< bitcoin-git>
bitcoin/master 9fc68fa Antoine Poinsot: script: match multisigs with up to MAX_PUBKEYS_PER_MULTISIG keys
< bitcoin-git>
bitcoin/master ae0429d Antoine Poinsot: script: allow up to 20 keys in wsh() descriptors
< bitcoin-git>
bitcoin/master 063df9e Antoine Poinsot: test/functional: standardness sanity checks for P2(W)SH multisig
< bitcoin-git>
[bitcoin] fanquake merged pull request #20867: Support up to 20 keys for multisig under Segwit context (master...descriptor_multi_wsh) https://github.com/bitcoin/bitcoin/pull/20867
< fanquake>
catgirladdict: can be blocked. Second time they've opened that PR.
< wumpus>
will do
< rebroad>
MarcoFalke i only just realised that BIP111 didn't get updated with the changes I proposed back in 2016... so yeah, nevermind... i guess i wasn't keeping my eye on the ball back in 2016!
< rebroad>
I ought to understand githb better. I'm never quite sure how to get notifications for when a pull is requiring my attention
< darosior>
Wallet question: Why does HasWalletSpend() only checks for the first output in mapTxSpends ?
< jonatack>
#21644 has been through several rounds of feedback and has acks by vasild, theStack and hebasto. It might be RFM and possible backport to 0.21.2.
< bitcoin-git>
[bitcoin] MarcoFalke opened pull request #21840: test: Misc refactor to get rid of &foo[0] raw byte pointers (master...2105-testRefactor) https://github.com/bitcoin/bitcoin/pull/21840
< bitcoin-git>
[gui] hebasto opened pull request #313: qt, build: Optimize string concatenation by default (master...210503-builder) https://github.com/bitcoin-core/gui/pull/313
< bitcoin-git>
[bitcoin] rebroad opened pull request #21841: Send fewer feefilter messages (avoid the wobbling number issue) (master...SteadierFeefilter) https://github.com/bitcoin/bitcoin/pull/21841
< gmaxwell>
rebroad: that 21841 PR will essentially do nothing. The underlying issue is that the 'rounding' uses a constantly changing random number, instead of like.. one per network identity. Which actually undermines the intended randomness.
< bitcoin-git>
[bitcoin] jonatack opened pull request #21843: net, rpc: enable GetAddr, GetAddresses, and getnodeaddresses by network (master...getnodeaddresses-by-network) https://github.com/bitcoin/bitcoin/pull/21843
< rebroad>
gmaxwell are you saying that the "rounding" undermines randonmess? I've updated my pull request to explain that it will make essentially no difference unless the algo for the minrelayfee is changed to make the numebr more static
< rebroad>
but I think it's good to account for possible changes elsewhere in the code and to not assume things will stay the same - these kind of assumptions are what caused the inflation bug, for example
< rebroad>
is there any particular reason a node won't use the first cmpctblock it received, but instead waits for the one it requested (from the node that first sent it the header)?
< rebroad>
I've just started using the netinfo RPC - very nice... is it documented anywhere? the rpc help doesn't go into detail about the different options for it
< rebroad>
e.g. the txn column is showing mostly zeros... what is this column?
< jonatack>
rebroad: bitcoin-cli -netinfo help
< jonatack>
(thanks!)
< rebroad>
ah..! thanks!! jonatack
< rebroad>
jonatack ah.. no, that doesn't work
< jonatack>
bitcoin-cli -netinfo help
< rebroad>
no, it doesn't work
< rebroad>
it just gives me the output from a regular netinfo command
< jonatack>
it was added after 0.21
< rebroad>
ah.. I'm running 0.21.1
< rebroad>
so i need to use master?
< jonatack>
ah, guess it wasn't backported. yes. oopsie
< rebroad>
oh. weird. on master netinfo 4 doesn't work but it does on 0.21.1
< jonatack>
or just grep bitcoin-cli.cpp on master for m_help_doc
< jonatack>
it should indeed work on master, it's continually in use for me on latest master and while testing pulls
< rebroad>
error: JSON value is not a boolean as expected
< jonatack>
are both client and server running master?
< rebroad>
ah. no client is master, server is 0.21.1
< rebroad>
oopsie
< jonatack>
thanks, that's an edge case i should check out.
< rebroad>
I would switch to master, but jonasschnelli hasn't rebased his adorable mempool fee graphs for master yet :(
< rebroad>
I tried rebasing them but clearly my C++ skills aren't up to scratch
< jonatack>
there's an error message if the server is pre-0.21 but i didn't test this, will have a look
< * jonatack>
lunch
< rebroad>
pull #9424 says that I can debug to multiple categories, but I don't see this done anywhere, nor documented how... could someone help me with the sytax please to log to two BCLog categories?
< gribble>
https://github.com/bitcoin/bitcoin/issues/9424 | Change LogAcceptCategory to use uint32_t rather than sets of strings. by gmaxwell · Pull Request #9424 · bitcoin/bitcoin · GitHub
< michaelfolkson>
rebroad: "The -debug=... command-line option controls debugging; running with just -debug or -debug=1 will turn on all categories (and give you a very large debug.log file)."
< fanquake>
If anyone with Apple M1 hardware is feeling adventurous; I've managed to get some binaries cross-compiled (minus Qt) using depends, however have no hardware to test them.
< fanquake>
I'll push a branch up soon, however at this stage build requires some manual steps, and Qt needs to be fixed. I assume there'll be some patches to backport.
< provoostenator>
I think the v0.21.1 dmg still needs Gatekeeper blessing
< rebroad>
ok, I notice #21841 is failing a test, so clearly my "make check" wasn't enough... what command do I run please to do all the tests?
< rebroad>
and why am I struggling to know where to look to find this kind of info...? Is there a developer FAQ somewhere? ideally one that can also be edited via github
< rebroad>
i mean, can we perhaps create a DEVELOPER.md in the docs directly as a place for people to find such info?
< rebroad>
ah.. there's a developer-nodes.md .... ooopse
< sipa>
rebroad: make check only runs the unit tests, it doesn't run various other things (in particular, functional tests in test/functional, and fuzz tests which require separate compilation options)
< sipa>
also things can fail in CI because it runs tests on many more varied architectures
< jonatack>
(rebroad, if useful here are some articles on these things; i try to keep them up to date, feel free to ping by direct irc with suggestions https://jonatack.github.io/articles)
< vincenzopalazzo>
hello guys, I'm looking the code of rpc interface and in particular of the command "gettxout" because I was not able to understend why the answer from rpc was empty. I noted that when the transaction is not in the mempool the answer is NONE
< vincenzopalazzo>
I want ask if there is any motivation that I'm missing at the moment to don't receive any "error" from the call in this case, like "utxo spent" and/or "utxo not found"
< sipa>
it's not an error because you didn't do anything wrong
< sipa>
the answer is just no utxo
< vincenzopalazzo>
sipa, yeah I assume that this this is one of the motivation
< gmaxwell>
Maybe make check should run the real tests. Make check has long been very limited compared to the real stuff.
< vincenzopalazzo>
I just asking because in my vision this case can be generalized like the case when the user will ask the block that is not synced in the node but is ready on the network, this is another case where the user doesn't make nothing wrong. I'm missing somethings?
< sipa>
vincenzopalazzo: the interface isn't always consistent
< vincenzopalazzo>
sipa: Do you things that a PR can be welcome to discuss this "problem" that is not really a problem?
< bitcoin-git>
[bitcoin] rebroad closed pull request #21841: Send fewer feefilter messages (avoid the wobbling number issue) (master...SteadierFeefilter) https://github.com/bitcoin/bitcoin/pull/21841
< vincenzopalazzo>
I want avoid stupid PR :)
< sipa>
vincenzopalazzo: the problem is that fixing it requires breaking compatibility
< vincenzopalazzo>
sipa: This is a big problem, also because there all the ln implementation that use this command
< sipa>
which we do sometimes, with proper deprecation and the ability to stick with existing behaviour etc... but i'm not sure it's worth it
< sipa>
vincenzopalazzo: i don't think there is a problem at all, except the fact that it"s confusing
< vincenzopalazzo>
sipa: Agree, no problems here. I just asking because I would happy to have more info on the command line when this error happen, if we have a message somethings the user (like me) can found a problem.
< vincenzopalazzo>
Before opening a PR to I prefererd to ask here. Any way I agree that this change bring more problem that solution
< bitcoin-git>
[bitcoin] rebroad reopened pull request #21841: Send fewer feefilter messages (avoid the wobbling number issue) (master...SteadierFeefilter) https://github.com/bitcoin/bitcoin/pull/21841