--- Log opened Tue Apr 20 00:00:31 2021 00:05 -!- asymptotically [asymptotic@gateway/vpn/mullvad/asymptotically] has joined #joinmarket 00:39 -!- jungly [~jungly@host-79-37-201-97.retail.telecomitalia.it] has joined #joinmarket 01:56 -!- undeath [~undeath@hashcat/team/undeath] has joined #joinmarket 02:53 -!- jungly [~jungly@host-79-37-201-97.retail.telecomitalia.it] has quit [Remote host closed the connection] 03:10 -!- jonatack [~jon@88.127.52.83] has quit [Ping timeout: 240 seconds] 03:11 -!- jonatack [jon@gateway/vpn/airvpn/jonatack] has joined #joinmarket 03:15 -!- proslogion [~proslogio@14.127.251.171] has joined #joinmarket 03:18 -!- Curt73Hand [~Curt73Han@static.57.1.216.95.clients.your-server.de] has joined #joinmarket 04:07 -!- mryandao_ [~mryandao@gateway/tor-sasl/mryandao] has joined #joinmarket 04:07 -!- mryandao [~mryandao@gateway/tor-sasl/mryandao] has quit [Ping timeout: 240 seconds] 05:39 -!- jungly [~jungly@host-79-37-201-97.retail.telecomitalia.it] has joined #joinmarket 07:08 -!- k3tan [~pi@gateway/tor-sasl/k3tan] has quit [Ping timeout: 240 seconds] 07:31 -!- luke-jr [~luke-jr@unaffiliated/luke-jr] has quit [Quit: ZNC - http://znc.sourceforge.net] 07:31 -!- luke-jr [~luke-jr@unaffiliated/luke-jr] has joined #joinmarket 08:35 -!- ongo [~weechat@static.179.176.203.116.clients.your-server.de] has joined #joinmarket 08:46 < waxwing> i wrote up a quick summary of things about longer term changes. i know there are several other things that could be entered in that list, that i omitted (obvious example: actual JM *protocol* changes), but these are just things that are on my mind: https://joinmarket.me/blog/blog/blue-sky-joinmarket-thoughts/ 08:47 < waxwing> feel free to tell me why i'm wrong :) 09:37 < openoms_> @waxwing thank you so much for this writeup! So much to learn where thngs are going. Currently can hardly think about anything else than LN channels, but as you write that is just more of a reason to improve the privacy of the base layer. 09:50 < belcher> waxwing thanks for the blog post 09:50 < belcher> it might be useful to define the phrase "negotiation mechanism" in a sentence or two for someone like me who hasnt read the coinshuffle++ paper yet 09:51 < belcher> coinjoinxt seems like a great idea to me, agree with everything you've written there, though note that the coinjoinxt proposed transaction graph should contain all segwit script types not just taproot so that it better blends in with all the other transactions out there 09:54 < belcher> it might be worth mentioning the possibility that teleport (my coinswap project) just takes over entirely from joinmarket, it has essentially the same target audience and is cheaper in miner fees 09:54 < belcher> obviously that would happen probably years from now if ever 10:00 < belcher> for languages i would also suggest rust, iv been finding it really great 10:12 < waxwing> openoms_, cool, it's not really 'where things are going' as much as what i'm wondering might be future directions, very much just personal thoughts there 10:13 < waxwing> belcher, re 'negotiation' and 'coordination' mechanisms, part of it is i want to distinguish those two, with JM having the economic part for coordination but still the negotiation part is distinct, but i think you're dead right it's not a very clear term. 10:13 < waxwing> so iow in coinshuffle++ the output amount and the set of participants are just treated as input to the protocol, whereas we need to *coordinate* that in the real world. 10:14 < waxwing> but the protocol itself handles how the tx gets set up before signing right. yeah maybe i can add a sentence although don't want to get in the weeds. 10:14 < belcher> hmm, how accurate is "coordination: the process of choosing the coinjoin amount, the set of participants and any fees paid to miners or between participants" 10:15 < waxwing> i suppose in our case we could specify it as that, yeah 10:15 < belcher> but actually yes dont let me put words in your mouth, best you find definitions you're happy with and edit the post 10:16 < waxwing> re: including other script types, that's a good point. i'm more having coinjoinxt in mind today, because taproot is almost certainly here in some months. 10:16 < waxwing> it's such a huge change of assumptions for anything contract-y regarding its privacy. 10:21 < belcher> it might be valuable to use multiparty-ecdsa as well for the coinjoinxt multisigs, because that means regular ecdsa single sig might also be the first coinjoinxt transaction 10:21 < belcher> though i appreciate its a huge ask and using taproot is much easier, hopefully taproot will be common enough that coinjoinxt always having a taproot tx as its first tx wont leak privacy too much 10:44 -!- jungly [~jungly@host-79-37-201-97.retail.telecomitalia.it] has quit [Ping timeout: 240 seconds] 10:54 < waxwing> yes. and not only first; every tx needs a multisig anchor unfortunately. 10:55 < waxwing> and also i'd like to stick to stuff we can pull out of libsecp if at all possible. 10:57 -!- proslogion [~proslogio@14.127.251.171] has quit [Quit: This computer has gone to sleep] 11:10 < belcher> oh 11:10 < belcher> yes thats right i forgot (re multisig anchor for everything) 11:10 < belcher> or maybe i never realized, because thinking about it now it seems like a big downside 11:11 < belcher> multiparty ecdsa seems much more essential then... otherwise without steganography coinjoinxt is just a more expensive version of equal-output coinjoins 11:56 < waxwing> i mean not 100% but 90% yeah 12:02 < belcher> what do you mean by 100% / 90% waxwing ? 12:06 < waxwing> well, i don't think the idea is completely useless with exposed multisigs. like, it isn't necessarily 100% clear what's going on just because there are multisigs in there. 12:07 < waxwing> also a thing that i have to go back and look at again is whether you *always* need one of the inputs to be multisig. the main point is that all participants' signatures must be required on every tx. it's a bit tricky as i recall. 12:07 < waxwing> so i might have oversimplified saying "every tx requires an N/N multisig input as a thread going through the whole chain/tree". i can't remember now. 12:08 < waxwing> anyway once you have the stega- property in there what i like is that it doesn't have to be more expensive (just like how coinswap can be cheaper due to stega- ), should probably be cheaper with a few participants, as i think i said in the blog. 12:09 < belcher> hmm yes you're right about the cheapness, coinjoinxt with just 2 or 3 makers still could provides pretty good privacy from a passive observer of the blockchain 12:11 < belcher> are you sure that every tx in a coinjoinxt PTG has to be a multisig? isnt there some trick that you have the transactions signed backwards, so the entire thing isnt valid until its all been signed 12:12 < belcher> im reading my notes here from 2018 and it really assumes that... maybe i was just wrong haha 12:12 < waxwing> yeah that's the basic idea. you've got to not only get both sides to agree to the tx set, but also have it so that there's not a possibility to double spend. 12:13 < waxwing> if you remember another element was 'promise utxos', where someone promises to provide *another* input from outside the set, later on. those require backouts in case they double spend them in advance. 12:13 < waxwing> without that element it's just a tree with a single root which is a bit too limited. 12:14 < waxwing> this was the original gist, it's still basically right iirc: https://gist.github.com/AdamISZ/a5b3fcdd8de4575dbb8e5fba8a9bd88c 12:15 < belcher> ah right, so adding new inputs into the PTG requires a multisig, but txes which dont add inputs can be single sig 12:18 < waxwing> well it's more like, if TXN in the set has 2 outputs: to Alice, to Bob, they could presign that both those outputs continue in the tree (as inputs to TXN+1), but they could double spend them. whereas if it goes to 1 output 2/2 A/B, then they can't double spend. 12:18 < belcher> aha yes thats right 12:19 < waxwing> so the whole flow along a path needs to be dual control, but for the promise utxos which by definition aren't dual control, we shove in a backout 12:31 -!- jungly [~jungly@host-79-37-201-97.retail.telecomitalia.it] has joined #joinmarket 12:58 -!- asymptotically [asymptotic@gateway/vpn/mullvad/asymptotically] has quit [Quit: Leaving] 13:07 -!- k3tan [~pi@gateway/tor-sasl/k3tan] has joined #joinmarket 13:37 -!- jungly [~jungly@host-79-37-201-97.retail.telecomitalia.it] has quit [Ping timeout: 245 seconds] 13:43 -!- luke-jr [~luke-jr@unaffiliated/luke-jr] has quit [Quit: ZNC - http://znc.sourceforge.net] 13:43 -!- luke-jr [~luke-jr@unaffiliated/luke-jr] has joined #joinmarket 15:19 -!- Curt73Hand [~Curt73Han@static.57.1.216.95.clients.your-server.de] has quit [Ping timeout: 246 seconds] 15:45 -!- ricks_ [~ricks_@gateway/tor-sasl/ricks/x-19497574] has joined #joinmarket 15:50 -!- ricks_ [~ricks_@gateway/tor-sasl/ricks/x-19497574] has quit [Remote host closed the connection] 16:51 -!- asoltys_ [~root@s207-81-214-2.bc.hsia.telus.net] has quit [Quit: leaving] 16:51 -!- asoltys [~root@s207-81-214-2.bc.hsia.telus.net] has joined #joinmarket 17:01 -!- belcher_ [~belcher@unaffiliated/belcher] has joined #joinmarket 17:04 -!- belcher [~belcher@unaffiliated/belcher] has quit [Ping timeout: 260 seconds] 17:09 -!- undeath [~undeath@hashcat/team/undeath] has quit [Quit: WeeChat 3.0.1] 17:11 -!- belcher_ is now known as belcher 17:36 -!- proslogion [~proslogio@14.127.251.171] has joined #joinmarket 17:47 -!- proslogion [~proslogio@14.127.251.171] has quit [Quit: This computer has gone to sleep] 21:39 -!- RusAlex [~Chel@unaffiliated/rusalex] has quit [Ping timeout: 265 seconds] 22:05 -!- RusAlex [~Chel@unaffiliated/rusalex] has joined #joinmarket 22:59 -!- Guest21684 [~GAit@101.ip-213-32-22.eu] has quit [Ping timeout: 245 seconds] 23:00 -!- Guest21684 [~GAit@unaffiliated/gait] has joined #joinmarket 23:07 -!- dave_uy [~david@108.61.193.26] has quit [Quit: The Lounge - https://thelounge.chat] 23:11 -!- dave_uy [~david@108.61.193.26] has joined #joinmarket 23:14 -!- midnight [~midnight@unaffiliated/midnightmagic] has quit [Ping timeout: 245 seconds] 23:28 -!- midnight [~midnight@unaffiliated/midnightmagic] has joined #joinmarket 23:50 -!- jungly [~jungly@host-79-35-191-191.retail.telecomitalia.it] has joined #joinmarket --- Log closed Wed Apr 21 00:00:32 2021