20:09:08 <bitconner> #startmeeting 20:09:08 <lightningbot> Meeting started Mon Sep 16 20:09:08 2019 UTC. The chair is bitconner. Information about MeetBot at http://wiki.debian.org/MeetBot. 20:09:08 <lightningbot> Useful Commands: #action #agreed #help #info #idea #link #topic. 20:09:13 <rusty> ... still asking (someone?) to join hangout apparently... 20:10:12 <t-bast> censorship? in this space? how can that be 20:12:13 <bitconner> #topic static remote key https://github.com/lightningnetwork/lightning-rfc/pull/642 20:18:53 <t-bast> am I the only one that hear only half what conner says? 20:19:04 <sstone> same here 20:19:54 <rusty> ARguing over game theory at the moment. Symmetry is better, but the conditions where you want to game the peer into closing is pretty narrow: if you can talk to them, you can just shutdown. If you can't, it's hard to make them unilateral close. 20:21:47 <bitconner> slow internetz, sorry :( 20:23:10 <t-bast> no problem, we'll interpolate the missing parts ;) 20:24:11 <rusty> bitconner reports roasbeef says c-lightning interop worked. W00t! 20:29:16 <rusty> Seems consensus is that option_simplified_commitment should consider symmetry for CSV (with constrained set of allowed CSV values). 20:29:25 <bitconner> #agreed push static remote key through as is, defer discussion of csv to option simplified commitment 20:30:22 <bitconner> #agree to merge pending roasbeef 20:30:26 <t-bast> ACK 20:30:31 <cdecker> ACK 20:30:48 <rusty> ack!! 20:31:18 <sstone> ack :) 20:32:04 <bitconner> #topic extending gossip queries https://github.com/lightningnetwork/lightning-rfc/pull/557 20:32:24 <bitconner> ack 20:33:19 <rusty> ack! 20:34:34 <t-bast> ack 20:36:11 <rusty> w00t! 20:36:48 <bitconner> #agree merge!!!! 20:37:18 <bitconner> #agreed mergee!!! 20:37:34 <t-bast> yay! 20:38:26 <bitconner> #topic flat feature bits https://github.com/lightningnetwork/lightning-rfc/pull/666 20:41:35 <roasbeef> rusty: re csv, we already stripped down the proposal to be as minimal as posisble so we can deploy 20:41:44 <bitconner> concept ACK, lprevent feature bit collision and can retroactively modify where feature bits are advertised 20:41:47 <rusty> roasbeef: agreed. 20:42:17 <rusty> bitconner: OK, will change mbz to allow non-zero, but mark it deprecated. 20:42:19 <bitconner> roasbeef, yeah we agreed to defer csv discussion to simplified commitment 20:42:33 <t-bast> ACK on #666, pending the change to allow for non-0 global features since master branches of eclair and lnd currently advertize it 20:42:44 <rusty> (Eclair and LND advertize global today in init in their master) 20:43:35 <roasbeef> what's teh feature bit change? 20:44:01 <t-bast> flattening everything 20:44:12 <t-bast> not separating between local / global 20:44:14 <bitconner> enforce that global features in init should always be zero-length, and put them all in current local field 20:44:27 <rusty> roasbeef: single feature field in init, each feature decides how and where it's advertized (init, node_announcement, channel_announcement). 20:47:09 <roasbeef> must this be done before rolling out the static key commitmetn stuff? feels liek it'll delay the deployment of a pretty critical safety improvement even further 20:47:37 <rusty> roasbeef: doesn't affect it AFAICT? 20:47:42 <roasbeef> this is mostly cosmetic 20:47:46 <roasbeef> well it would become a dep 20:47:54 <bitconner> #action continue work on PR, general agreement with direction 20:48:28 <bitconner> #topic base amp https://github.com/lightningnetwork/lightning-rfc/pull/643 20:53:01 <bitconner> all payments in the future will be amp or mpp payments and being dprecating current style of single-shot payments 20:53:20 <t-bast> agreed 20:53:34 <bitconner> allows bundling of invoice secret into mpp proposal and invoice changes 20:53:51 <bitconner> amp tlv payloads can just be a further extension of mpp fields 20:53:52 <t-bast> SGTM, feel free to rename it too, I just wanted to get the ball rolling on adding that field 20:54:07 <rusty> Yep, so the same feature bit means both. Also spec should spell out the implication that it means the final node accepts TLV, even if you didn't see a node_announcement. 21:03:00 <bitconner> #action rusty to think through overpayment scenarios 21:03:26 <rusty> Discussion was on whether we want to keep fuzzing amounts with MPP or not.. 21:04:44 <rusty> #action rusty to rebase 643 21:06:58 <rusty> Note: Propose meeting in 4 weeks moved earlier a few days, to 5pm Friday Berlin time, so we can do this face-to-face at TheLightningConference.com. 21:10:38 <bitconner> #action continue discussion of single tlv or split on pr 21:10:45 <bitconner> #endmeeting