< bitcoin-git>
bitcoin/master 5f019d5 251: Removes the boost/algorithm/string/join dependency...
< bitcoin-git>
bitcoin/master eeae471 MarcoFalke: Merge #13726: Utils and libraries: Removes the boost/algorithm/string/join dependency...
< bitcoin-git>
[bitcoin] MarcoFalke closed pull request #13726: Utils and libraries: Removes the boost/algorithm/string/join dependency (master...patch/remove_boost_join_dependency) https://github.com/bitcoin/bitcoin/pull/13726
< face>
Is everyting boost-related supposed to go? What about stuff that's not in c++11, like signals2?
< wumpus>
face: in the long run, that's the goal
< face>
That would mean reimplementing signals2 functionality? Why?
< wumpus>
some things that are not in c++11 will be in c++17, or later revisions
< wumpus>
we're not going for the approach of directly reimplementing signals2
< wumpus>
but the more traditional C++ way of using notification interfaces
< wumpus>
this move has already been underway, not even so much to get rid of boost, but because signal2 has some performance and debuggability issues
< face>
Yeah, I'm all for replacing boost:: with std:: but something like signals2 would probably never be in the standard and replacing it would be quite difficult/error-prone?
< bitcoin-git>
bitcoin/master e8c7434 MarcoFalke: Merge #13683: wallet: Introduce assertion to document the assumption that cache and cache_used are always set in tandem...
< bitcoin-git>
[bitcoin] MarcoFalke closed pull request #13683: wallet: Introduce assertion to document the assumption that cache and cache_used are always set in tandem (master...null-pointer-dereference) https://github.com/bitcoin/bitcoin/pull/13683
< wumpus>
face: I think I've already done so, above
< face>
wumpus: what would be a "more traditional C++ way" of doing thread-safe signals/slots system, beside rolling your own?
< face>
Just to be clear, I'm not trolling or anything. Just want to know wheather this has been discussed and where I can read up on it.
< wumpus>
only very little registering/unregistering happens, so protecting it with a mutex is sufficient
< wumpus>
that might not even be needed
< wumpus>
(e.g. if registeration is limited to init/shutdown)
< wumpus>
in any case, if you're really so interested in this, you might want to ask BlueMatt, he did some things in that direction
< bitcoin-git>
bitcoin/master 1be5c33 Ben Woosley: Avoid unnecessarily setting env variables on the lint build...
< bitcoin-git>
bitcoin/master 515348f Ben Woosley: Don't unnecessarily install shellcheck on the lint build...
< bitcoin-git>
bitcoin/master 751c958 Ben Woosley: Use the Travis python language feature on the lint build...
< wumpus>
go MarcoFalke \o/
< bitcoin-git>
[bitcoin] MarcoFalke closed pull request #13729: travis: Avoid unnecessarily setting env variables on the lint build (master...ci-lint-env) https://github.com/bitcoin/bitcoin/pull/13729
< face>
wumpus: Ok, thanks
< bitcoin-git>
[bitcoin] DrahtBot reopened pull request #12984: logs: Make the columns in subsequent UpdateTip log entries horizontally aligned (master...attention-to-detail) https://github.com/bitcoin/bitcoin/pull/12984
< bitcoin-git>
[bitcoin] DrahtBot closed pull request #12825: Only allocate a LevelDB block cache if LevelDB will actually use it (master...buffer-cache) https://github.com/bitcoin/bitcoin/pull/12825
< fanquake>
It would be good if the Drahtbot was a bit smarter about commenting on PRs. i.e It's a bit spammy how it basically comments on every depends PR to say that they conflict with each other.
< fanquake>
It also means the GH UI (if you use it) becomes fairly useless as any sort of gauge to what's recently happened in bitcoin/bitcoin.
< bitcoin-git>
bitcoin/master ac8a1d0 Conor Scott: [RPC] Remove field in getblocktemplate help that has never been used
< bitcoin-git>
bitcoin/master 10b9a81 MarcoFalke: Merge #12764: doc: Remove field in getblocktemplate help that has never been used....
< bitcoin-git>
[bitcoin] MarcoFalke closed pull request #12764: doc: Remove field in getblocktemplate help that has never been used. (master...remove_unused_gbt_field) https://github.com/bitcoin/bitcoin/pull/12764
< bitcoin-git>
bitcoin/master 01a06d6 practicalswift: Avoid locking mutexes that are already held by the same thread
< bitcoin-git>
bitcoin/master 05714f9 MarcoFalke: Merge #11762: Avoid locking mutexes that are already held by the same thread...
< bitcoin-git>
[bitcoin] MarcoFalke closed pull request #11762: Avoid locking mutexes that are already held by the same thread (master...remove-double-locks) https://github.com/bitcoin/bitcoin/pull/11762
< MarcoFalke>
> [16:44] <sipa> MarcoFalke: what is the time limit for DrahtBot to trigger a close/reopen?
< bitcoin-git>
[bitcoin] fanquake opened pull request #13741: Only allocate a LevelDB block cache if LevelDB will actually use it (master...eklitzke-buffer-cache-rebased) https://github.com/bitcoin/bitcoin/pull/13741
< bitcoin-git>
[bitcoin] fanquake closed pull request #12825: Only allocate a LevelDB block cache if LevelDB will actually use it (master...buffer-cache) https://github.com/bitcoin/bitcoin/pull/12825
< wumpus>
MarcoFalke: there's a compatibility issue with the new gitian build IIRC
< MarcoFalke>
I believe jonasschnelli solved it by compiling lxc 2.1.1. on the debian host
< wumpus>
oh!
< MarcoFalke>
>[13:47] <luke-jr> Surely there's another way to trigger a Travis rebuild?
< MarcoFalke>
Not that I am aware of
< MarcoFalke>
At least not in any way that uses the current travis.yml from master
< MarcoFalke>
> [09:33] <fanquake> It would be good if the Drahtbot was a bit smarter about commenting on PRs. i.e It's a bit spammy how it basically comments on every depends PR to say that they conflict with each other.
< MarcoFalke>
I am aware of that, but haven't found a good solution
< wumpus>
17:36 < sipa> how long was i offline? <- sounds like sipa has blackouts
< wumpus>
I have the same problem as jonasschnelli , my build host is debian jessie, not sure whether I'm going the manual-build LXC route too or bite the bullet and upgrade it to stretch
< masonicboom>
Is there a clean way to automatically detect code changes that affect consensus? Is all consensus code localized to src/consensus/? I'm looking to create a Twitter bot that broadcasts all changes and PRs that could alter consensus rules.
< sipa>
masonicboom: no
< sipa>
it
< sipa>
also validation, script/script, script/interpreter, coins, txdb, pubkey, versionbits, uint256, arith_uint256, chain, compressor, dbwrapper, hash, pow, prevector, checkqueue, ... and probably more
< sipa>
some things in crypto/
< sipa>
and libraries secp256k1 and leveldb
< sipa>
and of course your system's c and c++ libraries, kernel, hardware, ...
< masonicboom>
Ok thanks. From a glance it didn't seem like consensus could be all of it.
< masonicboom>
Think it's worthwhile to try making the bot? The idea is to make node operators better informed. So they don't just blindly trust core development, which makes core committers a ripe target for corruption
< sipa>
you'd need to report nearly every pull request...
< sipa>
in an ideal world consensus code would be so well encapsulated that there are basically no changes to that code apart from actual intentional consensus changes
< gmaxwell>
masonicboom: it's also really hard because there are many trivial changes that 'change' near consensus code but couldn't realistically have an effect. E.g. rearranging the order of #includes
< jonasschnelli>
wumpus: I had to compile LXC 2.1.1... which was hasselfree.
< jonasschnelli>
MarcoFalke: my build loop script was disabled...nighly build should compile again now.
< jonasschnelli>
*builds
< MarcoFalke>
thx!
< bitcoin-git>
[bitcoin] DrahtBot reopened pull request #13310: Report progress in ReplayBlocks while rolling forward (master...2018-05-replayblocks-progress) https://github.com/bitcoin/bitcoin/pull/13310
< bitcoin-git>
[bitcoin] DrahtBot reopened pull request #13280: [qt] Removed "Pay only the required fee" checkbox (master...custom-fee) https://github.com/bitcoin/bitcoin/pull/13280
< bitcoin-git>
[bitcoin] DrahtBot closed pull request #13247: Add tests to SingleThreadedSchedulerClient() and document the memory model (master...scheduler-tests) https://github.com/bitcoin/bitcoin/pull/13247
< bitcoin-git>
[bitcoin] DrahtBot closed pull request #13084: Do not turn OP_1NEGATE in scriptSig into 0x0181 in signing code (master...201804_keepnegone) https://github.com/bitcoin/bitcoin/pull/13084
< bitcoin-git>
[bitcoin] DrahtBot closed pull request #12911: wallet: Show fee in results for signrawtransaction* when known (master...sign-show-fees) https://github.com/bitcoin/bitcoin/pull/12911
< bitcoin-git>
[bitcoin] DrahtBot reopened pull request #12676: Show "bip125-replaceable" flag, when retrieving mempool entries (master...rpc-raw-replaceable-flag) https://github.com/bitcoin/bitcoin/pull/12676
< bitcoin-git>
[bitcoin] DrahtBot reopened pull request #12096: [rpc] [wallet] Allow specifying the output index when using bumpfee (master...better-bumpfee) https://github.com/bitcoin/bitcoin/pull/12096