--- Log opened Sun Nov 17 00:00:11 2019
00:07 -!- roconnor [~roconnor@host-45-78-202-124.dyn.295.ca] has joined #secp256k1
00:20 -!- belcher [~belcher@unaffiliated/belcher] has joined #secp256k1
01:22 < elichai2> I hate that in C `int` is used as a bool, it's both way bigger than usually needed and the size+alignment is platform dependant :/ (why not unsigned char :( )
01:37 < roconnor> my compiler says that sizeof(bool) = 1.
01:39 < roconnor> This is C99 though.
01:50 < gmaxwell> it's usually 1,  I think using 4 is a visual studioism.
01:56 < elichai2> Well C89 doesn't has bool, so people(libsecp included) use ints to represent booleans
01:57 < gmaxwell> there isn't any reason otherwise in libsecp256k1-- there isn't any place where there are many of them
01:57 < gmaxwell> and on some hardware (and historically) operations on ints are faster.
01:58 < elichai2> yeah but then bindings are a bitch. because you need to match to the same int
01:58 < gmaxwell> huh? no you don't.
01:58 < elichai2> and when they're in structs it's more space on the stack for no good reason (and then you complain :P)
01:59 < gmaxwell> It's normal for wrappers to change types.
01:59 < gmaxwell> elichai2: usualy they don't, due to alignment.
01:59 < elichai2> yeah, if you need to pass `int *has_quad_y` then you need to give it a valid C int, which is platform dependent
01:59 -!- jonatack [~jon@2a01:e35:8aba:8220:6627:dad:d967:649d] has quit [Ping timeout: 250 seconds]
02:00 < gmaxwell> Make the wrapper change the type, if thats required to give an ideomatic interface.
02:02 < elichai2> you still need to maintain a table of the types size per platform, which with unsigned char / uint8_t isn't needed. I'm mostly venting not asking for anything :)
02:03 < gmaxwell> what? table of the types size per platform? what broken language are you talking about. You should have access to the abi sizes through the ffi.
02:06 < elichai2> gmaxwell: https://doc.rust-lang.org/src/std/os/raw/mod.rs.html altough it seems that rust just doesn't support any platform which `int != i32` hmm
02:31 < elichai2> nickler: I wonder about the overhead of calling `secp256k1_xonly_pubkey_serialize` in `secp256k1_schnorrsig_verify` because verify already checks all the inputs and allocates a buffer on the stack, then `secp256k1_xonly_pubkey_serialize` does exactly the same all over again, which then calls `secp256k1_ec_pubkey_serialize` which does everything again. It's meaningless in the big scheme of time, just wanted to note
02:42 -!- jonatack [~jon@213.152.161.229] has joined #secp256k1
02:47 -!- roconnor [~roconnor@host-45-78-202-124.dyn.295.ca] has quit [Ping timeout: 245 seconds]
05:45 < nickler> elichai2: you mean the redundant allocating of a 33 byte buffer? Yeah, I left a TODO comment in the code. I think that could be fixed in a different PR because it'll need to change eckey_serialize by adding a new flag or we need more copy&paste.
06:09 < waxwing> in the musig paper p.12 in the section on Wagner on the naive version, i feel like it should say X_1 in several places where it says X_h(~). The logic seems to work like that.
06:18 -!- jonatack [~jon@213.152.161.229] has quit [Ping timeout: 265 seconds]
06:38 -!- roconnor [~roconnor@host-45-78-202-124.dyn.295.ca] has joined #secp256k1
06:55 -!- belcher [~belcher@unaffiliated/belcher] has quit [Quit: Leaving]
08:35 -!- roconnor [~roconnor@host-45-78-202-124.dyn.295.ca] has quit [Ping timeout: 265 seconds]
09:18 -!- jonatack [~jon@2a01:e35:8aba:8220:6627:dad:d967:649d] has joined #secp256k1
11:04 -!- belcher [~belcher@unaffiliated/belcher] has joined #secp256k1
17:41 -!- kallewoof [~quassel@240d:1a:759:6000:a7b1:451a:8874:e1ac] has joined #secp256k1
19:29 -!- belcher [~belcher@unaffiliated/belcher] has quit [Ping timeout: 246 seconds]
20:25 -!- roconnor [~roconnor@host-45-78-202-124.dyn.295.ca] has joined #secp256k1
21:02 -!- roconnor [~roconnor@host-45-78-202-124.dyn.295.ca] has quit [Ping timeout: 265 seconds]
22:22 -!- Madars [~null@unaffiliated/madars] has quit [Ping timeout: 276 seconds]
23:21 -!- belcher [~belcher@unaffiliated/belcher] has joined #secp256k1
23:38 -!- arubi [~ese168@gateway/tor-sasl/ese168] has quit [Ping timeout: 260 seconds]
--- Log closed Mon Nov 18 00:00:12 2019