💬 jamesob commented on pull request "policy: uncap datacarrier by default":
(https://github.com/bitcoin/bitcoin/pull/32406#discussion_r2076064182)
In a vacuum it might be nice as a `std::optional<unsigned int>`, but that probably causes too much downstream churn.

💬 maflcko commented on pull request "policy: uncap datacarrier by default":
(https://github.com/bitcoin/bitcoin/pull/32406#discussion_r2076086158)
However, if some people prefer the option to not be deprecated (for whatever reason), it seems pretty minimal maintenance to keep it around without deprecation for now. There is no rush for this pull request, as the 30.x release is months away and also no rush to remove or deprecate the setting. Even if it is only for the extremely unlikely case to avoid having to undeprecate it later on (see also https://github.com/bitcoin/bitcoin/pull/32359#issuecomment-2855558032). The option has existed for
...

💬 iFadi commented on pull request "Remove arbitrary limits on OP_Return (datacarrier) outputs":
(https://github.com/bitcoin/bitcoin/pull/32359#issuecomment-2855633720)
Concept NACK

💬 ryanofsky commented on issue "Depends toolchain doesn't contain enough info to build from depends on a fresh NixOS install":
(https://github.com/bitcoin/bitcoin/issues/32428#issuecomment-2855636472)
> The thing I don't understand is why setting CMAKE_PREFIX_PATH as well as CMAKE_FIND_ROOT_PATH seems to be necessary on nix while setting CMAKE_FIND_ROOT_PATH alone seems to work on other platforms.

Reason for this is that nix is patching cmake to remove the root path `/` from `CMAKE_SYSTEM_PREFIX_PATH` in https://github.com/NixOS/nixpkgs/blob/df545660856d8252b4856bede70d424ef4b47c64/pkgs/by-name/cm/cmake/001-search-path.diff

Because of this `CMAKE_FIND_ROOT_PATH` is not searched directly for
...

💬 willcl-ark commented on issue "Depends toolchain doesn't contain enough info to build from depends on a fresh NixOS install":
(https://github.com/bitcoin/bitcoin/issues/32428#issuecomment-2855642758)
Yes I agree @ryanofsky, thanks for double-checking!

I added the `if()` guard because I heard from @hebasto that we call the toolchain multiple times, and the `if` guard therefore avoids adding the path repeatedly.

📝 w0xlt opened a pull request: "docs: Improve `keypoolrefill` RPC docs"
(https://github.com/bitcoin/bitcoin/pull/32429)
Update `keypoolrefill` RPC docs to make it clear that descriptor wallets have 4 ScriptPubKeyManagers by default and each of them is updated in this command, as suggested https://github.com/bitcoin/bitcoin/issues/29924#issuecomment-2849321859

Closes https://github.com/bitcoin/bitcoin/issues/29924

💬 w0xlt commented on issue "`keypoolrefill` doesn't fill keypool to specified parameter":
(https://github.com/bitcoin/bitcoin/issues/29924#issuecomment-2855652872)
> it could make sense to clarify the RPC docs

Done in https://github.com/bitcoin/bitcoin/pull/32429

💬 ryanofsky commented on issue "Depends toolchain doesn't contain enough info to build from depends on a fresh NixOS install":
(https://github.com/bitcoin/bitcoin/issues/32428#issuecomment-2855669606)
> I added the `if()` guard because I heard from [@hebasto](https://github.com/hebasto) that we call the toolchain multiple times, and the `if` guard therefore avoids adding the path repeatedly.

I think this is half-true. I noticed by adding prints to the cmake file that it was called multiple times, but each time CMAKE_PREFIX_PATH was always set back to the original value, so I don't think the `if` guard did anything.

Also just to be sure another difference isn't glossed over: I think it proba
...

💬 achow101 commented on pull request "Remove the legacy wallet and BDB dependency":
(https://github.com/bitcoin/bitcoin/pull/28710#discussion_r2076116744)
Done both.

💬 achow101 commented on pull request "Remove the legacy wallet and BDB dependency":
(https://github.com/bitcoin/bitcoin/pull/28710#discussion_r2076119234)
Removed:
* `importmulti' in doc/descriptors.md
* bdb.cpp in linter
* `WriteHDCHain` and WriteCScript`

I would prefer to leave further cleanups for follow up RPs as I am sure there are plenty of things that were missed and trying to catch them all in this one PR is going to be very annoying and just delay this even more.

💬 achow101 commented on pull request "Remove the legacy wallet and BDB dependency":
(https://github.com/bitcoin/bitcoin/pull/28710#discussion_r2076119393)
Done

💬 willcl-ark commented on issue "Depends toolchain doesn't contain enough info to build from depends on a fresh NixOS install":
(https://github.com/bitcoin/bitcoin/issues/32428#issuecomment-2855759003)
> Also just to be sure another difference isn't glossed over: I think it probably is better to set CMAKE_SYSTEM_PREFIX_PATH instead of CMAKE_PREFIX_PATH since former seems to fix the original problem, while latter is more of a workaround.

That sounds reasonable.

> Also really appreciate you reporting this issue and finding a good workaround. This has been causing me headaches for a long time, and I've only had crappy workarounds and couldn't figure out how this was working at all on other syst
...

💬 maflcko commented on issue "Intermittent timeout in tsan feature_init.py":
(https://github.com/bitcoin/bitcoin/issues/30586#issuecomment-2855818587)
A workaround seems to be:

```diff
diff --git a/test/functional/feature_init.py b/test/functional/feature_init.py
index 15b3e85..1fef8e1 100755
--- a/test/functional/feature_init.py
+++ b/test/functional/feature_init.py
@@ -44,7 +44,7 @@ class InitTest(BitcoinTestFramework):
# bitcoind to perform any shutdown logic.
os.kill(node.process.pid, signal.CTRL_BREAK_EVENT)
else:
- node.process.terminate()
+ node.process.kill()
...

💬 ybaidiuk commented on pull request "Remove arbitrary limits on OP_Return (datacarrier) outputs":
(https://github.com/bitcoin/bitcoin/pull/32359#issuecomment-2855844462)
Why in 5 years btc tx should concur for space inside block with another data? blockchain is about bitcoin not other data, if you people want to store data it should be expensive and hard, or use side chains like liquid etc.

My 7 years old node go down today, and i'm switching to knots.

💬 janb84 commented on pull request "docs: clarify RPC credentials security boundary":
(https://github.com/bitcoin/bitcoin/pull/32424#issuecomment-2855887125)
Please [squash](https://github.com/bitcoin/bitcoin/blob/master/CONTRIBUTING.md#squashing-commits) your commits

💬 maflcko commented on pull request "docs: Improve `keypoolrefill` RPC docs":
(https://github.com/bitcoin/bitcoin/pull/32429#discussion_r2076251627)
Should probably use `DEFAULT_KEYPOOL_SIZE` instead of hardcoded 1000? Same for the `OUTPUT_TYPES`? Also, `ScriptPubKeyManagers` was never used in user-facing docs, except for one error message. Maybe just use "active ranged descriptor" (or similar)?

💬 maflcko commented on pull request "docs: Improve `keypoolrefill` RPC docs":
(https://github.com/bitcoin/bitcoin/pull/32429#discussion_r2076282332)
also, now that bdb was removed, no need to add docs for it?

📝 hebasto opened a pull request: "qt, docs: Unify term "clipboard""
(https://github.com/bitcoin-core/gui/pull/871)
A translator on Transifex noticed:
> The term "system clipboard" appears twice. The term "clipboard" appears 10 times. Perhaps we could standardize on just saying "clipboard"?

This PR addresses this issue.

💬 hebasto commented on pull request "qt, docs: Unify term "clipboard"":
(https://github.com/bitcoin-core/gui/pull/871#issuecomment-2855996352)
> A translator on Transifex noticed:
>
> > The term "system clipboard" appears twice. The term "clipboard" appears 10 times. Perhaps we could standardize on just saying "clipboard"?

Reported by @jesterhodl.

💬 crStiv commented on pull request "docs: clarify RPC credentials security boundary":
(https://github.com/bitcoin/bitcoin/pull/32424#issuecomment-2856003596)
> Please [squash](https://github.com/bitcoin/bitcoin/blob/master/CONTRIBUTING.md#squashing-commits) your commits

@janb84 yeah sure, np