20:38 <jb55> 15x is a lot!

20:38 <jb55> assuming its doing everything correctly

20:43 <darosior> This measurement is using their equivalent of -assumevalid

20:47 <achow101> "Note that for blocks below the milestone (see below for terminology), Libbitcoin will skip transaction validation (besides checking they were properly committed, i.e. no malleation of either transactions or witnesses). This is in my opinion a similar threat model to Bitcoin Core’s -assumevalid." <-- that seems to be quite different?

20:48 <achow101> no validation at all vs checking everything but scripts

20:48 <gmaxwell> its making 15x the network connections and it sounds like it also does agressive connection management to cull the slowest peers (instead of just stallers)

20:49 <gmaxwell> darosior: an idea what the performance difference is from just pulling from a couple lan-attached fast peers?

21:16 <bitcoin-git> [gui-qml] hebasto pushed 6 commits to main: https://github.com/bitcoin-core/gui-qml/compare/88ce5253a4a3...c117acc07ffa

21:16 <bitcoin-git> gui-qml/main fad9bd7 jarolrod: qml: signal based navigation in onboarding

21:16 <bitcoin-git> gui-qml/main 02a3b25 jarolrod: qml: use signal based navigation in settings with subpages

21:16 <bitcoin-git> gui-qml/main a77c7f1 jarolrod: qml: use signal based navigation in CreateWallet pages, introduce wizard

21:16 <bitcoin-git> [gui-qml] hebasto merged pull request #427: Introduce Signal Based Navigation Model to have Self-Contained Pages (main...contain-pages) https://github.com/bitcoin-core/gui-qml/pull/427

21:20 <bitcoin-git> [gui-qml] hebasto opened pull request #429: ci, gha: Update `actions/checkout` actions (main...241028-ci) https://github.com/bitcoin-core/gui-qml/pull/429

23:51 <darosior> achow101: the threat model isn't materially different i think. Just that for us it'd not be a big per improvement to validate out of order since we still need to update the UTxO set.

23:54 <darosior> gmaxwell: i think willcl-ark and others have been benching this, albeit with a much faster machine. I'd expect parallel validation (+ validation in parallel to download itself) would still make for a substantial improvement.

23:54 <darosior> achow101: s/per improvement/perf improvement/

23:56 <bitcoin-git> [bitcoin] ryanofsky opened pull request #31174: tinyformat: Add compile-time checking for literal format strings (master...pr/tcheck) https://github.com/bitcoin/bitcoin/pull/31174