💬 hebasto commented on pull request "wallet: Replace "non-0" with "non-zero" in translatable error message":
(https://github.com/bitcoin/bitcoin/pull/31987#issuecomment-2700679622)
> > Automatically generated by extract_strings_qt
>
> Yes, but it's committed, so shouldn't we update or regenerate it now?

Sure. It will regenerated during the [release process](https://github.com/bitcoin/bitcoin/blob/master/doc/release-process.md#before-branch-off), but not for 29.0 as we are behind the string freeze.

💬 fanquake commented on issue "Fuzz: Runtime errors when running fuzz tests on MacOs":
(https://github.com/bitcoin/bitcoin/issues/31591#issuecomment-2700685226)
@Prabhat1308 is your issue solved after using the new steps in #31954?

💬 fanquake commented on issue "`DEFAULT_TRANSACTION_MAXFEE` is 0.1 ₿":
(https://github.com/bitcoin/bitcoin/issues/31716#issuecomment-2700692797)
@achow101 @furszy @theStack any thoughts here?

💬 l0rinc commented on pull request "[WIP] refactor: migrate unit tests to Google Test":
(https://github.com/bitcoin/bitcoin/pull/31988#issuecomment-2700694714)
Cool proof of concept, eventually we might want to give another testing framework a try, but not sure we're there yet.
The limitations of Boost bothers me a bit every time I'm using it, but not enough to embark on a huge refactor yet.
Maybe it would still be easier to strangle out the old one by having two options temporarily - or to have an api with minimal changes compared to current one, supporting both.
Not sure, but my impression was that the pain isn't great enough yet.
But if you're p
...

💬 fanquake commented on issue "test: different error message fails rpc_signer.py":
(https://github.com/bitcoin/bitcoin/issues/31506#issuecomment-2700697810)
> cc @hebasto

@hebasto or maybe @achow101 can you take a look here? Seems like the bisect is pointing to breakage when boost::process was swapped out.

💬 hebasto commented on pull request "Drop `CCheckQueue::nIdle`":
(https://github.com/bitcoin/bitcoin/pull/26776#issuecomment-2700703633)
Benchmarked in https://github.com/bitcoin-dev-tools/benchcoin/pull/140.

💬 hebasto commented on issue "test: different error message fails rpc_signer.py":
(https://github.com/bitcoin/bitcoin/issues/31506#issuecomment-2700717633)
> > cc [@hebasto](https://github.com/hebasto)
>
> [@hebasto](https://github.com/hebasto) or maybe [@achow101](https://github.com/achow101) can you take a look here? Seems like the bisect is pointing to breakage when boost::process was swapped out.

I'm working on `src/util/subprocess.h` code.

🤔 marcofleon reviewed a pull request: "kernel: pre-29.x chainparams and headerssync update"
(https://github.com/bitcoin/bitcoin/pull/31978#pullrequestreview-2660913860)
Verified all the values for mainnet after syncing with `-assumevalid=0`. Also ran the headers sync script and got the same results.
```
du -csh ./blocks ./chainstate
681G ./blocks
11G ./chainstate
692G total
```

💬 stickies-v commented on pull request "validation: stricter internal handling of invalid blocks":
(https://github.com/bitcoin/bitcoin/pull/31405#discussion_r1981280478)
Oh sorry, I get it now, thanks. For some reason (I think I had just been reviewing the next section dealing with disconnecting the tip) I was confusing `pindex` for the chain tip here. I don't really have any doc suggestions, so can be marked as resolved.

💬 hebasto commented on pull request "[WIP] refactor: migrate unit tests to Google Test":
(https://github.com/bitcoin/bitcoin/pull/31988#issuecomment-2700746683)
> #### Motivation
>
> By migrating the unit tests from Boost to Google Test we can increase the chances of removing all dependencies from Boost in the future.

Depending on Boost.Test for test-only code does not seem like a problem to me.

💬 hebasto commented on issue "[rfc] build: Reject unclean configure?":
(https://github.com/bitcoin/bitcoin/issues/31942#issuecomment-2700755534)
> [@purpleKarrot](https://github.com/purpleKarrot) maybe you could provide guidance on how to properly solve issues like [#31959](https://github.com/bitcoin/bitcoin/issues/31959) or one of the others the project is experiencing, which lead to this RFC?

It would also be helpful if someone provided steps to reproduce those issues.

💬 laanwj commented on pull request "doc: Bring reduce-memory.md up to date":
(https://github.com/bitcoin/bitcoin/pull/31985#issuecomment-2700766903)
> Should we make this function like maxmempool, which will fail to accept a value lower than the minimim (5)? Seems we are somewhat inconsistent in our handling of these.

Yes, probably. It's pretty much always better for user communication to fail for wrong values than to silently interpret them as something else.

💬 Sjors commented on pull request "Add mainnet assumeutxo param at height 880,000":
(https://github.com/bitcoin/bitcoin/pull/31969#issuecomment-2700775387)
@jonatack wrote:

> must appear in the headers chain.

That's expected behavior. The snapshot will refuse to load until we have its header.

Though one thing that could be improved it to have it wait for headers to reach the given height instead of fail.

💬 Sjors commented on pull request "Add mainnet assumeutxo param at height 880,000":
(https://github.com/bitcoin/bitcoin/pull/31969#discussion_r1981300210)
We want this to be updated in master, so we have to pick something before the v29 branch-off.

💬 fanquake commented on issue "cmake inconsistently overriding `-O3` (sometimes)":
(https://github.com/bitcoin/bitcoin/issues/31491#issuecomment-2700782712)
@sipa I think you were going to follow up here?

💬 vasild commented on pull request "Split CConnman":
(https://github.com/bitcoin/bitcoin/pull/30988#discussion_r1981304441)
Right, only to the caller (same thread). Locked or not by other threads is irrelevant for `AssertLockNotHeld()`.

💬 fanquake commented on issue "Having qt(@6) breaks build for qt@5 on macOS 15.0 and 13.7":
(https://github.com/bitcoin/bitcoin/issues/31009#issuecomment-2700805847)
Looks like this is still an issue. So this should either be documented for macOS builders (that they'll need to uninstall `qt` if it happens to be installed, whch is decently likely (and increasingly so into the future)), or the underlying issue fixed?

💬 Sjors commented on pull request "guix: Notarize MacOS app bundle and codesign all MacOS and Windows binaries":
(https://github.com/bitcoin/bitcoin/pull/31407#issuecomment-2700813691)
@fanquake we don't publish them on the download site, but we do put them on `bitcoincore.org/bin`. It seems harmless, but not useful.

The unsigned downloads still work if users (or some docker automation) self-sign, though that's pointless now that we offer a correctly signed alternative.

Even if Apple ever revokes our certificate, afaik it's possible for users to self-sign and override our signature.[^1]

If the user does a guix build themselves and copies the unsigned binaries using (s
...

💬 Sjors commented on issue "Having qt(@6) breaks build for qt@5 on macOS 15.0 and 13.7":
(https://github.com/bitcoin/bitcoin/issues/31009#issuecomment-2700820604)
I ran into it[^1] again a week or so ago, so I don't think it's fixed.

[^1] though I didn't compare the exact error, building worked again once I uninstalled qt6

👍 willcl-ark approved a pull request: "[28.x] Backports"
(https://github.com/bitcoin/bitcoin/pull/31648#pullrequestreview-2660991916)
ACK 77c13c7d3e68464555c6e9a049782674bcd9ea67

All commits look good to me.

I guess the commit title "[WIP] Update release notes" gets updated later during finalisation?

💬 hodlinator commented on pull request "build, ci: Fix linking `bitcoin-chainstate.exe` to `bitcoinkernel.dll` on Windows":
(https://github.com/bitcoin/bitcoin/pull/31158#discussion_r1981191286)
nit: Less confusing for every branch in the block to result in `#define`:
```suggestion
#else
```