< bitcoin-git> [bitcoin] 34ro opened pull request #9744: Remove unused module from rpc-tests (master...remove-unused-modules-from-rpc-tests) https://github.com/bitcoin/bitcoin/pull/9744
< bitcoin-git> [bitcoin] 34ro opened pull request #9745: [RPC] Getting confirmations command (master...add-getconfirmations-to-rpc) https://github.com/bitcoin/bitcoin/pull/9745
< bitcoin-git> [bitcoin] kobake opened pull request #9747: Vs2015 (master...vs2015) https://github.com/bitcoin/bitcoin/pull/9747
< bitcoin-git> [bitcoin] kobake closed pull request #9747: Vs2015 (master...vs2015) https://github.com/bitcoin/bitcoin/pull/9747
< arubi> I think I'm building the crediting transaction from script_tests.json wrong. I can pass tests that don't have a checksig in them, but (e.g.) a p2pk test fails with a bad signature. here's an example of both cases: https://paste.fedoraproject.org/555262/92264814/raw/ . the tests describe the crediting tx structure right at the beginning of the script_tests.json file. would appreciate your input.
< bitcoin-git> [bitcoin] vosa88 opened pull request #9748: config.txt (master...patch-1) https://github.com/bitcoin/bitcoin/pull/9748
< bitcoin-git> [bitcoin] fanquake closed pull request #9748: config.txt (master...patch-1) https://github.com/bitcoin/bitcoin/pull/9748
< jonasschnelli> is std::unique_ptr<Secp256k1Init> ecc; not equivalent to ecc.reset(new Secp256k1Init());
< sipa> jonasschnelli: no
< jonasschnelli> I know I miss something... but I wonder what
< sipa> jonasschnelli: a unique_ptr is initialized to nullptr by default
< jonasschnelli> I stepped it in lldb and got a non nullptr after std::unique_ptr<Secp256k1Init> ecc;
< jonasschnelli> But maybe a -O0 issue?
< sipa> jonasschnelli: you must be doing something wrong :)
< jonasschnelli> heh.. okay. I see the point now.
< jonasschnelli> Ah. Yes. __first_ = 0x0000000000000000
< jonasschnelli> The question then remains, how could this have worked before your PR. :)
< sipa> because the secp library does right now not actually need a check the ctx pointer ther
< sipa> but it declares that argument is required to be nonnull
< sipa> so with sanitizers on, that's enforced at runtime