🤔 pablomartin4btc reviewed a pull request: "restrict std::cerr to errors; use std::cout for warnings and info"
(https://github.com/bitcoin/bitcoin/pull/32538#pullrequestreview-2847419694)
Concept ACK

I think some tests using `stop_node` with the warning in the args would fail as in `is_node_stopped` checks for the warning in `stderr` (eg `wallet_multiwallet.py`), it needs to be fixed too.

💬 pablomartin4btc commented on pull request "restrict std::cerr to errors; use std::cout for warnings and info":
(https://github.com/bitcoin/bitcoin/pull/32538#issuecomment-2887476146)
> This change means that warnings may be missed in tests/CI or elsewhere (see the failing CI).

That's my concern with this approach.

💬 purpleKarrot commented on issue "cmake: Cannot find Qt 6 on SunOS / illumos (OpenIndiana Distribution)":
(https://github.com/bitcoin/bitcoin/issues/32536#issuecomment-2887477698)
It sure is annoying when system packages are installed to a location where they cannot be found by default. On the other hand, package maintainers may have their reasons for doing that. There are probably multiple packages which depend on Qt6 and are built with CMake. It is very likely that package maintainers have their build environment set up for that.

Trying to work around that "issue" creates a maintenance burden for us and potentially for package maintainers too, because our workaround ma
...

🤔 l0rinc reviewed a pull request: "node: cap `-maxmempool` and `-dbcache` values for 32-bit"
(https://github.com/bitcoin/bitcoin/pull/32530#pullrequestreview-2847428745)
Looking at the code, I see that I misunderstood the description, I though we're capping the values in every circumstance, but now I see that the architectures are handles independently.

💬 achow101 commented on pull request "descriptors: MuSig2":
(https://github.com/bitcoin/bitcoin/pull/31244#discussion_r2093554238)
Done, added to musig.h

💬 achow101 commented on pull request "descriptors: MuSig2":
(https://github.com/bitcoin/bitcoin/pull/31244#discussion_r2093554514)
Updated the comments.

💬 achow101 commented on pull request "descriptors: MuSig2":
(https://github.com/bitcoin/bitcoin/pull/31244#discussion_r2093554645)
Done

💬 achow101 commented on pull request "descriptors: MuSig2":
(https://github.com/bitcoin/bitcoin/pull/31244#discussion_r2093554764)
Done

💬 achow101 commented on pull request "descriptors: MuSig2":
(https://github.com/bitcoin/bitcoin/pull/31244#discussion_r2093554945)
"is"

💬 achow101 commented on pull request "descriptors: MuSig2":
(https://github.com/bitcoin/bitcoin/pull/31244#discussion_r2093555270)
Added comments.

💬 achow101 commented on pull request "descriptors: MuSig2":
(https://github.com/bitcoin/bitcoin/pull/31244#discussion_r2093555374)
Done

💬 achow101 commented on pull request "descriptors: MuSig2":
(https://github.com/bitcoin/bitcoin/pull/31244#discussion_r2093555457)
any that we have.

💬 maflcko commented on pull request "node: cap `-maxmempool` and `-dbcache` values for 32-bit":
(https://github.com/bitcoin/bitcoin/pull/32530#issuecomment-2887503071)
No objection, but there is no hand-holding on 64-bit systems regarding OOM footguns, so doing it on 32-bit systems (the edge case ) seems fine, but a bit unexpected, given that OOM reports are generally on 64-bit systems? Again, no objection, just wondering if there is more background to it.

✅ furszy closed a pull request: "restrict std::cerr to errors; use std::cout for warnings and info"
(https://github.com/bitcoin/bitcoin/pull/32538)

💬 furszy commented on pull request "restrict std::cerr to errors; use std::cout for warnings and info":
(https://github.com/bitcoin/bitcoin/pull/32538#issuecomment-2887513237)
> When something literally starts with Warning: ..., I fail to see how it can be misinterpreted.

It is actually being misinterpreted by our own test framework. The framework does not look
at the stderr content, it only checks whether there is something inside stderr or not during
shutdown, failing when finds something there.

> This change means that warnings may be missed in tests/CI or elsewhere (see the failing CI).

That's ok. That's what the PR is proposing. To redirect them to std
...

💬 furszy commented on pull request "wallet: init, don't error out when loading legacy wallets":
(https://github.com/bitcoin/bitcoin/pull/32449#discussion_r2093584558)
> why the `.rpc.`?

because my subconscious still thinks about the legacy wallet proxy methods and we don't have them anymore.

💬 furszy commented on pull request "wallet: init, don't error out when loading legacy wallets":
(https://github.com/bitcoin/bitcoin/pull/32449#discussion_r2093585184)
removed.

📝 pinheadmz opened a pull request: "init: Configure reachable networks before we start the RPC server"
(https://github.com/bitcoin/bitcoin/pull/32539)
Closes https://github.com/bitcoin/bitcoin/issues/32433

`MaybeFlipIPv6toCJDNS()` relies on `g_reachable_nets` to distinguish between CJDNS addresses and other IPv6 addresses. In particular, [RFC4193](https://www.rfc-editor.org/rfc/rfc4193#section-3.1) address or "Unique Local Address" with the L-bit unset also begins with the `fc` prefix. #32433 highlights a use case for these addresses that have nothing to do with CJDNS.

On master we don't parse init flags like `-cjdnsreachable` until *aft
...

💬 pinheadmz commented on issue ""rpcallowip=" configuration directive doesn't accept RFC4193 addresses":
(https://github.com/bitcoin/bitcoin/issues/32433#issuecomment-2887572908)
https://github.com/bitcoin/bitcoin/pull/32539

💬 BitcoinMechanic commented on pull request "policy: uncap datacarrier by default":
(https://github.com/bitcoin/bitcoin/pull/32406#issuecomment-2887596317)
It seems if you want people to actually use this much OP RETURN data by relaxing the filter it'd make sense to simultaneously prevent "inscriptions" with this https://github.com/bitcoin/bitcoin/pull/28408