--- Log opened Tue Dec 28 00:00:14 2021
01:34 < RCasatta[m]> addressed feedback and squashed https://github.com/rust-bitcoin/rust-bitcoin/pull/672
11:05 -!- trev [~trev@user/trev] has quit [Quit: trev]
15:54 < ariard> BlueMatt: yeah you're thinking 30% as a default is too sensitive ? 50%, 70% ? maybe we should disable by default and experiment the feature with advanced users...
15:54 < ariard> gonna write the tests at least if you're ack
15:57 < BlueMatt[m]> ariard yea, I still feel like we should just expand the channel-close time if feerate goes up, instead of just flat-out always closing if the feerate goes up by NN%
15:57 < BlueMatt[m]> though I do like the min-increase-NN%-before-any-close
15:57 < BlueMatt[m]> because you're always at risk of closure if someone sends an HTLC right at the advertised minimum-cltv-delta advertised
15:58 < BlueMatt[m]> but it feels...wrong? if an htlc expires in a full day and a NN% increase in fees results in a close *now*
15:58 < BlueMatt[m]> or do you think that we need to because that's an indication (or at least may be an indication) that its gonna stay high?
16:49 < jeremyrubin> I think close reopen is the only safe option given if the time issue is in play then you need to fully revoke old states right
16:53 < BlueMatt[m]> I mean it’s lightning pre-package-relay, no option is actually safe, only shades of grey
17:55 < jeremyrubin> Haha. Presuming we want the code now to be the right model for later. But fair point, don't use lightning at all
18:00 < BlueMatt[m]> The right model for later is “remove all in-band fee updates and byo fee”, so it’s just a temporary thing either way.
18:03 < jeremyrubin> I agree. How do you feel on tx sponsors 😂
18:04 < jeremyrubin> I think we'll eventually have some primitives to help with it but I'm not bullish on how long it will take to recognize and solve the problem
18:19 < BlueMatt[m]> Package relay by itself basically fixes the issues. We don’t need consensus changes for this, luckily.
18:28 < jeremyrubin> Most of the stuff I've seen still has pinning issues, any links to the latest research?
18:29 < BlueMatt[m]> I’m not aware of any pinning issues in package relay, given reasonable limits and some kind of similar carve out(s) as we have now.
18:31 < BlueMatt[m]> But, we’ll have to get there and find out. It’s always a bit of a game of whack-a-mole with policy stuff. New frontiers and all that…
21:06 -!- trev [~trev@user/trev] has joined #bitcoin-rust
--- Log closed Wed Dec 29 00:00:15 2021