--- Day changed Thu Jul 26 2018
00:04 < andytoshi> dongcarl: easy to access and initialize and modify
00:04 < andytoshi> when you're building a transaction you're constantly adding and removing inputs and outputs, replacing scripts, etc
00:05 < andytoshi> and the docs are wrong - that should say "witness script" rather than "redeem script"
00:06 < dongcarl> haha okay cool, makes more sense
00:06 < andytoshi> "redeem script" is a p2sh term
00:06 < andytoshi> yeah, my bad
00:07 < dongcarl> I'll PR
00:08 < andytoshi> thanks
00:11 < dongcarl> https://github.com/rust-bitcoin/rust-bitcoin/pull/109
02:24 -!- treyzania [~treyzania@paphos.tr3y.io] has quit [Ping timeout: 264 seconds]
02:25 -!- treyzania [~treyzania@paphos.tr3y.io] has joined #rust-bitcoin
14:06 < andytoshi> BlueMatt: https://github.com/rust-bitcoin/rust-secp256k1/pull/39 bumps the version number of rust-secp and adds CHANGELOG.md to document the changes (i'm planning to do the same thing for rust-bitcoin on the next major version)
14:06 < andytoshi> can you ack it?
14:51 < BlueMatt> andytoshi: dongcarl I'm very confused
14:52 < BlueMatt> "redeemscript" is *definitely* the term used in segwit, too
14:52 < BlueMatt> "witness script" refers to the whole set of stack pushes in the witness
14:52 < BlueMatt> *not* the redeem script part that is being hashed there
15:11 < andytoshi> but bip141 says "The witnessScript (≤ 10,000 bytes) is popped off the initial witness stack. SHA256 of the witnessScript must match the 32-byte witness program."
15:12 < andytoshi> we have been using this terminology internally at blockstream (at least gwillen instagibbs nickler and myself, and surely sipa has noticed us doing so) and nobody has ever commented on this
15:13 < andytoshi> and only the top element if the witness stack is a script
16:03 -!- elichai2 [uid212594@gateway/web/irccloud.com/x-jbpgyhistpbhqcza] has joined #rust-bitcoin
16:08 -!- savil_ [~savil@2604:a880:2:d0::2119:6001] has quit [Quit: ZNC 1.7.1 - https://znc.in]
16:11 -!- savil[m] [savilmatri@gateway/shell/matrix.org/x-rhajeqmfmgutpydc] has joined #rust-bitcoin
16:19 < BlueMatt> andytoshi: funny I've never seen it, anyway, #bitcoin-core-dev seemed to agree it was at least confusing, even if accepted by some people
16:19 < BlueMatt> and given there are other options, best to use the non-confusing one
16:20 < BlueMatt> andytoshi: given there is *also* a scriptWitness
16:20 < BlueMatt> which refers to the whole witness
16:24 < andytoshi> yes, the scriptwitness is the witness to the script
16:24 < andytoshi> a witnessscript is a script that appears in the witness
16:24 < andytoshi> and a redeemscript has a meaning as per bip16 for many years
16:26 < andytoshi> i have a ton of source code that uses these terms the way bip141 does, i had never encountered any other interpretation until this morning
16:28 < andytoshi> i'm happy to use "witness redeem script" in docs, but there are variables all over liquid and icboc and probably rust-bitcoin called witness_script and redeem_script which have the bip141 meanings
16:35 < savil[m]> BlueMatt:  I'm looking at your suggestion to remove the "hex" dependency in https://github.com/rust-bitcoin/rust-bitcoin/pull/107.   Both hex::encode and hex::decode are used in non-test code.  The only way to make the dependency test-only, is to basically pull in the implementation into some utility file. What does that gain us, over referencing the original library?
16:36 < andytoshi> savil[m]: it means that the code is under our control and our review
16:37 < savil[m]> andytoshi: right, but its currently pegged to a commit number, so it won't be changing without us deciding to do so?
16:37 < andytoshi> i think a dep that pinned a git commit would be ok, except that git hashes are not collision-resistant, and it's still a review difficulty (an extra dep distracting people in Cargo.lock and teh build output)
16:38 < savil[m]> andytoshi: I don't have a strong opinion here. Should I just copy-pasta the implementation into a utility file? I can try to whittle it down but have to look deeper into that
16:39 < andytoshi> no, you can't copy and paste it, the licenses are incompatible
16:39 < andytoshi> you have to rewrite it (i'm pretty sure there is existing code somewhere in the codebase, or maybe in rust-jsonrpc or strason)
16:40 < savil[m]> andytoshi: hmm I see. Would you be opposed to landing this as is, and making a separate task to remove the hex-dependency?
16:41 < andytoshi> no, i think that's reasonable. i can remove the hex dependency later in a serde cleanup commit
16:41 < andytoshi> i'll post on the PR
16:41 < savil[m]> incremental steps and all that
16:41 < savil[m]> cool
16:49 < BlueMatt> savil[m]: my suggestion was that it was only used in non-test code in serde stuff
16:49 < BlueMatt> so if serde is made optional (which I think we wanted to do anyway) then you could make hex optional at the same time
16:54 < savil[m]> BlueMatt: ah, I see. And that is done by making a "serde" feature in Cargo.toml, and specifying that the hex-dependency only applies when that feature is enabled?
16:54 < andytoshi> we'll probably do that, yeah
16:55 < andytoshi> i'm not sure how to make cargo do that (making a dep optional based on a separate feature) but i assume it's easy
16:58 < andytoshi> oh, i figured it out
16:59 < savil[m]> yeah, https://doc.rust-lang.org/cargo/reference/manifest.html#the-features-section talks about it
17:18 < savil[m]> andytoshi: something like this I think https://gist.github.com/savil/1fafb6f4ab909f4e0264ceaf8ad33b97 (with a better name for "serde_feature". It can't be "serde", because that's the name of a dependency).
17:20 < savil[m]> anyway, I'll defer to you and pick up something else
17:37 < andytoshi> thanks!
17:37 < andytoshi> i think i might need a separate "serde_feature" anyway because i need to also turn on the rust-secp serde feature when serde is enabled for rust-bitcoin
18:40 < BlueMatt> andytoshi: https://github.com/rust-bitcoin/rust-secp256k1/pull/39#issuecomment-408195295
18:47 < dongcarl> andytoshi, BlueMatt: I'm happy to change the term to whatever is least confusing...
18:47 < BlueMatt> dongcarl: probably best to have this discussion more broadly (#bitcoin-core-dev and or the bitcoincore.org issue)
18:48 < BlueMatt> we need a standard here, cause it seems several have developed
18:50 < dongcarl> True
18:51 < BlueMatt> actually, bitcoin core irc meeting in 10 minutes
18:51 < BlueMatt> I'll bring it up then
18:51 < BlueMatt> not exactly the right venue, but whatever
20:16 < andytoshi> BlueMatt: re 1.0, :D
20:16 < andytoshi> but i think we should do 0.10.0 for now, make sure that we can upgrade rust-bitcoin and at least one actual project (icboc or liquid) to it
20:16 < andytoshi> then i'd be happy to publish 1.0 with no changes
20:17 < BlueMatt> yea, ok, fair
20:20 < andytoshi> rust-lightning would also count as an "actual project", though liquid has the nice property of actually using the serde feature
20:21 < BlueMatt> yea, I think I'm gonna be a bit to upgrade to it, sadly
20:21 < BlueMatt> the new api will take me a bit to check all the callsites
20:21 < andytoshi> yeah, ditto
20:21 < andytoshi> fortunately they changed in a way that you can't really do the wrong thing
20:26 < andytoshi> is there a way i can make crates.io use my yubikey instead of whatever it does now (i think it has a fixed token somewhere in my ~/.cargo?)
20:26 < BlueMatt> well, except use a too-large ctx :p
20:27 < andytoshi> heh, yes
20:27 < andytoshi> though arguably it's easier now to use the right size, since the compiler will bitch at you if you go too small
20:27 < andytoshi> anyway published
20:28 < BlueMatt> yes, totally agreed
20:28 < BlueMatt> really excited to do it
20:29 < andytoshi> ditto, it'll be fun, esp since some error paths actually went away
20:29 < andytoshi> like, `.unwrap()s` can be deleted
21:02 -!- savil [~savil@c-71-202-1-72.hsd1.ca.comcast.net] has joined #rust-bitcoin
21:02 -!- savil [~savil@c-71-202-1-72.hsd1.ca.comcast.net] has left #rust-bitcoin ["Textual IRC Client: www.textualapp.com"]
21:20 -!- savil[m] [savilmatri@gateway/shell/matrix.org/x-rhajeqmfmgutpydc] has left #rust-bitcoin ["User left"]
21:23 -!- elichai2 [uid212594@gateway/web/irccloud.com/x-jbpgyhistpbhqcza] has quit [Quit: Connection closed for inactivity]
21:29 -!- savil[m]2 [savilmatri@gateway/shell/matrix.org/x-rhajeqmfmgutpydc] has joined #rust-bitcoin