Skip to content

add last meeting minutes and logs #254

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
merged 1 commit into from
Nov 13, 2018
Merged

add last meeting minutes and logs #254

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
238 changes: 238 additions & 0 deletions minutes/2018-11-13.irc.log
View file
Open in desktop
Original file line number Diff line number Diff line change
@@ -0,0 +1,238 @@
2018-11-13 20:05:46 ~japaric hello everyone; let's start this meeting
2018-11-13 20:05:57 jamesmunns hello!
2018-11-13 20:06:28 ~japaric first, I'd like to discuss the procedure for making a roadmap for next year
2018-11-13 20:06:41 ~japaric I have a proposal that I have written in the dropbox paper
2018-11-13 20:06:51 @therealprof jamesmunns: I think it was Adam Green (the linkchecker)
2018-11-13 20:06:54 ~japaric the idea is to get input from the community using an informal survey
2018-11-13 20:07:37 ~japaric and then make a draft roadmap RFC by the end of January using that data and the input of all the teams
2018-11-13 20:07:51 @therealprof +1
2018-11-13 20:08:02 ~japaric then we can discuss feasibility of the roadmap with the Rust teams during RAH
2018-11-13 20:08:24 ~japaric at least for the parts that involve language / compiler work
2018-11-13 20:08:24 jamesmunns +1
2018-11-13 20:08:44 @korken89 +1
2018-11-13 20:09:54 ~japaric I expect that the Rust team will launch the call for 2019 blog posts later this year
2018-11-13 20:10:01 ~japaric we should incorporate those in the roadmap as well
2018-11-13 20:11:01 ~japaric ok, we don't have too many people today to vote but I think we start with the informal survey
2018-11-13 20:11:15 ~japaric we can start**
2018-11-13 20:11:56 jamesmunns (thanks therealprof)
2018-11-13 20:12:30 @adamgreig right, sorry, I'm here
2018-11-13 20:12:42 * cr1901 is present/listening as well
2018-11-13 20:13:18 jamesmunns japaric: are you asking for more informal wishlist items now?
2018-11-13 20:13:27 jamesmunns (making sure you aren't blocking on input)
2018-11-13 20:13:44 ~japaric no, not right now
2018-11-13 20:13:51 ~japaric we should also collect list of current TODO items from all the teams somewhere to have a high level view of the ongoing work
2018-11-13 20:14:19 ~japaric but I can come up with some dropbox paper structure for that by next week
2018-11-13 20:14:23 jamesmunns Also you mentioned blog posts, if any embedded people dont have their own blog, but want to post, guest posts on the embedded blog are totally welcome IMO
2018-11-13 20:14:33 jamesmunns (for the year end call for posts)
2018-11-13 20:14:37 @adamgreig +1 on starting to get a 2019 wishlist together
2018-11-13 20:14:39 jamesmunns (or any time, really)
2018-11-13 20:14:44 @adamgreig I think it would be easy to lose momentum once 2018 lands
2018-11-13 20:14:55 adamgr therealprof: I think japaric added the linkchecker stuff to CI. I just recently told it to ignore the print.html that mdbook is corrupting.
2018-11-13 20:14:57 ~japaric jamesmunns: guess posts sgtm
2018-11-13 20:15:34 ~japaric jamesmunns: we should advertise that on the newletters, in general and make it more promimently once the call for 2019 blog posts is out
2018-11-13 20:15:43 ~japaric that = guess posts are welcomed
2018-11-13 20:16:09 jamesmunns also welcome: posts from the "chip ecosystem islands", like stm32-rs, etc. I plan to add a post there talking about the re-release of the nrf52 ecosystem for 2018 edition (explaining all the crate renames, etc)
2018-11-13 20:17:10 @adamgreig on my stack is forming stm32-rs as an org rather than a repo and then it can have a nice introductory blog post
2018-11-13 20:17:31 ~japaric ok, I'll create the 2019 wishlist issue after this meeting
2018-11-13 20:17:35 @therealprof adamgreig: Yay!
2018-11-13 20:17:36 @adamgreig +1
2018-11-13 20:17:41 @korken89 that is really good, will show which are active as well
2018-11-13 20:17:49 cr1901 ^msp430 still needs some work before I'd be comfortable making blog posts
2018-11-13 20:17:52 @korken89 adamgreig: Awesome!
2018-11-13 20:18:18 jamesmunns adamgreig: +11
2018-11-13 20:18:21 @adamgreig glad everyone is very enthusiastic about helping :P
2018-11-13 20:18:55 ~japaric cr1901: (before I forget: I tested building rust-std (core & alloc) for msp430 with llvm-assertions enabled the other day and it worked so it may be possible to revive pftbest's old PR to enable builds of rust-std for msp430)
2018-11-13 20:19:34 cr1901 Yell at me if I haven't taken a look within the week
2018-11-13 20:20:07 cr1901 I've been juggling a number of things, fairly badly (what else is new?)
2018-11-13 20:20:17 jamesmunns cr1901: you're in good company, I feel
2018-11-13 20:20:32 cr1901 Good to know
2018-11-13 20:20:44 ~japaric ok, let's move to the next item in the agenda
2018-11-13 20:20:50 ~japaric 2018 edition finish line
2018-11-13 20:21:00 ~japaric I made a PR with the blog post I mentioned last week
2018-11-13 20:21:07 ~japaric https://github.com/rust-embedded/blog/pull/25
2018-11-13 20:21:43 ~japaric if you haven't reviewed it yet please give it a look
2018-11-13 20:21:56 ~japaric jamesmunns just opened a PR for the last blocker
2018-11-13 20:21:56 jamesmunns re: static guarantees, I moved type-state into static guarantees, but we need a paragraph or two explaining static guarantees in general for the "landing page"
2018-11-13 20:22:03 jamesmunns (it was already merged)
2018-11-13 20:22:19 jamesmunns and adamgreig opened a PR populating `tips for C/C++ developers`
2018-11-13 20:22:26 jamesmunns so that won't be empty for much longer
2018-11-13 20:22:47 @korken89 I feel quite out of touch with current issues after being gone for a while, I will try to jump in wherever I can help for now
2018-11-13 20:24:10 ~japaric @resources team: how do you feel about including the '"does this compile on stable?" community effort' in the blog post?
2018-11-13 20:24:33 @adamgreig is the question about doing it in the first place or including it in the blog post?
2018-11-13 20:24:34 @therealprof I mentioned my concern in the PR.
2018-11-13 20:24:40 jamesmunns japaric: maybe that can be a post-2018 goal?
2018-11-13 20:24:41 ~japaric adamgreig: both, actually
2018-11-13 20:24:55 @adamgreig I like the idea of having it
2018-11-13 20:24:57 ~japaric jamesmunns: yeah, we can start it by the next newsletter
2018-11-13 20:25:10 @adamgreig but could be convinced either way re blog post
2018-11-13 20:26:32 ~japaric the other question is when do y'all want to put up this blog post?
2018-11-13 20:26:57 ~japaric original ETA was tomorrow but I'm not sure if enough people have reviewed it yet
2018-11-13 20:27:38 @therealprof Quite a few people already chimed in, no?
2018-11-13 20:27:39 @adamgreig I don't think it's too controversial; maybe anotehr resources person could r+ it?
2018-11-13 20:27:59 jamesmunns will read shortly, was more focused on the checklist than the blog post
2018-11-13 20:28:06 @adamgreig well except for the issue on the stable awesome list thing
2018-11-13 20:28:14 ~japaric and we should include the 2019 wishlist in this blog post so the last part of it will have to be updated
2018-11-13 20:28:42 @adamgreig i propose we include that 2019 wishlist issue but do not include the stable awesome list
2018-11-13 20:28:47 @adamgreig and instead announce that once 2018 is actually landed
2018-11-13 20:29:00 ~japaric I can do that after this meeting (re 2019 wishlist modification)
2018-11-13 20:29:03 ~japaric adamgreig: sgtm
2018-11-13 20:29:05 @adamgreig we cna open the issue on awesome-embedded-rust and maybe start a few crates or something
2018-11-13 20:29:10 jamesmunns adamgreig: +1
2018-11-13 20:29:13 @korken89 +1
2018-11-13 20:29:17 @adamgreig but a lot of things just won't build on stable yet anyway
2018-11-13 20:29:21 @adamgreig on beta, maybe.. :p
2018-11-13 20:29:32 @therealprof Sounds great. Either way I r+d it.
2018-11-13 20:29:38 @adamgreig so this way we can make it a "rust stable is here! at last! embedded all the things! let's build a list of what now works on stable"
2018-11-13 20:29:45 @adamgreig and it will give us something to do immediately post 2018 :p
2018-11-13 20:30:06 ~japaric adamgreig: I like how that sounds
2018-11-13 20:30:10 @therealprof adamgreig: I have the feeling a lot of stuff will compile on stable.
2018-11-13 20:30:22 @adamgreig therealprof: well volunteered ;)
2018-11-13 20:30:27 @adamgreig I mean stable right now, i.e. without const fn
2018-11-13 20:30:43 ~japaric if something compiles on beta we can tag it as "compiles on 1.31"
2018-11-13 20:30:48 @therealprof Drivers don't usually need it.
2018-11-13 20:30:57 @adamgreig but even then a lot of things still have vestigial #[feature] things around
2018-11-13 20:31:02 @therealprof It's applications that suck without.
2018-11-13 20:31:18 @therealprof True.
2018-11-13 20:31:38 ~japaric we are at the half hour mark; let's not go too off topic
2018-11-13 20:31:49 @adamgreig can we resolve to open the issue on awesome-embedded but not put it in the blog post, then?
2018-11-13 20:31:54 ~japaric ok then let's include the wishlist in the blog post and release it tomorrow
2018-11-13 20:31:56 @adamgreig and we'll blog about it later, after 2018 lands
2018-11-13 20:31:59 @adamgreig great
2018-11-13 20:32:19 ~japaric any volunteer to open the issue in the awesome list repo?
2018-11-13 20:32:38 @adamgreig will do
2018-11-13 20:32:43 ~japaric adamgreig: thanks
2018-11-13 20:32:54 @therealprof … and make sure it happens, too? :-p
2018-11-13 20:33:01 @adamgreig I will open the issue :p
2018-11-13 20:33:26 @therealprof Always rooting for the easy stuff, eh? ;)
2018-11-13 20:34:01 ~japaric ok, let's do a quick past over the remaining book issues
2018-11-13 20:34:16 ~japaric issue #2: getting started chapter
2018-11-13 20:34:21 ~japaric the semihosting section landed
2018-11-13 20:34:27 ~japaric and the registers section was updated
2018-11-13 20:34:33 ~japaric the only thing left to cover is the #[interrupt] macro
2018-11-13 20:34:46 cr1901 There's an interrupt macro ._.?
2018-11-13 20:34:50 @therealprof Has it landed?
2018-11-13 20:34:59 @adamgreig lol
2018-11-13 20:35:30 @adamgreig https://github.com/rust-embedded/cortex-m-rt/blob/master/CHANGELOG.md#v065---2018-10-23
2018-11-13 20:35:39 @therealprof cr1901: There's always been one. But now it's shiny and new and useful.
2018-11-13 20:35:52 @adamgreig it's now an attribute instead of a macro-by-example
2018-11-13 20:36:05 cr1901 I don't understand the distinction
2018-11-13 20:36:07 @adamgreig and it's in cortex-m-rt instead of being made by svd2rust
2018-11-13 20:36:19 cr1901 actually, I'll ask later
2018-11-13 20:36:19 @adamgreig previously you wrote interrupt!(tim2, TIM2); fn tim2() { ... }
2018-11-13 20:36:26 @adamgreig now you write #[interrupt] fn TIM2() { ... }
2018-11-13 20:36:40 @therealprof Is there an issue for it?
2018-11-13 20:36:42 @adamgreig (and it does some lovely things with static muts)
2018-11-13 20:36:43 cr1901 huh...
2018-11-13 20:36:51 @therealprof I can take it.
2018-11-13 20:37:11 ~japaric re #[interrupt] macro we can add a section now using the hand written lm3s6965 crate, or add it after svd2rust is released and the stm32f30x crate is updated
2018-11-13 20:37:34 ~japaric any preference? and any volunteer?
2018-11-13 20:37:47 @therealprof Yeah, I thought we need a new svd2rust version.
2018-11-13 20:38:14 @therealprof japaric: Didn't quite catch what you meant re lm3s6965, but I can take it.
2018-11-13 20:38:33 ~japaric therealprof: https://github.com/japaric/lm3s6965
2018-11-13 20:38:47 ~japaric "device crate" but only has interrupt stuff in it; no registers
2018-11-13 20:39:00 ~japaric wrote it to test cortex-m-rtfm
2018-11-13 20:39:00 jamesmunns (I'd like to discuss svd2rust at the end of the meeting, re: https://github.com/rust-embedded/svd2rust/pull/254)
2018-11-13 20:39:03 @therealprof Hm…
2018-11-13 20:39:11 @therealprof Not my forte.
2018-11-13 20:39:38 @therealprof Prefer working with real hardware.
2018-11-13 20:39:39 ~japaric therealprof: then you volunteer to write it after svd2rust v0.14 is out?
2018-11-13 20:39:50 @adamgreig no reason not to write it in advance of svd2rust
2018-11-13 20:39:51 @therealprof japaric: Yes.
2018-11-13 20:39:54 @adamgreig we know exactly what it'l look like
2018-11-13 20:40:08 @therealprof adamgreig: I agree.
2018-11-13 20:40:12 ~japaric therealprof: thanks
2018-11-13 20:40:41 ~japaric moving on, issue #5 HAL
2018-11-13 20:40:46 ~japaric therealprof: jamesmunns any update?
2018-11-13 20:41:10 jamesmunns none from my side
2018-11-13 20:41:21 @therealprof Yeah, forgot about it.
2018-11-13 20:41:29 @therealprof Will do some writing tonight I guess.
2018-11-13 20:42:13 jamesmunns I probably wont have time in the upcoming week to write, but will gladly review
2018-11-13 20:42:14 ~japaric therealprof: +1
2018-11-13 20:42:22 ~japaric jamesmunns: +1
2018-11-13 20:42:48 @adamgreig https://github.com/rust-embedded/awesome-embedded-rust/issues/115 is open
2018-11-13 20:42:52 ~japaric ok, next issue, #6, is static guarantees
2018-11-13 20:43:04 ~japaric type state was moved under it
2018-11-13 20:43:14 ~japaric we have to write some text for the static guarantees landing page
2018-11-13 20:43:42 ~japaric (if I understood jamesmunns correctly)
2018-11-13 20:43:48 jamesmunns Really just 2-3 paragraphs intro'ing all the cool stuff that can be checked at compile time would be perfect
2018-11-13 20:43:52 jamesmunns help very much wanted
2018-11-13 20:44:04 jamesmunns japaric: yup exactly
2018-11-13 20:44:35 ~japaric jamesmunns: I can give it a go, I guess
2018-11-13 20:44:57 jamesmunns japaric: +1
2018-11-13 20:45:28 jamesmunns Doesn't have to be comprehensive, just an intro. I think the intro will help others write larger subsections, like type-state-programming
2018-11-13 20:45:49 ~japaric next issue #7 tips for C/C++ programmers
2018-11-13 20:45:56 ~japaric adamgreig: I saw a PR
2018-11-13 20:46:15 jamesmunns https://github.com/rust-embedded/book/pull/81
2018-11-13 20:47:15 @therealprof I'll do a pass later.
2018-11-13 20:47:23 @adamgreig yep, it's up at last
2018-11-13 20:47:46 @adamgreig this is rougher than the concurrency pr was, it will need a bit more work
2018-11-13 20:47:56 ~japaric adamgreig is looking for feedback so please give it a read and comment
2018-11-13 20:47:58 @adamgreig (and I already remembered the repr() stuff from last meeting needed to add on)
2018-11-13 20:48:13 ~japaric adamgreig: thanks for working on this
2018-11-13 20:49:03 ~japaric issue #9, book outline: done; the PR just landed
2018-11-13 20:49:42 ~japaric issue #10, the debugonomicon; ryankurte is not around so we'll check next week
2018-11-13 20:49:50 ~japaric but the book was renamed
2018-11-13 20:50:35 ~japaric last 10 minutes is open floor for updates
2018-11-13 20:50:42 ~japaric jamesmunns: you wanted to discuss a svd2rust PR
2018-11-13 20:50:59 jamesmunns Yup: https://github.com/rust-embedded/svd2rust/pull/254
2018-11-13 20:51:19 jamesmunns But this may tie in to the post-2018 discussion
2018-11-13 20:51:25 jamesmunns More abstractly:
2018-11-13 20:51:43 jamesmunns "How do we see ourselves developing and maintaining embedded crates in 2019"
2018-11-13 20:51:44 cr1901 japaric: Now that msp430 is working again, I can looking into AT2XT for the showcase. However, panic_implementation was removed recently, so I need to fix that.
2018-11-13 20:52:05 ~japaric cr1901: $ sed -i s/panic_implementation/panic_handler
2018-11-13 20:52:08 jamesmunns This PR introduces feature gates for each peripheral generated by svd2rust, which speeds compile time
2018-11-13 20:52:13 cr1901 yes, sorry
2018-11-13 20:52:21 @adamgreig cr1901: no, that's the fix
2018-11-13 20:52:37 jamesmunns however, therealprof has pointed out that this can make it harder to abstract over -pac crates from higher levels
2018-11-13 20:52:54 cr1901 adamgreig: Oh, whoops
2018-11-13 20:53:22 jamesmunns If anyone has feelings on that, feel free to comment on that PR
2018-11-13 20:53:38 jamesmunns This is a breaking change, so it either needs to go in now, or wait for 0.15.0.
2018-11-13 20:53:50 jamesmunns (or never)
2018-11-13 20:54:15 cr1901 Yea, that would be the msp430-rt crate, and pftbest has been busy (and will continue to be busy) with meatspace AFAIK
2018-11-13 20:54:42 jamesmunns But in general, I think we should sit down as a WG/community, and come up with a more structured plan of how we /want/ the ecosystem to look. I should probably write a blog post about this
2018-11-13 20:54:51 @adamgreig yes please
2018-11-13 20:55:01 @adamgreig I think that's really what's underlying a lot of these issues
2018-11-13 20:55:02 @therealprof +1
2018-11-13 20:55:10 jamesmunns There are a couple really good efforts, stm32-rs in particular that are trying new things
2018-11-13 20:55:28 jamesmunns But I don't think we have a really good picture how things work "at scale"
2018-11-13 20:55:30 @adamgreig if you want "trying new things" I have a very different stm32 project coming soon ;)
2018-11-13 20:55:45 @adamgreig but yea
2018-11-13 20:55:58 @adamgreig we're starting to build up a picture of pacs -> hals -> bsps, embedded-hal, and drivers
2018-11-13 20:56:21 @adamgreig I think the features-per-periph in svd2rust is contentious because it's adding complexity to try and escape long compile times
2018-11-13 20:56:32 @adamgreig but in a way that's not necessarily what we want long-term
2018-11-13 20:56:33 ~japaric jamesmunns: seems like this PR is going to require quite some discussion to make a decision so I'd like to propose to postpone it for v0.15 (or never depending on the outcome)
2018-11-13 20:56:57 jamesmunns japaric: yeah, I felt like that might be the decision.
2018-11-13 20:57:18 @adamgreig it's kind of a shame because reducing compile times would be really nice for a lot of users very quickly
2018-11-13 20:57:33 @adamgreig but we maybe need to resolve the philosophical argument around it first
2018-11-13 20:57:50 @adamgreig the svd2rust-generated crate is always by far the longest compile in any embedded apps i'm building
2018-11-13 20:57:51 @therealprof +1
2018-11-13 20:58:02 jamesmunns Yeah, its a big win for some people, myself in particular (takes my 8 minute build down to 30-40 seconds for one crate)
2018-11-13 20:58:03 @therealprof … and CI.
2018-11-13 20:58:40 jamesmunns but this might be a patch/hack, instead of a needed architectural change
2018-11-13 20:58:40 @therealprof However whether it's a win at all depends on whether you'd like to opt for specific peripherals.
2018-11-13 20:58:50 cr1901 I thought the idea was SVD crates were meant to be fairly static
2018-11-13 20:59:06 @therealprof I think that a lot of people (me included) would be rather on the "give me all" side of things.
2018-11-13 20:59:09 jamesmunns cr1901: things like changing build profiles can force a rebuild
2018-11-13 20:59:19 @adamgreig and every ci run
2018-11-13 20:59:21 jamesmunns e.g. going from opt level 0 to 1 or back
2018-11-13 20:59:27 jamesmunns and CI runs, which are always clean builds
2018-11-13 20:59:28 cr1901 hmmm
2018-11-13 20:59:40 * cr1901 hasn't run into this problem yet
2018-11-13 20:59:44 jamesmunns its papercuts, for sure
2018-11-13 20:59:47 @adamgreig possibly smaller device crates?
2018-11-13 20:59:52 @adamgreig it scales poorly with large mcus
2018-11-13 20:59:53 jamesmunns yeah
2018-11-13 21:00:01 @adamgreig well I mean it scales linearly but is still annoying
2018-11-13 21:00:02 jamesmunns the fact that its a single compilation unit is really rough
2018-11-13 21:00:10 @adamgreig hah that too
2018-11-13 21:00:13 ~japaric ok, we are almost out of time so I'll make a PR to release v0.14.0 (w/o jamesmunns PR) and we can vote on it using reviews
2018-11-13 21:00:25 @adamgreig 17 other cpu threads sitting idle while one desperately reads megabytes of rust
2018-11-13 21:00:28 @therealprof There's a heavily discussed Rust RFC about introducing bitfields.
2018-11-13 21:00:39 cr1901 adamgreig: I mean, msp430 micros aren't exactly large
2018-11-13 21:00:44 jamesmunns therealprof: oh really? link please?
2018-11-13 21:00:48 @adamgreig cr1901: yea, which is probably why you've not run into it so much
2018-11-13 21:00:53 @adamgreig on an stm32f4 crate it might take 1-2min to build
2018-11-13 21:00:55 * cr1901 has a visceral reaction to bitfields
2018-11-13 21:00:57 cr1901 hmmm
2018-11-13 21:01:15 @therealprof :-D
2018-11-13 21:01:20 ~japaric ok, I'm going to oficially end this meeting. Thanks everyone for attending!
Loading