Fix: Change WitnessTransaction inv type to use txid #424
Conversation
Codecov Report
@@ Coverage Diff @@
## master #424 +/- ##
=======================================
Coverage 86.71% 86.71%
=======================================
Files 40 40
Lines 8689 8689
=======================================
Hits 7535 7535
Misses 1154 1154
Continue to review full report at Codecov.
|
|
Hmm, @apoelstra this is clearly a bug. How easy is it for you to make patch releases forking from the release point? |
|
If you want, I could make a release branch and quickly try backport all the non-backwards incompatible changes to it to make a minor release. Not sure it's worth it for us, though. |
|
Just saw that there are others waiting for a next major release, though. |
|
Yeah, it was my bad messing with that when was doing hashtype system...
... and others waiting: |
…and sign_ecdsa_recoverable_with_noncedata f93ca81 Add sign_ecdsa_with_noncedata and sign_ecdsa_recoverable_with_noncedata (junderw) Pull request description: Fixes rust-bitcoin#424 As discussed on [IRC](https://gnusha.org/bitcoin-rust/2022-03-19.log) (starts at 09:14). These methods will allow for users to generate multiple signatures with the same private key and message by utilizing one of the `Variants` mention in RFC6979 which is exposed by libsecp256k1 via the `noncedata` argument. The reasoning behind adding this is to allow our library to migrate from using the -sys crate. Currently we support using this noncedata argument, and would like to continue doing so while at the same time migrating away from -sys crate. ACKs for top commit: apoelstra: ACK f93ca81 Tree-SHA512: 494d4f9046960779e199b18ff908fe74feda66a5cfc066c9ae6f3836fcaabd56defaa2138a913b25f1af3aa7dd48986e058804223224b76b303837c0c7adbaed
From BIP144: "MSG_WITNESS_TX getdata requests should use the non-witness serialized hash."