--- Log opened Mon Jun 28 00:00:51 2021
01:55 -!- michaelfolkson2 [~michael@2a00:23c7:6080:9301:992:7ee3:6f49:438] has joined #bitcoin-rust
02:04 -!- michaelfolkson2 [~michael@2a00:23c7:6080:9301:992:7ee3:6f49:438] has quit [Quit: I am away from my computer]
05:22 -!- tibo_ [~tibo@2400:4050:2a83:7000:743a:bf27:3a73:9040] has quit [Remote host closed the connection]
05:33 -!- michaelfolkson2 [~michael@2a00:23c7:6080:9301:f025:227d:3e76:811c] has joined #bitcoin-rust
05:34 < sebx2a> Since some of us are worried about GitHub actions lock-in, this might alleviate it a bit: https://github.com/nektos/act it would be at least possible to call this on a real VM in another CI environment xD
06:56 -!- michaelfolkson2 [~michael@2a00:23c7:6080:9301:f025:227d:3e76:811c] has quit [Quit: I am away from my computer]
07:09 -!- michaelfolkson2 [~michael@2a00:23c7:6080:9301:f025:227d:3e76:811c] has joined #bitcoin-rust
07:19 -!- michaelfolkson2 [~michael@2a00:23c7:6080:9301:f025:227d:3e76:811c] has quit [Quit: I am away from my computer]
07:34 -!- michaelfolkson2 [~michael@2a00:23c7:6080:9301:f025:227d:3e76:811c] has joined #bitcoin-rust
08:29 < gene> sebx2a: yeah, I agree after looking at the lib more closely. my thinking was that anyhow is likely to be the approach that gets used in std, at least there was some conversation about it
08:30 < gene> wanted to future proof whatever gets adopted, to prevent some of the churn BlueMatt talked about with switching out error types
08:32 < sebx2a> gene: "at least there was some conversation about it" you happen to remember where? The reasoning and scope would be interesting.
08:33 < gene> sure, it was in a thread I linked about std::io in core, one moment
08:34 < sebx2a> `std::io::Error` is like that a bit already by having a `Box<dyn Error>` cause. But getting rid of the kind would be super annoying imo, unless there was a trait bound for error traits that exposes a kind.
08:35 < gene> https://github.com/rust-lang/rust/issues/48331#issuecomment-576942330
08:36 < gene> https://github.com/rust-lang-nursery/portability-wg/issues/12#issuecomment-382072568
08:37 -!- michaelfolkson2 [~michael@2a00:23c7:6080:9301:f025:227d:3e76:811c] has quit [Quit: I am away from my computer]
08:37 < BlueMatt> sebx2a: I dunno why we need to be worried about actions lock-in? We don't currently do anything actions-specific, the yml file is just a glorified shell script with a bunch of extra metadata.
08:37 < BlueMatt> I *do* think we should avoid changing that, though, of course.
08:38 < sebx2a> BlueMatt: Ah, I see, just remembered you arguing against fancy use of it
08:38 < BlueMatt> yes, I strongly disagree with "fancy" use of github actions
08:39 < BlueMatt> as long as I can roughly copy-paste the contents of the actions yml into a shell I'm happy
08:40 < gene> another possible alternative to core2: https://github.com/Kixunil/genio
08:40 < gene> 6 contributors, looks like more than the std copy-pasta that is core2
08:40 < sebx2a> I just thought being able to run it locally would move the equilibrium of too much/too little GH actions usage (rebuilding e.g. the build matrix in bash wouldn't be all that great either imo)
08:44 < gene> looks like genio is high-churn right now, though
08:45 < sebx2a> gene: yeah, genio would be the obvious choice (I think most here even know kinuxil), but at least last time I looked it seemed quite unfinished
08:45 < sebx2a> and we'd probably also want less dependencies
08:45 < gene> ugh, that bbqsrc guy is kind of offputting
08:46 < gene> closes my PR, and applies the fix himself
08:46 < gene> fucking annoying
08:48 < gene> looks like he took my commit, but manually merged it
08:48 < gene> what a weird way to do things...
08:49 < gene> maybe I was too harsh, but I don't understand doing things that way
08:50 < sebx2a> Ehm, from what I see your PR wasn't building. He had to fix it and then just merged it locally instead of pushing to your branch (quite some hassle) and going through the UI.
08:51 < sebx2a> Anyway, regarding the std error thing: I hope they don't go the anyhow path. If we want useful errors in core it can't require alloc afaik and that's incompatible with that approach.
08:52 < sebx2a> I'm also a bit confused by the nightly backtrace stuff in that regard, because that has the same problem. Why would backtraces be part of the central error trait? Can't they be an add-on?
08:54 < gene> like having two traits: Error + Backtrace?
08:54 < sebx2a> Yeah, I'd prefer that, then the Backtrace impl could be contingent on alloc
08:55 < gene> that sounds great
08:56 < sebx2a> As it is you'd probably have to impl it such that without alloc the method returns None, which is kinda weird. Would be nicer if the build breaks if one forgets to activate the right feature.
08:56 < gene> maybe there could even be a const generics impl, where you have a fixed buffer for backtraces (assuming the buffer is what requires alloc)
08:58 -!- michaelfolkson2 [~michael@2a00:23c7:6080:9301:f025:227d:3e76:811c] has joined #bitcoin-rust
08:58 < sebx2a> that sounds quite interesting too, maybe they actually intend to solve it that way and the current impl is just a MVP https://doc.rust-lang.org/src/std/backtrace.rs.html#139-143
08:59 < sebx2a> uh, otoh it seems to have nested vecs and dynmic names and stuff, maybe not
09:00 < gene> maybe not, yeah
09:01 < gene> could see doing a reimpl with const generics, and sane defaults for buffer size
09:01 < gene> probably wouldn't ever have more than 128 relevant frames in a backtrace, for example
09:02 < gene> filename could similarly be set to MAX_FILENAME_LENGTH (pseudo-limit)
09:05 < gene> would probably be easier to just do the split impl like you suggested, though
09:08 < gene> they even have the disabled feature, which you could turn into a feature-gated thing
09:09 < gene> have a mock `Vec` type for disabled platforms, so the build works
09:09 < gene> use alloc::Vec and std::vec::Vec for the others
09:16 < devrandom> I think github auto-closes a PR if it sees the commit merged via a manual push
09:21 < gene> more about how error-handling wg is actually doing no-std Error: https://github.com/rust-lang/project-error-handling/issues/20#issuecomment-738471153
09:22 < gene> devrandom: yeah, I was just confused because I had not experienced that workflow before
09:23 < gene> was to hasty in being annoyed, I understand better now
09:52 < devrandom> OK, sounds good :)
09:54 -!- Guest64 [~textual@c-73-189-219-70.hsd1.ca.comcast.net] has joined #bitcoin-rust
09:54 < sebx2a> gene: 128*MAX_FILENAME_LENGTH might be a high overhead to pay every time a result is returned …
09:54 < sebx2a> I don't fully get what the error wg is proposing, but good that they thought of it at least ^^
09:58 -!- michaelfolkson2 [~michael@2a00:23c7:6080:9301:f025:227d:3e76:811c] has quit [Quit: I am away from my computer]
10:15 -!- michaelfolkson2 [~michael@2a00:23c7:6080:9301:f025:227d:3e76:811c] has joined #bitcoin-rust
10:25 -!- michaelfolkson2 [~michael@2a00:23c7:6080:9301:f025:227d:3e76:811c] has quit [Quit: I am away from my computer]
10:54 < gene> sbx2a: true, the const generics impl is probably a bad idea
10:54 < gene> especially with dynamic sized types
10:55 < gene> the TinyBox solution seems interesting, especially if it doesn't require alloc
11:33 -!- michaelfolkson2 [~michael@2a00:23c7:6080:9301:f025:227d:3e76:811c] has joined #bitcoin-rust
11:49 -!- Guest64 [~textual@c-73-189-219-70.hsd1.ca.comcast.net] has quit [Quit: My MacBook Air has gone to sleep. ZZZzzz…]
11:59 -!- michaelfolkson2 [~michael@2a00:23c7:6080:9301:f025:227d:3e76:811c] has quit [Quit: I am away from my computer]
12:00 -!- Guest64 [~textual@c-73-189-219-70.hsd1.ca.comcast.net] has joined #bitcoin-rust
12:09 -!- michaelfolkson2 [~michael@2a00:23c7:6080:9301:f025:227d:3e76:811c] has joined #bitcoin-rust
12:21 -!- michaelfolkson2 [~michael@2a00:23c7:6080:9301:f025:227d:3e76:811c] has quit [Quit: I am away from my computer]
13:01 -!- michaelfolkson2 [~michael@2a00:23c7:6080:9301:f025:227d:3e76:811c] has joined #bitcoin-rust
13:31 -!- michaelfolkson2 [~michael@2a00:23c7:6080:9301:f025:227d:3e76:811c] has left #bitcoin-rust [Textual IRC Client: www.textualapp.com]
13:33 -!- belcher_ is now known as belcher
15:52 -!- Guest64 [~textual@c-73-189-219-70.hsd1.ca.comcast.net] has quit [Quit: My MacBook Air has gone to sleep. ZZZzzz…]
16:56 -!- andytoshi [~apoelstra@s66-183-0-205.bc.hsia.telus.net] has quit [Ping timeout: 244 seconds]
16:56 -!- andytoshi [~apoelstra@s66-183-0-205.bc.hsia.telus.net] has joined #bitcoin-rust
16:57 -!- b42 [~mmilata@b42.cz] has quit [Ping timeout: 244 seconds]
16:57 -!- tibo_ [~tibo@2400:4050:2a83:7000:74a4:d079:c39e:d07e] has joined #bitcoin-rust
16:58 -!- b42 [~mmilata@b42.cz] has joined #bitcoin-rust
17:21 -!- belcher_ [~belcher@user/belcher] has joined #bitcoin-rust
17:24 -!- belcher [~belcher@user/belcher] has quit [Ping timeout: 268 seconds]
17:28 -!- Guest64 [~textual@c-73-189-219-70.hsd1.ca.comcast.net] has joined #bitcoin-rust
17:52 -!- Guest64 [~textual@c-73-189-219-70.hsd1.ca.comcast.net] has quit [Quit: My MacBook Air has gone to sleep. ZZZzzz…]
18:03 -!- Guest64 [~textual@c-73-189-219-70.hsd1.ca.comcast.net] has joined #bitcoin-rust
18:10 -!- Guest64 [~textual@c-73-189-219-70.hsd1.ca.comcast.net] has quit [Quit: My MacBook Air has gone to sleep. ZZZzzz…]
23:24 -!- Guest64 [~textual@c-73-189-219-70.hsd1.ca.comcast.net] has joined #bitcoin-rust
23:29 -!- Guest64 [~textual@c-73-189-219-70.hsd1.ca.comcast.net] has quit [Quit: My MacBook Air has gone to sleep. ZZZzzz…]
--- Log closed Tue Jun 29 00:00:51 2021