03:18 < bitcoin-git> [bitcoin] jnewbery opened pull request #18401: Refactor: Initialize PrecomputedTransactionData in CheckInputScripts (master...2020-03-precomputedtransactiondata) https://github.com/bitcoin/bitcoin/pull/18401

06:55 < meshcollider> promag: done, thanks

11:39 < yevaud> as a thought experiment mostly, what is the most compact encoding of a block header that is stateless?

11:40 < yevaud> the version is mostly random due to asicboost so there's no clear wins to be had there.

11:41 < kallewoof> Travis gave me a single log line "An error occurred while generating the build script." and failed my build for some reason..

11:42 < yevaud> we save a little over 9 bytes if we run length encode the previous block hash as a 4 bit integer.

11:43 < aj> yevaud: https://lists.linuxfoundation.org/pipermail/bitcoin-dev/2018-March/015851.html ?

11:43 < yevaud> we optionally save 3 bytes if we assume knowledge of the previous timestamp. we optionally save 32 bytes if we assume knowledge of the previous header.

11:45 < yevaud> aj: interesting, a few assumptions in that document are violated today.

11:45 < yevaud> "the version often matches that of a recent ancestor block"

11:48 < yevaud> the main one it looks like I'd forgotten is that the difficulty changes infrequently, but that's assuming some amount of state as well. I was thinking you could have the client recover that value, but it clearly can't know the resulting difficulty without the input to the function.

11:49 < yevaud> ie, if a header you receive is discontinuous *and* on a difficulty change boundary, you don't know the value. you also don't know the height for an out of order and non connectable header, unfortunately.

11:50 < yevaud> though that said, you know the height implicitly due to the unique difficulty, unless it is on a fork *and* on the difficulty change boundary.

12:09 < bitcoin-git> [bitcoin] jonatack opened pull request #18402: gui: display mapped AS in peers info window (master...gui-peers-display-mapped_as) https://github.com/bitcoin/bitcoin/pull/18402

12:09 < bitcoin-git> [bitcoin] vova557 opened pull request #18403: litecoin (master...master) https://github.com/bitcoin/bitcoin/pull/18403

12:20 < bitcoin-git> [bitcoin] MarcoFalke closed pull request #18403: Биткоин (master...master) https://github.com/bitcoin/bitcoin/pull/18403

14:50 < bitcoin-git> [bitcoin] hebasto opened pull request #18404: build: Drop unused ${WRAP_DIR}/${HOST} directory (master...20200322-gitian-win) https://github.com/bitcoin/bitcoin/pull/18404

18:08 < sipa> Anyone know what is causing the "An error occurred while generating the build script." for the macOS 10.14 native Travis job?

18:09 < sipa> eg here https://travis-ci.org/github/bitcoin/bitcoin/jobs/665432889?utm_medium=notification&utm_source=github_status

18:13 < hebasto> sipa: it seems it started just after their maintenance...

18:49 < bitcoin-git> [bitcoin] hebasto opened pull request #18405: build: Drop ZeroMQ patch for Mingw-w64 < 4.0 (master...20200322-zmq-mingw) https://github.com/bitcoin/bitcoin/pull/18405

19:35 < promag> hebasto: mind taking a look at #18338?

19:35 < gribble> https://github.com/bitcoin/bitcoin/issues/18338 | Fix wallet unload race condition by promag · Pull Request #18338 · bitcoin/bitcoin · GitHub

19:36 < hebasto> promag: doing right now, not sure if could complete it today

21:00 < promag> hebasto: I'd be happy to address a partial review

22:35 < bitcoin-git> [bitcoin] theStack opened pull request #18406: test: add executable flag for rpc_estimatefee.py (master...20200322-test-add-executable-flag-for-rpc-estimatefee) https://github.com/bitcoin/bitcoin/pull/18406