--- Log opened Wed Apr 29 00:00:06 2020
00:30 -!- jonatack_ [~jon@37.165.16.42] has joined #rust-bitcoin
00:34 -!- jonatack [~jon@37.165.71.66] has quit [Ping timeout: 264 seconds]
00:59 -!- ghost43 [~daer@gateway/tor-sasl/daer] has quit [Remote host closed the connection]
01:00 -!- ghost43 [~daer@gateway/tor-sasl/daer] has joined #rust-bitcoin
01:06 -!- midnight [~midnight@unaffiliated/midnightmagic] has quit [Ping timeout: 252 seconds]
01:15 -!- midnight [~midnight@unaffiliated/midnightmagic] has joined #rust-bitcoin
02:50 -!- mauz555 [~mauz555@2a01:e0a:56d:9090:b94c:baeb:34ca:6eb6] has joined #rust-bitcoin
03:24 -!- ccdle12_ [~ccdle12@cpc139350-aztw33-2-0-cust310.18-1.cable.virginm.net] has joined #rust-bitcoin
03:24 -!- ccdle12 [~ccdle12@cpc139350-aztw33-2-0-cust310.18-1.cable.virginm.net] has joined #rust-bitcoin
03:25 -!- ccdle12_ [~ccdle12@cpc139350-aztw33-2-0-cust310.18-1.cable.virginm.net] has quit [Remote host closed the connection]
04:45 -!- jonatack_ [~jon@37.165.16.42] has quit [Read error: Connection reset by peer]
04:46 -!- jonatack [~jon@213.152.161.25] has joined #rust-bitcoin
04:47 -!- jonatack [~jon@213.152.161.25] has quit [Read error: Connection reset by peer]
04:53 -!- jonatack [~jon@37.165.16.42] has joined #rust-bitcoin
04:58 -!- jonatack [~jon@37.165.16.42] has quit [Ping timeout: 256 seconds]
04:59 -!- jonatack [~jon@109.232.227.149] has joined #rust-bitcoin
05:58 -!- ccdle12 [~ccdle12@cpc139350-aztw33-2-0-cust310.18-1.cable.virginm.net] has quit [Quit: WeeChat 1.9.1]
08:20 < wumpus> elichai2: that's definitely interesting to me, the device I'm using rust-libsecp on has 8MB of memory in total
08:21 < elichai2> :) So after some of the currently open PRs in libsecp will get merged I'll push to update it in rust-secp
09:37 -!- michaelfolkson [~textual@2a00:23c5:be01:b201:2576:934:d62b:3df0] has joined #rust-bitcoin
09:53 < BlueMatt> ariard: yea, I'll probably do that.
09:53 < BlueMatt> unless you have any objections
09:53 < BlueMatt> tbh the total diff isnt all that much
11:02 < BlueMatt> ariard: do you want to try to do a few rounds back and forth on 610 and see if we can land it today/tomorrow or should we do the rustfmt first and make you rebase it? :p
11:02 -!- DeanGuss [~dean@gateway/tor-sasl/deanguss] has joined #rust-bitcoin
11:07 < BlueMatt> elichai2: any furthe comments/takes on https://github.com/rust-bitcoin/rust-secp256k1/pull/208 ?
11:07 < BlueMatt> (or andytoshi?) would be nice to merge it :)
11:10 < elichai2> I couldn't get any better answer than "as far as I know, no work has been done on C interop with wasm32-unknown-unknown"
11:10 < elichai2> but I guess if it works then I'm fine with it
11:13 < BlueMatt> I mean it does run the tests on travis....
11:13 < BlueMatt> so, thats at least somewhat compelling
11:13 < BlueMatt> but, I mean, there *is* work on interop in the sense that you can compile C to wasm32-unknown-unknown in clang
11:14 < BlueMatt> like, thats a supported thing, no?
11:14 < elichai2> BlueMatt: just a small comment on "it would be super, duper strange for clang-LLVM to have different int sizes from rust-LLVM for the same target" even though it's not the same because __int128 isn't standard but an extension, it doesn't match i128/u128 and they have different ABIs (https://github.com/rust-lang/unsafe-code-guidelines/issues/119)
11:14 < BlueMatt> hmm, interesting, right
11:14 < elichai2> BlueMatt: right. you can compile C into the target and Rust into the target, but I'm not sure anyone checked if the ABI match
11:14 < BlueMatt> but, yea, ok, not officially part of the C abi stuff
11:14 < BlueMatt> right
11:15 < BlueMatt> i mean I guess I'm ok with it if travis runs the tests and they pass?
11:16 < elichai2> for example the `libc` crate is empty under the wasm32-unknown-unknown
11:16 < elichai2> Yeah I agree about tests passing. it's not that everyone is going to use this
11:17 < BlueMatt> right, yea, I mean fair enough....still, the LLVM IR for such things should be well-defined?
11:17 < BlueMatt> i mean hopefully its something that gets resolved with the standardization of wasi and such, no
11:17 < BlueMatt> cause, like, at that point wasm32 has a pretty well-defined abi in terms of int sizes
11:18 < BlueMatt> cause almost the only thing we care about here is int/pointer size
11:18 < elichai2> and alignment of types
11:18 < BlueMatt> right
11:18 < BlueMatt> well, int alignment
11:22 < elichai2> just noticed you're passing `CC=clang-9` does it not work on older clangs? what about newer ones?
11:22 -!- DeanGuss [~dean@gateway/tor-sasl/deanguss] has quit [Remote host closed the connection]
11:22 -!- DeanGuss [~dean@gateway/tor-sasl/deanguss] has joined #rust-bitcoin
11:23 < BlueMatt> pre-clang-8 doesnt support wasm32, I thought?
11:23 < andytoshi> BlueMatt: oh i missed this
11:23 < BlueMatt> see linked issue?
11:23 < andytoshi> reading the history now
11:25 < andytoshi> i don't understand the stuation well enough
11:25 < andytoshi> i'll merge it if elichai2 says it's ok
11:25 < andytoshi> i don't see anything wrong with it
11:25 < BlueMatt> heh, alright.
11:30 < elichai2> can we somehow sanity check the size+alignment of int? like add a static called `static unsigned char INT_SIZE = sizeof(int)` same for alignment and for pointer sizes
11:31 < elichai2> maybe add a function to the fake sysroot that return those and assert that in the wasm bindgen test
11:31 < elichai2> even maybe run that test in context creation under wasm32
11:32 < elichai2> (either via function call or via extern static)
11:37 < elichai2> BlueMatt what do you say?
11:40 < BlueMatt> elichai2: yea, if you see an easy way to do that we should def do it
11:40 < BlueMatt> check alignment and size of void*, size_t, int, etc
11:41 < elichai2> I'll try to quickly prototype something and I'll see how it looks :)
11:46 < BlueMatt> kool
12:03 < elichai2> BlueMatt: We can assume unsigned char is always size 1 align 1, right?
12:03 < BlueMatt> uhhhhhhhhhhh
12:03 < BlueMatt> i mean might as well test it?
12:03 < BlueMatt> but, I'd think so?
12:04 < andytoshi> the C standard requires it be size 1
12:05 < BlueMatt> but align 1?
12:05 < BlueMatt> i kinda assume so cause otherwise every char string would break
12:05 < BlueMatt> but.....
12:17 < elichai2> BlueMatt: https://github.com/TheBlueMatt/rust-secp256k1/compare/202-04-wasm...elichai:2020-04-sanity_wasm
12:17 -!- DeanGuss [~dean@gateway/tor-sasl/deanguss] has quit [Ping timeout: 240 seconds]
12:17 < elichai2> practically this is it: https://github.com/elichai/rust-secp256k1/commit/04db9ea67d342ffd1f47c7646f6c05a373de84b4
12:17 < elichai2> there's another commit that introduces a `size_t` type
12:19 < elichai2> andytoshi: feel free to comment on that test too :)
12:20 -!- DeanGuss [~dean@gateway/tor-sasl/deanguss] has joined #rust-bitcoin
12:22 < elichai2> this is with extra spaced removed :P https://github.com/elichai/rust-secp256k1/commit/1b0b22175a82e3767d6744492552e0eab2c2ecaf
12:22 < BlueMatt> elichai2: why not compare usize to size_t?
12:23 < BlueMatt> elichai2: but, yea, that looks awesome! do you want to open a pr with the whole thing or should i take the commits?
12:23 < elichai2> Felt it's more inline with the rest of the types that we typedefed
12:23 < BlueMatt> eh, yea, fair enough, whatever.
12:24 < elichai2> I can either open a PR into your branch via github or you can cherry-pick them however you prefer
12:24 < elichai2> https://github.com/elichai/rust-secp256k1/tree/2020-04-sanity_wasm
12:29 < BlueMatt> elichai2: done!
12:30 < BlueMatt> elichai2: oh, wait, no, we cant include stddef, I dont think
12:30 < BlueMatt> I had issues #includeing anything
12:31 < BlueMatt> elichai2: for size_t I used unsigned long
12:31 < BlueMatt> which I know isnt technically correct, but...
12:31 < elichai2> that's included in all of secp headers
12:31 < elichai2> we use size_t quite a lot
12:31 < BlueMatt> oh, hmmm, I thought I'd avoided
12:31 < BlueMatt> it
12:31 < BlueMatt> no, you're right
12:32 < BlueMatt> one sec
12:33 < BlueMatt> ok, used it correctly in the dummy string.h
12:33 -!- jonatack [~jon@109.232.227.149] has quit [Ping timeout: 244 seconds]
12:34 < BlueMatt> elichai2: plz2ack 208
12:34 < elichai2> andytoshi: I ACKed, altough I see now that it doesn't count in rust-secp
12:35 -!- jonatack [~jon@37.170.24.56] has joined #rust-bitcoin
12:38 < BlueMatt> lol ffs
12:39 < elichai2> Why can I delete a comment but not an accidental review :/
12:39 < BlueMatt> cause github's review tools are shit
12:39 < andytoshi> lol
12:39 < andytoshi> ok taking a look
12:41 < andytoshi> ok will merge when travis finishes
12:46 < BlueMatt> yay
12:48 < BlueMatt> any reason to not do a release?
13:08 < andytoshi> no reason, we can do it
13:08 < andytoshi> elichai2: do you know offhand if we need to do a major release, and of the full lib or just -sys?
13:08 < andytoshi> if not i can investigate it
13:10 < elichai2> hmm We need major for the main lib because of ecdh_raw breakage
13:10 < elichai2> about the sys, I still haven't understood when it should be minor and when major
13:11 < elichai2> nothing was really changed I think
13:11 < elichai2> and if we do another minor for `-sys` maybe bump the underlying libsecp too? it can cut `lowmemory` mem usage by half
13:12 < elichai2> But we can postpone this to later (real_or_random is trying to push to an official release of libsecp soon)
13:25 < BlueMatt> if were doing a major bump we could take this too, mod travis failures https://github.com/rust-bitcoin/rust-secp256k1/pull/154
13:31 -!- DeanGuss [~dean@gateway/tor-sasl/deanguss] has quit [Ping timeout: 240 seconds]
13:34 -!- jonatack [~jon@37.170.24.56] has quit [Quit: jonatack]
13:36 -!- DeanGuss [~dean@gateway/tor-sasl/deanguss] has joined #rust-bitcoin
13:46 -!- jonatack [~jon@2a01:e0a:53c:a200:bb54:3be5:c3d0:9ce5] has joined #rust-bitcoin
14:41 -!- michaelfolkson [~textual@2a00:23c5:be01:b201:2576:934:d62b:3df0] has quit [Quit: Sleep mode]
14:43 -!- michaelfolkson [~textual@2a00:23c5:be01:b201:2576:934:d62b:3df0] has joined #rust-bitcoin
15:01 -!- mauz555 [~mauz555@2a01:e0a:56d:9090:b94c:baeb:34ca:6eb6] has quit [Remote host closed the connection]
15:02 -!- mauz555 [~mauz555@2a01:e0a:56d:9090:5d17:dfd5:86c9:227] has joined #rust-bitcoin
15:21 -!- michaelfolkson [~textual@2a00:23c5:be01:b201:2576:934:d62b:3df0] has quit [Ping timeout: 272 seconds]
15:37 -!- DeanGuss [~dean@gateway/tor-sasl/deanguss] has quit [Remote host closed the connection]
15:37 -!- DeanGuss [~dean@gateway/tor-sasl/deanguss] has joined #rust-bitcoin
15:56 -!- prologas [4e3a633f@78-58-99-63.static.zebra.lt] has joined #rust-bitcoin
16:10 -!- DeanGuss [~dean@gateway/tor-sasl/deanguss] has quit [Ping timeout: 240 seconds]
16:22 -!- DeanGuss [~dean@gateway/tor-sasl/deanguss] has joined #rust-bitcoin
16:26 -!- prologas [4e3a633f@78-58-99-63.static.zebra.lt] has quit [Remote host closed the connection]
16:36 -!- mauz555 [~mauz555@2a01:e0a:56d:9090:5d17:dfd5:86c9:227] has quit [Read error: Connection reset by peer]
16:44 < ariard> BlueMatt: yeah 5min for 610, almost done
17:38 < ariard> BlueMatt: okay found failure reason in 610, but that's something which is already here
17:39 < ariard> BlueMatt: look in Channel::build_htlc_transaction, it uses a bool local flag
17:41 < ariard> which doesn't make sense because you never sign local htlc transaction for offchain
17:41 < ariard> and this bool is turn to true which means we use their_to_self delay intead of our_to_self_delay
17:42 < ariard> their_to_self_delay is provided by them, we should never us it
17:54 < ariard> BlueMatt: wait in fact that's correct because it's called in Channel::commitment_signed so when we verify remote signatures on oour
17:54 < ariard> local transactions with remote csv requirement
17:58 < BlueMatt> ariard: hmmm, wait, whats up?
17:58 < BlueMatt> I'm confused
18:00 < ariard> BlueMatt: code is right, but variable names are so confusing that I inverted both delays in one of the top commit of 610
18:00 < BlueMatt> ah lol....well local and remote are consistently confusing everywhere.......
18:01 < ariard> yes, I add a commit to rename them, at least in ChannelMonitor
18:25 < BlueMatt> ariard: lmk when tests pass and I'll go review it again :p
18:35 < ariard> BlueMatt: yeah test were locally good on last push, use last commit I mention in PR :)
18:36 < ariard> c7761f, tests are good, just a 1.22.0 issue, hmmm
18:42 < ariard> BlueMatt: tip d9ab6f6
19:58 -!- midnight is now known as midnightmagic
19:59 -!- midnightmagic is now known as midnight
20:30 -!- DeanGuss [~dean@gateway/tor-sasl/deanguss] has quit [Remote host closed the connection]
20:30 -!- DeanGuss [~dean@gateway/tor-sasl/deanguss] has joined #rust-bitcoin
--- Log closed Thu Apr 30 00:00:08 2020