01:14 <rmsyn[m]> would anyone in the rust-embedded working group be interested in setting up an org to get membership in various vendor programs to get access to documentation (e.g. Synopsys, MIPI, etc.)?

01:15 <rmsyn[m]> then, HAL and driver authors could request membership (possibly with a donation to hellp cover org membership costs)

01:16 <rmsyn[m]> I would be more than happy to contribute funds to an effort like that

10:20 <holo[m]> <holo[m]> "Hey im trying to use enums to..." <- > <@holo:matrix.org> Hey im trying to use enums to handle multiple actions on my pico w. On pico W it looks its sendting everything like it should https://pastebin.com/cQiv2yY7 and this is handle by this code: https://github.com/holotrack/switch-embassy-rs/blob/main/... (full message at <https://catircservices.org/_matrix/media/v3/download/catircservices.org/bcoAagWQyZgYQBKcuIKyoecm>)

11:08 <therealprof[m]> <rmsyn[m]> "would anyone in the rust-..." <- Hm, sounds tricky and expensive. Would have to review all licenses (probably with a lawyer) and check whether that kind of sharing is allowed. Also, who decides which standards to sign up for?

11:09 <therealprof[m]> therealprof[m]: Seems a lot cheaper and straight forward to individually apply or go through the grey channels to get access to those documents.

11:43 <whitequark[cis]> i do have a bunch of docs available, hmu if you have want a pdf

11:44 <whitequark[cis]> i've got some DesignWare pdfs that i've never seen circulating in the west, for their basic IP (SPI, DMA, etc)

17:22 <rmsyn[m]> therealprof: have tried to contact a few vendors, most seem to want some form of organization to interact with their sales people. appears a lot harder for an individual to get any traction with these people. my thoughts were that if we setup something that is not exactly open-access (to appease the confidentiality agreements), we could have something with a little more accessibility to FOSS developers

17:23 <rmsyn[m]> if we spread the cost among a large group of us, and everyone in the group gains access to the org's docs, we could greatly expand access to documentation (without breaking whatever IP vendors want to protect)

17:24 <dngrs[m]> rmsyn[m]: Have you tried pretending you're a company? As long as they don't require a DUNS number or something I've had good experiences just making shit up

17:24 <dngrs[m]> Might need some generic website but hey there's templates for that

17:25 <JamesMunns[m]> also, many of those vendors that expect you to sign an NDA have terms like "only the people in your company and/or those under similarly strict NDAs"

17:25 <rmsyn[m]> as far as what vendors to get access to, that could be decided by the members of the group. hedging toward the big players might be a good first bet

17:25 <JamesMunns[m]> I dunno if I can cosign "do a fraud" wrt pretending, so no comment on that one :)

17:26 <rmsyn[m]> jamesmunns: that's what I mean, instead of individuals signing those NDAs, whatever org we setup would be the "company"

17:26 <dngrs[m]> I'm obviously not endorsing anything either, especially no accidental anonymous data leaks

17:26 <rmsyn[m]> we could decide on a code-of-conduct that covers: if you break the NDA, you're out

17:27 <rmsyn[m]> personally, NDAs are kind of a gross-but-necessary-evil type thing

17:27 <JamesMunns[m]> yeah... idk. I actually run a company, and have had experience (and lawyers assistance with reading docs), I don't think things work the way you might assume

17:27 <JamesMunns[m]> breaking of NDAs usually comes with liabilities and damages, and probably they practically won't care, but I'm not gunna put my name down for someone else in that regard.

17:28 <rmsyn[m]> sure, not a lawyer, and corporate IP law is probably some of the most arcane stuff out there. just looking for a solution to what seems like an unnecessarily opaque doc situation

17:29 <JamesMunns[m]> IMO using grey sourced docs is generally better - you can't be held to the terms of an NDA you didn't sign

17:29 <dngrs[m]> I see it this way: even the C standard is officially not available for free, but you just so happen to find it easily on the web

17:29 <JamesMunns[m]> but if you DID sign something, AND you publicly spread those docs, you're probably liable for it. Again, I'd say probably most companies won't care, but I don't want to find out which ones do.

17:30 <dngrs[m]> Yeah before you sign an NDA pay very close attention to the damages part

17:30 <dngrs[m]> And probably put an LLC between you and the damages

17:31 <rmsyn[m]> right, I wouldn't want anyone exposed to unnecessary liability. kind of wish they would release a version of the docs "here are the registers, here's how they interact. btw no netlists"

17:32 <JamesMunns[m]> dngrs[m]: also lol at american NDAs which always state "if you leak one sentence of this doc you have done irreparable harm that cannot be cured with money alone"

17:32 <dngrs[m]> I've signed some nasty shit in the past (most egregious was a few millions I think?) because in Germany you're typically protected from that unless it's specifically gross negligence or active malice but your local jurisdiction might digger

17:32 <dngrs[m]> s/digger/differ/

17:34 <FlixtheNewbie[m]> JamesMunns[m]: The American work culture is ridiculous, TBH. I've been laid off several times, and it's always been normal. I got laid off from an American company a few days ago, and all my accesses were cut off immediatly, I couldn't speak to anyone, etc. for “security reason”. I asked if I was being fired because I did something wrong, but no, it's just their regular protocol.

17:34 <dngrs[m]> Yeah it's nuts

17:37 <rmsyn[m]> the situation just seems silly, when there are ARM peripherals with full docs, freely available. then Synopsys wants a blood pact to get any DesignWare docs

17:40 <dngrs[m]> Yet another engineering problem ... just social 😬

17:42 <dngrs[m]> I find it a bit annoying there seems to be no off the shelf "general gantry" product where you'd attach an extruder hot end, CNC motor, pen plotter, laser etc

17:42 <dngrs[m]> I mean there's specific requirements for each of those (enclosure, fume extraction, build plate etc) so I get it

17:42 <dngrs[m]> But it seems not totally unsolvable either

17:43 <dngrs[m]> Sorry wrong room

17:53 <therealprof[m]> @rmsyn:matrix.org Yeah, but you can essentially think of us as "being a random bunch of guys who want to do embedded stuff in Rust", while there are a few employees here working on that stuff during business hours, for many it's more of a hobby. Those documents and the agreements behind them are a huge can of worms and doing this in any meaningful way would require a person dedicating a serious amount of time into that plus

17:53 <therealprof[m]> lawyers and whatnot (so immense monetary resources required) and even then it's not guaranteed we'd get the results you're hoping for.

17:54 <therealprof[m]> As alluded to here before. If you need something, just ask around and there's a good chance a document will fall of a truck and might land right in front of your feet. ;)

17:55 <rmsyn[m]> :) may all doc carrying trucks lose their cargo

17:56 <therealprof[m]> If you scroll up 6h, you might find a truck...

17:58 <rmsyn[m]> I appreciate the immense resources involved with interfacing with some of these vendors, which is why a sort of collective way to distribute the costs would be awesome. not just confined to rust-embedded folks, but something that includes the larger FOSS ecosystem. however much of a long off hope that might be...

17:58 <JamesMunns[m]> Yeah, tbh their whole business model is they don't want collective public action, sadly.

18:04 <rmsyn[m]> would be nice to design/have a SoC/SoM with all open IP (opencores, ARM, etc.) and RISC-V cores

18:05 <rmsyn[m]> currently to dumb at circuit design to pull this off myself, may be a good next project though :p

18:57 <whitequark[cis]> we're doing something that might help you over in the Amaranth community

18:58 <whitequark[cis]> an RFC just landed with a CSR API specifically designed with Rust and its ownership model in mind

18:58 <whitequark[cis]> (most SoCs don't really bother thinking about concurrency requirements for peripherals, sadly)

19:01 <whitequark[cis]> it's usually fine but sometimes you really want to split your UART in half for example, and Amaranth SoC framework has guidelines for what to do if you want that to work. AFAIK no one else is doing this sort of work

19:02 <whitequark[cis]> it's really early days though

19:13 <rmsyn[m]> whitequark: that sounds amazing, will check it out!