< jeremyrubin> hm ok I know what the problem is more or less... now just to solve it :p
< jeremyrubin> CMutableTransaction tx(*block.vtx[0]);
< jeremyrubin> tx.vout.resize(1);
< jeremyrubin> block.vtx[0] = MakeTransactionRef(std::move(tx));
< jeremyrubin> That's the fix for it to work in TestChain100Setup
< jeremyrubin> (And then a call to GenerateCoinbaseComittment)
< jeremyrubin> Maybe I should pr that...
< bitcoin-git> [bitcoin] MarcoFalke opened pull request #9920: [0.14] Allow any subkey in verify-commits (BlueMatt) (0.14...Mf1703-verifySubkeys) https://github.com/bitcoin/bitcoin/pull/9920
< wumpus> would P2P over UNIX sockets be useful? it obviously can't be used for anything like address advertisement, but if you want to set up a local test network (as the tests do) with -connect/-addnode it may make sense
< wumpus> also maybe Tor hidden services could connect to us over a UNIX socket?
< Tinglysning> you guys are amazing. have a nice day.
< Tinglysning> cya :)
< bitcoin-git> [bitcoin] laanwj opened pull request #9921: build: Probe MSG_DONTWAIT in the same way as MSG_NOSIGNAL (master...2017_03_cloudabi_netcompat) https://github.com/bitcoin/bitcoin/pull/9921
< bitcoin-git> [bitcoin] laanwj pushed 2 new commits to master: https://github.com/bitcoin/bitcoin/compare/53c300fb525a...01b7cda91684
< bitcoin-git> bitcoin/master fa7da3f MarcoFalke: [contrib] gh-merge: Move second sha512 check to before signing
< bitcoin-git> bitcoin/master 01b7cda Wladimir J. van der Laan: Merge #9905: [contrib] gh-merge: Move second sha512 check to the end...
< bitcoin-git> [bitcoin] laanwj closed pull request #9905: [contrib] gh-merge: Move second sha512 check to the end (master...Mf1703-gh-merge) https://github.com/bitcoin/bitcoin/pull/9905
< bitcoin-git> [bitcoin] laanwj closed pull request #9920: [0.14] Allow any subkey in verify-commits (BlueMatt) (0.14...Mf1703-verifySubkeys) https://github.com/bitcoin/bitcoin/pull/9920
< MarcoFalke> wumpus: Sorry, I missed that #9884 needs backport as well.
< gribble> https://github.com/bitcoin/bitcoin/issues/9884 | An error has occurred and has been logged. Please contact this bot's administrator for more information.
< MarcoFalke> Oh wait, you already did.
< wumpus> yep I backported that one
< BlueMatt> wtf is the travis failure oon 0.14
< BlueMatt> wumpus: see #9880's latest commit which will fix 0.14 verify-commits
< gribble> https://github.com/bitcoin/bitcoin/issues/9880 | Verify Tree-SHA512s in merge commits, enforce sigs are not SHA1 by TheBlueMatt · Pull Request #9880 · bitcoin/bitcoin · GitHub
< MarcoFalke_lab> BlueMatt: As the script will now accept different subkeys, you might have to track the which subkey was revoked for revsigs.
< MarcoFalke_lab> I think the easiest solution is to just revert my pull on 0.14
< BlueMatt> hmm?
< BlueMatt> no, gpg knows whats revoked
< BlueMatt> and gpg prints the same VALIDSIG line whether it was signed by a revoked key or not
< MarcoFalke_lab> We don't need the commit for 0.14 as only laanwj commits on that branch and his key is all fine
< BlueMatt> yea, but pieter has merged things on that branch with a subkey that is now revoked?
< BlueMatt> (same commit as on the master branch)
< MarcoFalke_lab> Yeah, but that subkey is in trusted keys and all commits are already marked as trusted as well
< MarcoFalke_lab> We only need that commit if pieter uses a different subkey on 0.14
< BlueMatt> well we need the revsig checking fix either way
< BlueMatt> ?
< MarcoFalke_lab> verify-commits in case of a revsig, will only check against the hardcoded keys in trusted-keys
< MarcoFalke_lab> so if the keys are replaced by the root keys, it will no longer detect the revsig
< MarcoFalke_lab> BlueMatt: Jup, on master we need the fix at some time
< BlueMatt> ahh, es, ok, so travis wil pass by not backporting that fix....I mean, sure, but I'd rather fix it so that others can merge on 0.14?
< MarcoFalke_lab> I don't think there is a need for others to merge on 0.14 right now. Though you are still welcome to fix that until tomorrow :P
< BlueMatt> MarcoFalke_lab: well the commit i just pushed should fix it
< BlueMatt> or, it fixes the current failure for me
< wumpus> ok, going to try that
< wumpus> cherry-picked that commit to 0.14, looks like verify-commits is taking a very long time
< BlueMatt> it always does
< wumpus> ok
< BlueMatt> hum, failed
< BlueMatt> wtf
< BlueMatt> oh, you didnt push yet
< BlueMatt> heh, nvm
< BlueMatt> its massively slow with the --tree-checks option after the new pr
< BlueMatt> because git-cat is slow as fuck cat'ing every file :(
< BlueMatt> (other option is to do a checkout as you go and then hash, which is faster, but does a checkout in your repo which really sucks)
< gmaxwell> wumpus: p2p on domain sockets would be useful to avoid port conflicts when running test tnings. oh man, and tor using it would give a nice way of distinguishing tor HS without needing to listen on another port the a user might forget to firewall off.
< gmaxwell> (I think the biggest argument against having a HS listening port is that the user will not firewall it off and it'll be awful for privacy.)
< bitcoin-git> [bitcoin] SCDeveloper opened pull request #9924: [UI Styling] Left-Align Tab Bar (master...patch-1) https://github.com/bitcoin/bitcoin/pull/9924
< bitcoin-git> [bitcoin] SCDeveloper opened pull request #9925: [UI Styling] Left-Align Tab Bar (master...patch-2) https://github.com/bitcoin/bitcoin/pull/9925
< wumpus> BlueMatt: i didn't push yet, was waiting for it to succeed first. But it does: "There is a valid path from HEAD to 82bcf405f6db1d55b684a1f63a4aabad376cdad7 where all commits are signed!", so pushing
< bitcoin-git> [bitcoin] laanwj pushed 1 new commit to 0.14: https://github.com/bitcoin/bitcoin/commit/43a987c1cbcd04b9c59052606c3c31c56abc826e
< bitcoin-git> bitcoin/0.14 43a987c Matt Corallo: Fix regsig checking for subkey sigs in verify-commits...
< Victorsueca> is there going to be a rc4 so far or you're going for the final release unless some bug is found?
< sipa> if no bug is found, the rc becomes the final release
< luke-jr> why is it not sufficient that only the tip be signed?