< bitcoin-git>
bitcoin/master 0089905 practicalswift: Add compile time checking for cs_main locks which we assert at run time
< bitcoin-git>
bitcoin/master 5473e25 MarcoFalke: Merge #14444: Add compile time checking for cs_main locks which we assert at run time...
< bitcoin-git>
[bitcoin] MarcoFalke closed pull request #14444: Add compile time checking for cs_main locks which we assert at run time (master...move-runtime-checking-to-compile-time-checking) https://github.com/bitcoin/bitcoin/pull/14444
< bitcoin-git>
[bitcoin] achow101 opened pull request #14558: rpc: Require solvability in importmulti if importing more than the scriptPubKey (master...importmulti-solvability) https://github.com/bitcoin/bitcoin/pull/14558
< promag>
ken2812221: instead of adding a generic GetRelativePath (just because we have to support boost 1.47), and since we need the relative path in the recursive_directory_iterator loop, I've removed the walletdir there
< bitcoin-git>
[bitcoin] practicalswift closed pull request #14560: wallet: Avoid calling uiInterface.LoadWallet(...) with cs_wallet held (master...uiInterface.LoadWallet-without-cs_wallet) https://github.com/bitcoin/bitcoin/pull/14560
< bitcoin-git>
bitcoin/master 9dcf6c0 Wladimir J. van der Laan: build: Add --disable-bip70 configure option...
< bitcoin-git>
bitcoin/master 38b9850 Wladimir J. van der Laan: qt: cleanup: Move BIP70 functions together in paymentserver...
< bitcoin-git>
bitcoin/master fbb643d James Hilliard: Add BIP70 deprecation warning
< bitcoin-git>
[bitcoin] laanwj closed pull request #14451: Add BIP70 deprecation warning and allow building GUI without BIP70 support (master...deprecate-bip70) https://github.com/bitcoin/bitcoin/pull/14451
< fanquake>
wumpus are we meant to be able to build bitcoin-qt without protobuf after #14451?
< gribble>
https://github.com/bitcoin/bitcoin/issues/14451 | Add BIP70 deprecation warning and allow building GUI without BIP70 support by jameshilliard · Pull Request #14451 · bitcoin/bitcoin · GitHub
< fanquake>
I'm unable to, and commented about 30s to late.. Although might be misunderstanding.
< fanquake>
np, cleared up by Marco
< luke-jr>
not building bitcoin-qt seems a bit extreme though; might make sense to refactor configure to just change the bip70 option
< bitcoin-git>
[bitcoin] promag closed pull request #13637: wallet: Add GetBalances to calculate all balances (master...2018-06-wallet-getbalance) https://github.com/bitcoin/bitcoin/pull/13637
< wumpus>
promag: I'm a bit scared of that one, let me see again
< promag>
wumpus: thanks
< wumpus>
it's one of those things that have been 'fixed' too many times
< luke-jr>
is there a reason we distribute tar.gz instead of tar.xz?
< luke-jr>
is a 10 GB blockchain benchmark significant?
< luke-jr>
or do I need to get real storage involved?
< gmaxwell>
luke-jr: tar.xz is probably a PITA to open on windows/mac/openbsd. and who cares if the download is a couple MB bigger when the sync going to be 250GB...
< luke-jr>
gmaxwell: well, they're Linux binaries inside
< gmaxwell>
there is also source inside.
< luke-jr>
a different tarball
< luke-jr>
anyway, I see 2.2% performance gain over the first 10 GB of blocks using BlueMatt's POWER8 optimisation stuff (on binaries built for POWER7)
< luke-jr>
is that significant?
< * luke-jr>
has no idea why -mcpu=power7 performs better than -mcpu=power9 on POWER9..
< sipa>
there may be a larger benefit to larger blocks
< gmaxwell>
luke-jr: the interesting question is did something break the speedup of sse4/avx over plain C.
< gmaxwell>
The microbenchmark on that power8 optimized sha2 is a comparable speedup to sse4 on x86.
< luke-jr>
well, even if there's another bottleneck, it probably helps free up CPU resources for other processes?
< BlueMatt>
heh could try running ibd with some other cpu-but-not-memory-intensive process running on the same core on another thread and see if you see a difference lol
< luke-jr>
hmm, interesting idea.
< gmaxwell>
or just profile an figure out why its not speeding things up. :P
< jnewbery>
Just running test_runner and the new progress output is thrilling. Glad we didn't go for the boring spinner option.
< instagibbs>
it should make some sort of progressive ascii art
< bitcoin-git>
[bitcoin] instagibbs opened pull request #14563: Pin flake8 version to earlier version to avoid pyflakes update (master...fix_flake) https://github.com/bitcoin/bitcoin/pull/14563
< bitcoin-git>
[bitcoin] instagibbs closed pull request #14563: Pin flake8 version to earlier version to avoid pyflakes update (master...fix_flake) https://github.com/bitcoin/bitcoin/pull/14563