--- Log opened Sun Aug 25 00:00:50 2019
03:30 -!- jonatack [~jon@2a01:e35:8aba:8220:6627:dad:d967:649d] has quit [Quit: jonatack]
05:33 -!- ghost43 [~daer@gateway/tor-sasl/daer] has quit [Ping timeout: 260 seconds]
05:36 -!- ghost43 [~daer@gateway/tor-sasl/daer] has joined #rust-bitcoin
05:38 -!- jonatack [~jon@2a01:e35:8aba:8220:6627:dad:d967:649d] has joined #rust-bitcoin
09:44 -!- elichai2 [uid212594@gateway/web/irccloud.com/x-sxmexnyzysywrqnl] has joined #rust-bitcoin
11:03 < dpc> https://github.com/rust-lang/wg-governance/issues/12 - there's already an open Cryptography WG application, and I guess cryptocurrencies would fall under it.
11:53 -!- elichai2 [uid212594@gateway/web/irccloud.com/x-sxmexnyzysywrqnl] has quit [Quit: Connection closed for inactivity]
12:23 < andytoshi> heh, i think more than half the people in that thread are people i'd prefer not to interact with if i can avoid it
12:25 -!- elichai2 [uid212594@gateway/web/irccloud.com/x-palnbqflrnedxzmq] has joined #rust-bitcoin
12:26 < elichai2> andytoshi: you're starting to influence burntoshi :)
12:26 < elichai2> https://github.com/BurntSushi/quickcheck/issues/241
12:27 < andytoshi> elichai2: heh, burntsushi has always been super conservative
12:28 < andytoshi> i think his comments on our "tempfile broke us" thread were because he was mad at me for shittalking npm, rather than because he actually thought it was ok to break things
12:30 < elichai2> makes sense heh
12:30 < elichai2> have you seen the npm ads thing?
12:30 < andytoshi> lol yeah
12:32 < andytoshi> so, i think there's wide agreement that "no MSRV, everyone has to use the latest" is not reasonable, so i think there'd be wide support for a cargo proposal that let projects set a MSRV
12:33 < andytoshi> so even though nobody likes our "support 1.22.0 for years" policy, i think we'll nonetheless eventually get cargo support for it
12:33 < andytoshi> and i really would like to help push that forward, but the withdrawal of the rust community from the public internet has been a blocker
12:34 < andytoshi> for my participation, at least
12:35 < andytoshi> offtopic: lol https://github.com/git/git/blob/master/banned.h
12:35 < elichai2> yeah I think the argument currently is about how old should we support and if that should be a social thing or a cargo thing
12:36 < elichai2> lol, do they have examples of why they're easy to misuse? (though everything string related in C is easy to misuse)
12:37 < andytoshi> well the current cargo story is "you can't support any compiler at all unless it happens to be the very latest release and happens to have no regressions in it"
12:37 < andytoshi> elichai2: git has its own string-handling library
12:38 < andytoshi> i think we're always going to be more-or-less on our own supporting 1.22 (with a few exceptions like serde (but not serde-derive), byteorder, etc., which never change and are widely used)
12:38 < elichai2> byteorder is pretty much dead in new rust versions
12:38 < elichai2> since the built in functions for be/le
12:40 < andytoshi> yeah, it'll be nice when we can support those
12:40 < andytoshi> though 'byteorder' is probably the only crate in our dep tree that i've actually read 100% of
12:40 < andytoshi> so of all the things to lose, it's probably the least helpful :)
12:41 < elichai2> haha, that's pretty cool, I read only some parts when I tried to realise why does it return results
12:42 < elichai2> because the built in rust ones literally just transmutes everything after calling `to_be()`/`to_le()`
12:43 < andytoshi> the result stuff is only for things that go through the io::Write and io::Read APIs
12:43 < andytoshi> which tragically are in std but not core
12:49 < elichai2> do we have good uses for a non cryptographic rng crate? i'm thinking of writing one for fun, should be pretty simple (either chacha or if I value speed PCG style or a simle Xor+Shift, the biggest part would be seeding from OS). that way I can slowly provide a more stable thinner alternatice to `rand`
12:49 < andytoshi> i mean, it would be nice to have a `Rng` trait without any actual rngs at all
12:50 < andytoshi> i implemented fortuna for rust-crypto way back in the day
12:50 < andytoshi> when dagenix was maintaining it
12:52 < elichai2> hmm and then people can implement it on top of either a cryptographically secure one or a fast non crypto
12:53 < elichai2> what's sepcial about Fortuna?
12:55 < andytoshi> this is interesting and relevant to us https://nullprogram.com/blog/2019/07/10/
12:55 < andytoshi> elichai2: i forget why i chose fortuna
15:31 < elichai2> andytoshi: hmm I started looking into stuff,  and it looks like to generate floating points `rand` is doing something not so uniformly random according to some people: https://docs.rs/rand/0.7.0/src/rand/distributions/float.rs.html?search=#104
15:32 < elichai2> https://mumble.net/~campbell/2014/04/28/uniform-random-float they're basically doing `random_real_53` here: https://mumble.net/~campbell/2014/04/28/random_real.c
17:44 -!- TamasBlummer1 [~Thunderbi@p200300DD671264815035A63851C67AC1.dip0.t-ipconnect.de] has joined #rust-bitcoin
17:46 -!- TamasBlummer [~Thunderbi@p200300DD67126418A47E9396C884C452.dip0.t-ipconnect.de] has quit [Ping timeout: 264 seconds]
17:46 -!- TamasBlummer1 is now known as TamasBlummer
18:26 < elichai2> andytoshi: i'll publish it tomorrow. feel free to give feedback :) https://github.com/elichai/Random-trait
20:51 -!- ghost43 [~daer@gateway/tor-sasl/daer] has quit [Remote host closed the connection]
20:51 -!- ghost43 [~daer@gateway/tor-sasl/daer] has joined #rust-bitcoin
22:23 -!- mryandao [~mryandao@gateway/tor-sasl/mryandao] has quit [Remote host closed the connection]
22:23 -!- mryandao [~mryandao@gateway/tor-sasl/mryandao] has joined #rust-bitcoin
--- Log closed Mon Aug 26 00:00:52 2019