--- Log opened Wed Oct 19 00:00:52 2022
03:30 -!- jon_atack [~jonatack@user/jonatack] has joined ##miniscript
03:32 -!- jonatack [~jonatack@user/jonatack] has quit [Ping timeout: 268 seconds]
07:16 <@sipa> Here is a writeup with the changes to descriptors/miniscript we discussed in person at tabconf: https://gist.github.com/sipa/06c5c844df155d4e5044c2c8cac9c05e
07:16 <@sipa> darosior and others ^
07:18 < darosior> Thanks for the writeup, will go through it
07:44 < darosior> Score is nea
07:44 < darosior> t
07:45 <@sipa> if it works... it'll need some coding/fuzzing to figure out if it's actually a sufficient rule
08:09 -!- jon_atack [~jonatack@user/jonatack] has quit [Ping timeout: 264 seconds]
09:02 < andytoshi> sipa: idk if you looked at maaku's recent email about just "using the CMS dummy push to bitmap which keys are used, rather than adding CHECKSIGADD", but he says that this was never considered, and i think we did consider it (if only for a couple minutes)
09:02 < andytoshi> sorry to bring something borderline-trollish into here
09:03 < andytoshi> it's pretty ugly, would eliminate any hope of doing weighted thresholds, but it _would_ have saved blockspace and probably we should've added a footnote or something to bip341 about it
09:04 < instagibbs> it was definitely considered
09:04 < instagibbs> alas, yes should be documented
09:07 < andytoshi> istr that our attitude was roughly "if you're doing shit directly on the blockchain rather than using musig/frost, you're already taking a space hit in exchange for noninteractivity, so an extra space hit isn't the end of the world"
09:07 < andytoshi> at least, that was my attitude
09:07 < andytoshi> which is potentially-controversial enough that we should've been more explicit
09:08 < andytoshi> (and yes, mark would've vehemently disagreed, but at least it would've been explicit)
09:12 <@sipa> @andytoshi Yeah, I think we didn't feel like opening up the bikeshedding for designing some bitvector approach, as if you care about efficiency, you'd probably be using some key aggregation scheme anyway.
09:12 <@sipa> There are OP_SUCCESSx, so something like that can be added later with no loss in efficiency.
09:13 < andytoshi> yeah, good point
09:13 < andytoshi> BTW can i have ops (i forget if this is one of the channels where we need to maintain ops at all times or we lose it forever)
09:13 < andytoshi> sipa: i think mark's email was actually pretty concilitory -- he said that he was just bringing it up so that we don't forget that it's a technique in the future
09:14 < andytoshi> it may be worthwhile to reply saying yeah, this is true, and if there are people who simultaneously need noninteractivity and efficiency, that it may be worth SUCCESSx'ing it in in the future
09:14 < andytoshi> though i don't particularly want to reply, and i'm sure you don't either :)
09:14 <@sipa> Yes, agree. It'd certainly have been worth pointing out this possibility in a footnote or so (or we could still do so).
11:50 < sanket1729> andytoshi: I hope you are not preparing new release notes :P
11:50 < sanket1729> I already have them in the release PR
11:55 < andytoshi> sanket1729: what release PR?
11:57 < andytoshi> sanket1729: yes, i've written up release notes .. i am about 75% done. it looks like most of the remaining PRs for me to note you've covered so maybe i'll just copy your stuff
11:57 < andytoshi> but my version has more links and exposition so i'd like to keep it :P
11:58 < andytoshi> a couple things i'm stuck on -- we were thinking about reverting 473. should we do that?
11:58 < andytoshi> and then secondly, in rust-bitcoin we use TimeLock but in rust-miniscript we use Locktime and Timelock .. should we unify the capitalization somehow?
12:03 < sanket1729> I don't think we were ever reverting 473?
12:04 < sanket1729> It was more like not merging 474
12:05 < sanket1729> Also, IMO we should cut release if we are this close. I don't mind fixing the weird edge case calculations of varint sizes in witness cost estimations in later minor release
12:05 < sanket1729> Can sneak those in as minor bug fixes
12:06 < sanket1729> darosior: I think the multi-path stuff would have to wait for this release :( . Would need more detail review for the large PR. 
12:07 < sanket1729> I have not yet completely figured out how to do it, but perhaps we can unify the Xpub and MultiXub instead of them requiring separate structs
12:12 < andytoshi> ok agreed
12:13 < andytoshi> i need like an hour though
12:13 < andytoshi> sanket1729: if you want to make a release PR can you just copy the first couple paragraphs from https://gist.github.com/apoelstra/e24863f702ebe35515a774eca44d4e2d ?
12:13 < andytoshi> or else i'll finish up my changelog and make my own
12:13 < andytoshi> when i get a change
12:13 < andytoshi> chance
12:14 < sanket1729> I can work on the release PR.
12:14 < sanket1729> I think we should get 475
12:14 < sanket1729> which fixes a reachable panic
12:50 < sanket1729> pushed a release PR
13:10 < andytoshi> ok reviewing 475
13:12 < andytoshi> sanket1729: can we add a method to DescriptorPublicKey which converts to a DefiniteDescriptorKey without providing a dummy index?
13:12 < andytoshi> oh there is one
13:13 < andytoshi> DefiniteDescriptorKey::ne
13:13 < andytoshi> new
13:13 < andytoshi> ok, will run tsets on 475 then merge, and then same with the release PR
13:15 -!- jon_atack [~jonatack@user/jonatack] has joined ##miniscript
13:20 -!- jonatack1 [~jonatack@user/jonatack] has joined ##miniscript
13:21 -!- jon_atack [~jonatack@user/jonatack] has quit [Ping timeout: 255 seconds]
13:26 < andytoshi> ok, running tests on 462 now. your changelog oloks great! thanks for clarifying the pkh stuff
13:27 -!- jonatack1 [~jonatack@user/jonatack] has quit [Ping timeout: 255 seconds]
13:30 -!- jonatack1 [~jonatack@user/jonatack] has joined ##miniscript
13:36 < sanket1729> awesome
13:36 -!- jonatack1 [~jonatack@user/jonatack] has quit [Read error: Connection reset by peer]
13:40 < sanket1729> I am working on upping elements-miniscript to everything latest
13:56 < andytoshi> oh thank you!
13:58 < andytoshi> running final tests before merging 462
13:58 < andytoshi> then i'll publish
13:58 < andytoshi> it's been six months less a day since 7.0.0 .. let's try to get a new release out much faster (maybe 3 monthns not 6)
14:01 < sanket1729> Also merged the rust-elements release PR
14:02 < sanket1729> release would be required for elements-miniscript
14:06 < andytoshi> published and tagged rust-minisrcipt
14:06 < andytoshi> kk will publish/tag rust-elemnts
14:07 < andytoshi> done
14:19 < sanket1729> Thanks
14:24 < sanket1729> andytoshi: How do you get more descriptive merge commit message from the PR being merged?
14:24 < sanket1729> If I type 'git merge <commit-hash>' the default message is "Merged <hash> into <branch>"
14:25 < andytoshi> sanket1729: i use the bitcoin core merge script
14:26 < andytoshi> https://github.com/bitcoin-core/bitcoin-maintainer-tools/blob/main/github-merge.py
14:26 < sanket1729> I don't mean the merge script
14:26 < sanket1729> I mean while merging commits from rust-miniscript into elements-miniscript
14:26 < andytoshi> ohh yeah
14:26 < andytoshi> lol i did that manually
14:26 < sanket1729> Nice, Thanks for the hard work :)
14:26 < andytoshi> sorry to disappoint
14:27 < andytoshi> :P no worries. i would like to write a nice script for it. for Bitcoin Core->Elements merging I have one, maybe i should try to generalize that..
14:27 < andytoshi> i like the way the Core merge script uses git variables (e.g. testcmd) to configure itself for each repo
15:21 < jeremyrubin> re maku email, there is tremendous value in improving the epistemology of why things are the way they are, especially if we're not rationalizing decisions that were made without knowledge of the given answer. 
15:22 < jeremyrubin> the effort for any particular trivia might not be worth it, but in aggregate it makes it easier for new folks to learn up from our written records
19:05 -!- jonatack1 [~jonatack@user/jonatack] has joined ##miniscript
19:27 -!- jonatack1 [~jonatack@user/jonatack] has quit [Read error: Connection reset by peer]
21:04 <@sipa> andytoshi darosior sanket1729 achow101 Do you think we should outlaw musig() inside pkh? Or outlaw pkh altogether in tapscript descriptors?
21:05 <@sipa> (because 80 bit collision security)
--- Log closed Thu Oct 20 00:00:52 2022