Sebastian Falbesoner
Sebastian Falbesoner
Concept ACK, thanks for picking this up.
Rebased on master (necessary since #26606 was merged and also touched the same functional test file).
Thanks for the review! Rebased on master (as I suspect CI would be very unhappy otherwise) and addressed comments https://github.com/bitcoin/bitcoin/pull/30125#discussion_r1813809067, https://github.com/bitcoin/bitcoin/pull/30125#discussion_r1813825732 and https://github.com/bitcoin/bitcoin/pull/30125#discussion_r1813830674
Concept ACK Now that `encode_op_n` handles also -1, the following call-site in `CScript.__coerce_instance` can be simplified: https://github.com/bitcoin/bitcoin/blob/31be1a47675e4449f856e61beb2b4bfc228ea219/test/functional/test_framework/script.py#L446-L450
Concept ACK In order to avoid changes in functional tests and keep the patch small and simple, it probably would makes sense to set `LEAF_VERSION_TAPSCRIPT` as default value for parameter...
@LarryRuane: Interesting problem (and very creative solution with the RBF tx :)). Fwiw, on one of my machines running OpenBSD 7.5, the same behaviour as you described occurs on commit...
Suggestion for adding corresponding `decodescript` test coverage: Patch ```diff diff --git a/test/functional/rpc_decodescript.py b/test/functional/rpc_decodescript.py index f37e61ab50..a105b3d3b4 100755 --- a/test/functional/rpc_decodescript.py +++ b/test/functional/rpc_decodescript.py @@ -187,6 +187,16 @@ class DecodeScriptTest(BitcoinTestFramework): assert_equal('1 ' + xonly_public_key,...
Concept ACK
@laanwj: Fair points! My reasoning to consider this (somewhat hacky) "import code from test framework" approach as acceptable was the fact that we already do it in other contrib scripts...
I can reproduce this issue locally by passing the randomseed as shown in the CI logs, i.e. ``` $ ./build/test/functional/feature_bip68_sequence.py --randomseed 6169832640268785903 ``` Seems that the number of UTXOs generated...