--- Log opened Wed Dec 26 00:00:05 2018
01:01 -!- elichai2 [uid212594@gateway/web/irccloud.com/x-nrgbenpcgnfxidig] has joined #rust-bitcoin
01:50 -!- e4xit [~e4xit@cpc123762-trow7-2-0-cust7.18-1.cable.virginm.net] has joined #rust-bitcoin
06:04 -!- e4xit [~e4xit@cpc123762-trow7-2-0-cust7.18-1.cable.virginm.net] has quit [Quit: quit]
06:07 -!- e4xit [~e4xit@cpc123762-trow7-2-0-cust7.18-1.cable.virginm.net] has joined #rust-bitcoin
06:31 -!- belcher [~belcher@unaffiliated/belcher] has quit [Ping timeout: 250 seconds]
09:12 -!- grubles [~grubles@unaffiliated/grubles] has quit [Remote host closed the connection]
10:23 < real_or_random> dongcarl: ah I just wanted to point out that debian stable has rust 1.24 now... but you were quicker
10:24 < dongcarl> Oh wow it’s the legendary real_or_random
10:24 < real_or_random> Oo
10:26 < dongcarl> Haha yeah... I’ve bootstrapped up to 1.20.0with mrustc
10:26 < dongcarl> But it can technically be done all the way up to nightly
10:27 < dongcarl> Given enough time haha
10:28 < andytoshi> at risk of goalpost-moving, i would like to use mrustc (plus one or two, maybe, if there's a good reason) as our target rustc rather than debian
10:28 < andytoshi> since mrustc is what people have to use to avoid "trusting trust" attacks
10:31 < dongcarl> andytoshi: let’s post the discussion on GitHub
10:31 < BlueMatt> dongcarl: lol take a vacation, bro
10:32 < BlueMatt> dongcarl: cause the ripemd fuzzer uses the length in the final result, but sha256/512 do not (because thats how they were originally written, which sucks, but oh well)
10:32 < dongcarl> I’m happy to do so, but we need to talk about the steps... Who’s uploading the binary to Travis? I don’t think mrustc is deterministically built yet so some governance decision needs to be made
10:33 < dongcarl> BlueMatt: Ah... it seems the fuzzing tests are still working for me tho?
10:33 < dongcarl> As in, passing
10:33 < BlueMatt> dongcarl: there are no fuzz tests? but the patch looks fine, the final length just gets ignored by the sha256/512 hashers
10:35 < BlueMatt> andytoshi: mrustc seems like a good target to me. If it supports 1.20, 1.20 sgtm
10:35 < BlueMatt> rust-lightning does 1.22, though
10:35 < dongcarl> It doesn’t support 1.20.0
10:36 < dongcarl> It supports 1.19.0
10:36 < BlueMatt> ah
10:36 < dongcarl> And I use that to compile 1.20.0
10:36 < dongcarl> We can do the nesting doll all the way up tho
10:36 < BlueMatt> yea
10:37 < dongcarl> I have the ability to produce the tarballs... I just need to know what the governance will be
10:37 < dongcarl> Who’s uploading
10:37 < dongcarl> Where are we building
10:37 < dongcarl> Etc
10:37 < dongcarl> Mostly cuz it’s not deterministic :-/
10:37 < andytoshi> BlueMatt: so, 1.20 gets us something big over 1.19 (iirc it's associated constants) so IMO it's worth the extra inconvenience to make people boostrap an entire compiler version
10:37 < BlueMatt> well in terms of policy around version we dont need policy around mrustc?
10:37 < BlueMatt> andytoshi: sgtm
10:38 < BlueMatt> 1.20 is how old now?
10:38 < andytoshi> lol, like 18 monts
10:38 < andytoshi> https://blog.rust-lang.org/2017/08/31/Rust-1.20.html aug 31
10:38 < andytoshi> 2017
10:38 < BlueMatt> hmm, not as old as i thought
10:38 < BlueMatt> man the rust ecosystem moves to fast :(
10:38 < andytoshi> yeah, holy shit eh
10:39 < BlueMatt> heh, my gut would be "two years", but that leaves us at 1.14 lol
10:40 < BlueMatt> but I'm ok with 1.20, assuming we're not about to jump and do it again in a few months
10:41  * BlueMatt -> out
12:11 -!- elichai2 [uid212594@gateway/web/irccloud.com/x-nrgbenpcgnfxidig] has quit [Quit: Connection closed for inactivity]
14:41 < BlueMatt> note that cc 1.0.26 *also* doesn't build on rust 1.14 on windows
21:49 -!- grubles [~grubles@unaffiliated/grubles] has joined #rust-bitcoin
22:01 -!- grubles [~grubles@unaffiliated/grubles] has quit [Remote host closed the connection]
22:02 -!- grubles [~grubles@unaffiliated/grubles] has joined #rust-bitcoin
--- Log closed Thu Dec 27 00:00:04 2018