< bitcoin-git>
bitcoin/master 8455e36 practicalswift: [test] Avoid reading a potentially uninitialized variable in tx_invalid-test...
< bitcoin-git>
bitcoin/master 19be26a MarcoFalke: Merge #9555: [test] Avoid reading a potentially uninitialized variable in tx_invalid-test (transaction_tests.cpp)...
< bitcoin-git>
[bitcoin] MarcoFalke closed pull request #9555: [test] Avoid reading a potentially uninitialized variable in tx_invalid-test (transaction_tests.cpp) (master...avoid-ub-in-tx_invalid-test) https://github.com/bitcoin/bitcoin/pull/9555
< gmaxwell>
dear god, UIs are hard.
< gmaxwell>
19:02 < pawn> gmaxwell: I have the new version. I opened the console. I ran dumpwallet, and it's not giving anything back to me. I passed wallet.dat and /Users/pawn/Library/Application\ Support/Bitcoin/wallet.dat as an argument.
< sipa>
:'(
< gmaxwell>
I had him back up first.
< gwillen>
that is ... not a failure mode I ever would have thought of. Users find very creative ways to hurt themselves. I'm glad you had him back up.
< gmaxwell>
saw things similar prior. (button for a most common symbol in my words isn't working)
< achow101>
Does anyone else have a compiler warning for wallet/db.cpp:620?
< phantomcircuit>
achow101, compiler warnings are hard to reproduce because they all have different ideas of what to warn about
< gmaxwell>
achow101: whats the warning?
< gmaxwell>
if it's a shadow warning, mention it on #9911
< achow101>
looks like paveljanik already got it and fixed it
< paveljanik>
yup :-)
< bitcoin-git>
[bitcoin] practicalswift opened pull request #9936: [trivial] Fix three typos introduced into walletdb.h in yesterday's merge of commit 7184e25 (master...fix-typos-introduced-in-commit-7184e25c) https://github.com/bitcoin/bitcoin/pull/9936
< bitcoin-git>
[bitcoin] laanwj opened pull request #9937: rpc: Prevent `dumpwallet` from overwriting files (master...2017_03_walletdump_nooverwrite) https://github.com/bitcoin/bitcoin/pull/9937
< bitcoin-git>
[bitcoin] laanwj closed pull request #9928: Allow verify-commit.sh to just verify the HEAD commit (Use non-recursive verification by default) (master...2017/03/vc_simple) https://github.com/bitcoin/bitcoin/pull/9928
< wumpus>
going to tag 0.14.0 final in a bit
< wumpus>
gah, the github-merge script is not working anymore
< bitcoin-git>
bitcoin/master f20e664 Matt Corallo: Check gpg version before setting --weak-digest
< bitcoin-git>
bitcoin/master b3ec305 Matt Corallo: Fix bashisms in verify-commits and always check top commit's tree
< bitcoin-git>
bitcoin/master 00c13ea Wladimir J. van der Laan: Merge #9932: Fix verify-commits on travis and always check top commit's tree...
< bitcoin-git>
[bitcoin] laanwj closed pull request #9932: Fix verify-commits on travis and always check top commit's tree (master...2017-03-fix-verify-commits) https://github.com/bitcoin/bitcoin/pull/9932
< bitcoin-git>
bitcoin/master c6b82d1 Russell Yanofsky: Add tests for CWalletTx::nTimeSmart
< bitcoin-git>
bitcoin/master 1f98abe Russell Yanofsky: Factor out CWallet::nTimeSmart computation into a method....
< bitcoin-git>
bitcoin/master 6c996c2 Russell Yanofsky: Add documentation describing CWallet::nTimeSmart....
< morcos>
wumpus: #9602 has 5+ ACK's can we go ahead and merge while its clean? The only quibbling seems to be about more comment cleanups and some outstanding lack of consensus on breaking priortisetransaction API. That change can always be modified later.
< BlueMatt>
wumpus: hmm, the travis error on 309bf16 i can reproduce - I dont match your sha512 tree
< bitcoin-git>
[bitcoin] TheBlueMatt opened pull request #9940: Fix verify-commits on OSX, update for new bad Tree-SHA512, point travis to different keyservers (master...2017-03-verify-commits-no-recursion) https://github.com/bitcoin/bitcoin/pull/9940
< wumpus>
morcos: ok
< wumpus>
BlueMatt: I've done nothing special, just used github-merge.py, no clue why that SHA512 wouldn't match :/
< wumpus>
BlueMatt: maybe we should revert the SHA512 change until we have a clue what is happening here
< wumpus>
let's see if I can reproduce it myself, anyhow
< wumpus>
309bf16 doesn't even have a SHA512
< wumpus>
it was a quick fixup (without github-merge) to get github-merge to work again
< achow101>
github as this review feature where you can submit all of your comments as one review and that goes out to recipients as one email. I ask that people please use that to avoid spamming people's inboxes with emails for every single comment you make
< achow101>
s/as this/has this/
< BlueMatt>
achow101: sadly you cant respond in batch
< BlueMatt>
so you get a batch review, and then a million response emails :(
< BlueMatt>
also, I'd generally recommend a threaded email client to fix this issue :P
< achow101>
well that's annoying then.
< sipa>
you can't respond in batch?
< sipa>
really?
< BlueMatt>
sipa: i dont think so? Did I miss something?
< BlueMatt>
maybe i did...hmmm
< achow101>
BlueMatt: if you respond in the "view changes" tab, you can respond in a batch with the "start a review" button when you click to reply
< kanzure>
should decoderawtransaction calculate total input amounts?
< sipa>
no
< kanzure>
is there another way for an rpc user to do basic sanity checks on transactions like that?
< sipa>
if they have the inputs, sure
< sipa>
which should be trivial if it's their own
< sipa>
and nearly impossible otherwise
< sipa>
(would require a fully indexed blockchain)
< kanzure>
which rpc command though?
< achow101>
kanzure: there was a proposal for a verifyrawtransaction but it was controversial so it got dropped
< achow101>
the only way to check if a transaction is valid is to send it
< sipa>
kanzure: if you have -txindex, use getrawtransaction
< sipa>
kanzure: otherwise, don't
< kanzure>
getrawtransaction seems to only take txid
< kanzure>
oh you mean check inputs via getrawtransaction
< sipa>
yes, you fetch all the inputs
< kanzure>
ok ok fine, thanks
< gmaxwell>
we need to specify a new raw transaction format that can carry inputs... but without segwit it's insufficient to verify things.
< sipa>
kanzure: or use gettransaction
< luke-jr>
"(note: this is a temporary file, to be added-to by anybody, and moved to release-notes at release time)" made it into the final tag XD
< sipa>
(which only works if it's your own transaction)
< gmaxwell>
but it should be segwit only because of that.
< kanzure>
context: i am doing a simple user interface for a 8-of-12 signing ritual, checking some other nodes to make sure the total inputs add up to something reasonable and expected
< gmaxwell>
top
< kanzure>
sipa: no, iterating over inputs via getrawtransaction is good enough for now. also giving inputs in the transaction format sounds good to me.
< sipa>
kanzure: still, it requires a full txindex node
< sipa>
kanzure: unless all the outputs are available
< sipa>
which i guess will be the case for you
< sipa>
s/outputs are available/inputs are unspent/
< luke-jr>
hm, we lost bitcoin.conf.5 (manpage)
< BlueMatt>
any idea in what pr?
< BlueMatt>
its there there in contrib/debian
< bitcoin-git>
[bitcoin] jnewbery opened pull request #9945: Improve logging in bctest.py if there is a formatting mismatch (master...improve_bctest_logging) https://github.com/bitcoin/bitcoin/pull/9945
< luke-jr>
BlueMatt: not in 0.14/master?
< BlueMatt>
oh, nvm, I'm thinking of contrib/debain/examples/
< bitcoin-git>
[bitcoin] pinheadmz opened pull request #9946: Fix build errors if spaces in path or parent directory (master...dirspaces2) https://github.com/bitcoin/bitcoin/pull/9946