chat-rooms link in the header?

22:32 <kenny> Not sure if there's a better channel for this, but I'm trying the lpcomp code in the nrf-hal-common, and the inputs _have_ to be Floating. I didn't find a mention of this in the datasheet, so I was wondering if anyone knew if this was a hardware requirement, or a bug in the code and Input<PullUp> should be supported?

22:35 <re_irc> <@jamesmunns:beeper.com> Might be worth checking the errata?

22:37 <re_irc> <@jamesmunns:beeper.com> Which nRF are you looking at? Tho I don't see any errata

22:39 <re_irc> <@jamesmunns:beeper.com> I don't see any significant difference in the pin config of pullup vs floating: https://github.com/nrf-rs/nrf-hal/blob/master/nrf-hal-common/src/gpio.rs#L187-L211

22:41 <re_irc> <@jamesmunns:beeper.com> Though if the internal pullups (relatively small, maybe something like 30k?) are strong enough to interfere with your external signal, that could be a problem I guess?

22:59 <re_irc> <@peter9477:matrix.org> : 11-16k Ohms, 13k typical

23:03 <kenny> I'm looking at the nrf52840. The code that will only accept Floating is at https://github.com/nrf-rs/nrf-hal/blob/master/nrf-hal-common/src/lpcomp.rs#L269

23:06 <kenny> (what I'm looking to do is get an event on rising and lowering edges. I first tried the gpiote code, but it seems that only does one edge (although I saw some people saying using two works on the same pin, even if not officially supported). I thought maybe lpcomp work be better, but then ran in to the pull-up issue)

23:09 <re_irc> <@peter9477:matrix.org> kenny: Was curious too, so I found this, which "confirms" (to the extent we trust the Nordic engineers fully, which we don't) that the pulls can stay active when a pin is used as an AIN: https://devzone.nordicsemi.com/f/nordic-q-a/23212/adc-with-internal-pull-up/91305

23:10 <re_irc> <@peter9477:matrix.org> Actually not confident he's with Nordic, but his response seems legit. There was another forum thread where an actual engineer claimed otherwise, but the OP in that case followed up and said after checking, it was clear the engineer was wrong...

23:11 <kenny> lol. Okay, thanks. I guess I'm in to undocumented either way. Maybe I'll give it a try

23:15 <re_irc> <@jamesmunns:beeper.com> Ah! I missed that it was "doesn't compile" and not "compiles but doesn't actually work"

23:15 <re_irc> <@jamesmunns:beeper.com> then yes, that just looks like a limitation in the hal code :D

23:20 <re_irc> <@peter9477:matrix.org> Hehe, found a post where the OP linked to some Embassy code, and the Nordic guy responded "We don't support RUST"... (but was otherwise helpful... he just meant he's not going to go read a bunch of Rust to help the guy with the source code)