💬 fanquake commented on pull request "Wallet: Zero out wallet master key upon locking so it doesn't persist in memory":
(https://github.com/bitcoin/bitcoin/pull/27080#issuecomment-1429473503)
Added this to #26878 for backporting to 24.x.

💬 MarcoFalke commented on pull request "verify-commits: Bump trusted git root to after most recent laanwj merge":
(https://github.com/bitcoin/bitcoin/pull/27076#issuecomment-1429480424)
lgtm, Approach ACK

I haven't tested this, because I use a trusted git root set by myself anyway, but I can't see a reason not to do this. This was also done last time in commit d4b3dc5b0a726cc4cc7a8467be43126e78f841cf, so it makes sense to do the same approach again.

💬 willcl-ark commented on pull request "Convert ArgsManager::GetDataDir to a read-only function":
(https://github.com/bitcoin/bitcoin/pull/27073#issuecomment-1429498959)
I don't particularly want to feature-creep too much, but another adjacent issue is https://github.com/bitcoin/bitcoin/issues/19990

Which could potentially be addressed in this PR too if desired?

📝 theStack opened a pull request: "script: remove unused bitwise `CScriptNum` operators"
(https://github.com/bitcoin/bitcoin/pull/27096)
These overloaded operators were introduced almost seven years ago in the implementation of OP_CHECKSEQUENCEVERIFY (BIP112, PR #7524, commit 53e53a33c939949665f60d5eeb82abbb21f97128), have never used since outside of fuzzing. One could argue to keep those around solely for consistency reasons, but I think its preferable to get rid of methods that are likely never used. Even the so-far only use of the & operator (checking the OP_CSV argument for the SEQUENCE_LOCKTIME_DISABLE_FLAG), could be remove
...

💬 stickies-v commented on pull request "wallet: SecureString to allow null characters":
(https://github.com/bitcoin/bitcoin/pull/27068#issuecomment-1429565498)
> I would be surprised if any password manager produced a passphrase that included non-typeable characters.

I googled "random password generator" and with the [second result](https://passwordsgenerator.net/) I was able to generate a null-character containing password (\039"`\9%|+']1+5) in less than a minute (disabled using letters to speed it up, but otherwise using default settings). It seems like indeed a fair amount of generators do not use backslashes as symbols (e.g. Bitwarden, Lastpass)
...

✳️ fanquake pushed commits to a branch: bitcoin/bitcoin:master
(https://github.com/bitcoin/bitcoin/compare/2c1fe27bf3c1...d6ef44cccbdf)
Merge bitcoin/bitcoin#27081: Modernize rpcauth.py

e4e17907b686c25dda91e569645a8f501ca48f90 Modernize rpcauth.py and its tests (Pieter Wuille)

Pull request description:

Use Python3 constructions, and f-strings.

ACKs for top commit:
jamesob:
Github ACK https://github.com/bitcoin/bitcoin/pull/27081/commits/e4e17907b686c25dda91e569645a8f501ca48f90

Tree-SHA512: 005573d967e04400fec727f45739f138879be703e692745c0a639272d37d221d230f388de23f2615cb954bb47179fb46e53da0410ae9f0865319b91bb2dc01f4

🚀 fanquake merged a pull request: "Modernize rpcauth.py"
(https://github.com/bitcoin/bitcoin/pull/27081)

💬 hebasto commented on pull request "guix: consolidate to glibc 2.27 for Linux builds":
(https://github.com/bitcoin/bitcoin/pull/27029#issuecomment-1429668279)
Guix builds:
```
851f4116c395a6b89475871a9763a111abad3951424276135931d80b1c09615e guix-build-d5d4b75840b4/output/aarch64-linux-gnu/SHA256SUMS.part
bb9d3867b8f7dae5f9f8c326aa1b0e9fa28964659b3c7187e743152044c9dbe8 guix-build-d5d4b75840b4/output/aarch64-linux-gnu/bitcoin-d5d4b75840b4-aarch64-linux-gnu-debug.tar.gz
e4c0ab4883993362aea104ea776102a5d5beaa39a22e9aa22fd27a81d36415dc guix-build-d5d4b75840b4/output/aarch64-linux-gnu/bitcoin-d5d4b75840b4-aarch64-linux-gnu.tar.gz
01b239015610aea0c47
...

💬 Ayms commented on issue "Allow several OP_RETURN in one tx and no limited size":
(https://github.com/bitcoin/bitcoin/issues/27043#issuecomment-1429678640)
@MarcoFalke super simple...

@Semisol as mentionned in the thread there are so many ways to store things that there is no rationale to impose a limit in OP_RETURN, this is cleary a handicap for bitcoin and future uses/evolutions, bitcoin can clearly do far better what ethereum and all the messy stuff around pretend to do but we need this change

💬 furszy commented on pull request "refactor: wallet, remove global 'ArgsManager' dependency":
(https://github.com/bitcoin/bitcoin/pull/26889#discussion_r1105772605)
np, small miscommunication. Done now, all of them tackled now. Thanks!

💬 Sjors commented on pull request "p2p, validation: Don't download witnesses for assumed-valid blocks when running in prune mode":
(https://github.com/bitcoin/bitcoin/pull/27050#issuecomment-1429695634)
Making witness skipping opt-in makes sense. In addition our GUI initial setup wizard could help the user with this (as it does with suggesting pruning).

💬 furszy commented on pull request "refactor: wallet, remove global 'ArgsManager' dependency":
(https://github.com/bitcoin/bitcoin/pull/26889#issuecomment-1429700293)
Feedback tackled.
Moved remaining variables to signed type. Small [diff](https://github.com/bitcoin/bitcoin/compare/fafaa7a7e15a49da0214ce9546a42117c63e611f..d90b54b4aa18d60faa0075348fbb29a513558098).

💬 vasild commented on pull request "build: Add CMake-based build system (1 of N)":
(https://github.com/bitcoin/bitcoin/pull/27060#discussion_r1105847171)
Do autotools really pick `/foo/bin/clang` (assuming `/foo/bin` is in `$PATH`)? I tried that but for me it picked `/usr/bin/c++` nevertheless, even though I had a `clang` executable in a directory earlier than `/usr/bin` in `$PATH`.

💬 vasild commented on pull request "build: Add CMake-based build system (1 of N)":
(https://github.com/bitcoin/bitcoin/pull/27060#discussion_r1105862197)
nit:

```suggestion
message(WARNING "${warning}")
```

https://cmake.org/cmake/help/latest/command/message.html

👍 vasild approved a pull request: "build: Add CMake-based build system (1 of N)"
(https://github.com/bitcoin/bitcoin/pull/27060)

💬 vasild commented on pull request "Handle CJDNS from LookupSubNet()":
(https://github.com/bitcoin/bitcoin/pull/27071#discussion_r1105886420)
Done.

💬 willcl-ark commented on issue "libevent: Error from accept() call: Invalid argument on Android termux (arm)":
(https://github.com/bitcoin/bitcoin/issues/19432#issuecomment-1429822788)
Since this was reported:

* There have been no additional reports of occurance
* Lineage OS has been updated from v16.0 to v20.0
* Bitcoin Core has been updated from v0.20 to v24.0.1
* libevent in Bitcoin Core depends was [updated](https://github.com/bitcoin/bitcoin/pull/21991) to 2.1.12-stable (which their script [appears to use](https://github.com/termux/termux-packages/blob/master/packages/bitcoin/build.sh#L25))

(and is anyone still running a full node on Android phone)

@MarcoFalke
...

💬 vasild commented on pull request "Handle CJDNS from LookupSubNet()":
(https://github.com/bitcoin/bitcoin/pull/27071#discussion_r1105887566)
There is no reason for 0x55 specifically. I added a comment to clarify:

`// Just an arbitrary number, anything != 0xFC would do.`

💬 john-moffett commented on pull request "validation: Improve error handling when VerifyDB fails due to insufficient dbcache":
(https://github.com/bitcoin/bitcoin/pull/25574#issuecomment-1429827332)
ACK https://github.com/bitcoin/bitcoin/commit/080f83e43ddc1fd2ddc57f6590b9166a373d38c2 apart from the incidental test failure, which you've probably already caught:

```diff
diff --git a/test/functional/feature_pruning.py b/test/functional/feature_pruning.py
index 664ed779db2..03f5f2c8037 100755
--- a/test/functional/feature_pruning.py
+++ b/test/functional/feature_pruning.py
@@ -226 +226 @@ class PruneTest(BitcoinTestFramework):
- with self.nodes[2].assert_debug_log(expected_msgs
...

💬 MarcoFalke commented on issue "libevent: Error from accept() call: Invalid argument on Android termux (arm)":
(https://github.com/bitcoin/bitcoin/issues/19432#issuecomment-1429830980)
Is this still an issue with a recent version of Bitcoin Core? If yes, what are the steps to reproduce?