| 20 Jan 2020 |
| 22:14:54 |  mcbridejc | And I think that at least some of it could go away if I instead provide boxed traits instead of fully defined types. |
| 22:15:13 | mcbridejc | But I haven't fully fleshed that out yet |
| 22:15:22 |  jschievink | you can put them as associated type behind a driver-specific trait |
| 22:17:54 |  yatekii | agreed, that hides some of the type insanity :D |
| 22:22:18 | mcbridejc | I don't follow. What type are you proposing could have associated types? |
| 22:23:43 | jschievink | mcbridejc: define a custom trait that has all the types you need as associated types |
| 22:24:00 | jschievink | eg. https://jonas-schievink.github.io/rubble/rubble/config/trait.Config.html |
| 22:28:24 | mcbridejc | Ah, thank you for the example |
| 22:30:32 | yatekii | hmm does anyone know if there is a way to store a function handle in a const? |
| 22:31:03 |  HeroicKatora | As yatekii said there is static-alloc which provides a bump allocator that basically serves memory from the .bss segment. |
| 22:32:55 | HeroicKatora | I would be glad about feedback from people trying to use it in embedded |
| 23:02:21 |  Danilo | Using the blue pill, it seems that enabling PA0 and PA3 doesn't work with the stm32f103xx hal. PA1 and PA2 work. Does this sound familiar to anyone? |
| 23:26:08 | Danilo | Never mind, I destroyed the GPIO pins when the controller for the nixie tube blew up :D |
| 23:26:26 |  thalesfragoso | Danilo: I just checked some old projects, and I have used all those pins without any problems... |
| 23:26:45 | thalesfragoso | oh |
| 21 Jan 2020 |
| 00:33:18 |  adamgreig | jschievink: i hoped that would be the deref issue rather than empty loop, but i'll take it :p |
| 00:34:22 | jschievink | adamgreig (@freenode_adamgreig:matrix.org): a deref soundness bug? |
| 00:34:42 | adamgreig | the problem we have where if you have a &bla it might get read any time |
| 00:35:09 | adamgreig | causing our underlying mmio abstraction to be potentially troublesome for side-effecting reads |
| 00:35:10 | jschievink | Ah, that's not a rust bug |
| 00:35:16 | adamgreig | indeed |
| 00:35:31 | adamgreig | nevertheless i was hopeful it might somehow have been fixed :p |
| 00:44:46 | thalesfragoso | btw, is there any ongoing work on that ? It seems like a problem that should be solved before 1.0ing things |
| 00:45:10 | thalesfragoso | I mean the problem with VolatileCell and mmio stuff |
| 02:29:59 | adamgreig | little announcement, https://github.com/stm32-rs/stm32-rs-nightlies now exists and contains up-to-date builds of the stm32-rs crates |
| 02:30:09 | adamgreig | so you can depend on it from cargo.toml and you can view generated source in your browser |
| 02:41:06 |  jacobrosenthal | Well done. Stm32 team is always pushing the bounds. |
| 03:45:34 |  baloo_ | adamgreig: I'd be interested to have a discussion about https://github.com/rust-embedded/embedded-hal/issues/172 on the next meeting |
| 03:48:11 | baloo_ | I know I'm expected to add a comment for that, but I'm not sure what state the rfc has to reach to be discussed. |
| 03:48:32 | baloo_ | that's something I'm hacking around a bit myself and I'd be interested in pushing forward |
