michagogo has quit [Quit: Connection closed for inactivity]
AaronvanW has quit [Remote host closed the connection]
AaronvanW has joined #bitcoin-core-dev
yanmaani has quit [Remote host closed the connection]
yanmaani has joined #bitcoin-core-dev
AaronvanW has quit [Ping timeout: 240 seconds]
sipsorcery has quit [Ping timeout: 252 seconds]
belcher has quit [Ping timeout: 252 seconds]
AaronvanW has joined #bitcoin-core-dev
belcher has joined #bitcoin-core-dev
AaronvanW has quit [Ping timeout: 248 seconds]
tripleslash has joined #bitcoin-core-dev
AaronvanW has joined #bitcoin-core-dev
Yihen has joined #bitcoin-core-dev
Yihen has left #bitcoin-core-dev [#bitcoin-core-dev]
Yihen has joined #bitcoin-core-dev
bitcoin-git has joined #bitcoin-core-dev
<bitcoin-git>
[bitcoin] ben-kaufman opened pull request #22775: rpc: Add option to list transactions from oldest to newest in `listtransactions` RPC command (master...listtransactions-set-order) https://github.com/bitcoin/bitcoin/pull/22775
bitcoin-git has left #bitcoin-core-dev [#bitcoin-core-dev]
bitcoin-git has left #bitcoin-core-dev [#bitcoin-core-dev]
bitcoin-git has joined #bitcoin-core-dev
<bitcoin-git>
[bitcoin] fanquake merged pull request #22275: A few follow-ups for taproot signing (master...202106_taproot_sign_followup) https://github.com/bitcoin/bitcoin/pull/22275
bitcoin-git has left #bitcoin-core-dev [#bitcoin-core-dev]
bitcoin-git has left #bitcoin-core-dev [#bitcoin-core-dev]
babasancheti has joined #bitcoin-core-dev
SpellChecker has quit [Remote host closed the connection]
SpellChecker has joined #bitcoin-core-dev
lkqwejhhgasdjhgn has joined #bitcoin-core-dev
AaronvanW has joined #bitcoin-core-dev
Henrik has joined #bitcoin-core-dev
AaronvanW has quit [Ping timeout: 250 seconds]
johnzweng has quit [Quit: Leaving...]
johnzweng has joined #bitcoin-core-dev
AaronvanW has joined #bitcoin-core-dev
bitcoin-git has joined #bitcoin-core-dev
<bitcoin-git>
[bitcoin] jnewbery opened pull request #22777: net processing: don't request tx relay on feeler connections (master...2021-08-feeler-no-frelay) https://github.com/bitcoin/bitcoin/pull/22777
bitcoin-git has left #bitcoin-core-dev [#bitcoin-core-dev]
bitcoin-git has joined #bitcoin-core-dev
<bitcoin-git>
[bitcoin] glozow closed pull request #22252: policy: Trim Packages when transaction with same txid exists in mempool (master...2021-06-mempool-matches) https://github.com/bitcoin/bitcoin/pull/22252
bitcoin-git has left #bitcoin-core-dev [#bitcoin-core-dev]
bitcoin-git has left #bitcoin-core-dev [#bitcoin-core-dev]
bitcoin-git has joined #bitcoin-core-dev
<bitcoin-git>
[bitcoin] MarcoFalke merged pull request #22622: util: Check if specified config file cannot be opened (master...check_conf_file) https://github.com/bitcoin/bitcoin/pull/22622
bitcoin-git has left #bitcoin-core-dev [#bitcoin-core-dev]
AaronvanW has joined #bitcoin-core-dev
jonatack has quit [Ping timeout: 246 seconds]
Guyver2 has joined #bitcoin-core-dev
babasancheti has quit [Quit: Client closed]
b10c has joined #bitcoin-core-dev
<laanwj>
robertspigler: good to see at least they're honest about it--a web ui stack imports so many dependencies, through different package managers, etc, it's very hard to minimize the risk of a third-party compromise affecting the software supply chain
sipsorcery has joined #bitcoin-core-dev
AaronvanW has quit [Ping timeout: 250 seconds]
vysn has quit [Remote host closed the connection]
<laanwj>
qt's dependendency management is delightfully dumb and simple in comparison
<fanquake>
you're only one npm install away from downloading a malicious left-pad
Yihen has quit [Remote host closed the connection]
Guyver2 has quit [Quit: Going offline, see ya! (www.adiirc.com)]
bitcoin-git has joined #bitcoin-core-dev
<bitcoin-git>
[bitcoin] jnewbery opened pull request #22778: net processing: Reduce resource usage for inbound block-relay-only connections (master...2021-02-tx-relay-init) https://github.com/bitcoin/bitcoin/pull/22778
bitcoin-git has left #bitcoin-core-dev [#bitcoin-core-dev]
AaronvanW has joined #bitcoin-core-dev
muhblockchain has quit [Ping timeout: 248 seconds]
bitdex has quit [Ping timeout: 244 seconds]
Henrik has quit [Quit: My MacBook Air has gone to sleep. ZZZzzz…]
Henrik has joined #bitcoin-core-dev
AaronvanW has quit [Ping timeout: 252 seconds]
<michaelfolkson>
Has anyone in the past tried to add encryption to the RPC interface? Or is there a particular reason why that shouldn't be done?
<michaelfolkson>
Couldn't find anything from a cursory PR search
grettke has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
<sipa>
there used to be!
<sipa>
it was deleted because it's pointless complexity
<sipa>
it's intended for communication with trusted, local software
bitcoin-git has left #bitcoin-core-dev [#bitcoin-core-dev]
AaronvanW has joined #bitcoin-core-dev
<michaelfolkson>
sipa: Thanks
Randolf has joined #bitcoin-core-dev
videre_e has joined #bitcoin-core-dev
<videre_e>
is it possible to do mining bitcoins on a nvidia card?
<Randolf>
videre_e: This is the development channel. For mining questions, please ask in the #bitcoin-mining channel instead: /join #bitcoin-mining
Guyver2 has joined #bitcoin-core-dev
Henrik has quit [Quit: My MacBook Air has gone to sleep. ZZZzzz…]
jonatack has joined #bitcoin-core-dev
vnogueira has joined #bitcoin-core-dev
<laanwj>
bitcoind's control socket is supposed to be local-only, i wish we'd used a UNIX socket from the beginning (as clightning does)
<laanwj>
and encryption is pointless when communicating to a local service, the OS can intercept everything anyhow
<laanwj>
fanquake: unfortunately, rust has the same problem, i guess it's possible to use it without cargo almost everything uses that, it's possible to keep dependencies to a minimum but the indirect dep tree blows up fast
<Randolf>
Is there a serious interest in adding UNIX sockets support?
<Randolf>
I do like the idea of using this because it's more efficient -- I find with SMTP server milters that UNIX sockets perform better than TCP/IP connections (even through localhost).
<laanwj>
i have some ancient PRs where i implemented UNIX socket support, it's easy for RPC at the server side, at the client side it is blocked on a libevent issue in libevent http, which is still not solved upstream
<laanwj>
so we basically cannot do it, at least in bitcoin-cli, without some really ugly hacks around libevent
<Randolf>
Yuck.
<Randolf>
So this seems more like a libevent issue.
<laanwj>
if you don't care about bitcoin-cli (e.g. use it from python) server-side might be enough thogh
<Randolf>
Perhaps it should be implemented on the server-side anyway, and then eventually the cli could be fixed over time?
<laanwj>
yes, i filed a PR to implement the necessary support in libevent but it was never merged
<Randolf>
Which PR?
<laanwj>
it's kind of an annoyance for me so i don't really want to go back to it
<Randolf>
I want to follow it, at least. I don't know if I can pick it up though as I'm not a seasoned C/C++ developer yet.
<laanwj>
i did suggest doing only the server side at the time, but this made it controversial somehow
<laanwj>
i'm actually really bad at getting my own PRs in open source projects merged
bitcoin-git has joined #bitcoin-core-dev
<bitcoin-git>
[bitcoin] darosior opened pull request #22779: policy: correct (lower) the dust threshold for Taproot outputs (master...taproot_dust_limit) https://github.com/bitcoin/bitcoin/pull/22779
bitcoin-git has left #bitcoin-core-dev [#bitcoin-core-dev]
<laanwj>
another reason why i wish we had used it from the beginning because in that case, support for UNIX sockets in the HTTP lib would have been a hard requirement so we'd never have chosen libevent's in the first place-
<sipa>
iirc, my suggestion at the time was just dropping libevent for bitcoin-cli, as bitcoind-compatible-httpd should be very simple to implement by hand :)
<sipa>
of course, suggestions are easy if you don't follow up with code
<laanwj>
i dread getting into implementing a HTTP client, it sounds easy in theory but also i have enough experience with the avalanche of inadvertent complexity involved in such projects by now
videre_e has left #bitcoin-core-dev [leave!]
<laanwj>
can't we just use libcurl *ducks*
<Randolf>
laanwj: Even with Qt5's libraries that take care of various tedious aspects of the implementation details, it's still hard (so I agree with you).
* Randolf
laughs
<sipa>
laanwj: a fully compliant HTTP client sounds terrible, but my point is that we don't actually need that, just something that's compatible with bitcoind. but also, i'm not actually volunteering to write this :)
<laanwj>
sipa: yeah, true, it's just that we now have squashed most of the usability issues in bitcoin-cli (after the libevent switch), that would basically start over
<laanwj>
most actual work will subtle things like error handling, timeouts, etc
<laanwj>
then again, if someone wants to take this on, i'm not exactly stopping them
<Randolf>
That doesn't sound like a minor undertaking.
<sipa>
we could use boost::beast *ducks*
<laanwj>
it doesn't, and seeing that it's not been picked up after so many years, it probably isn't really that much "serious interest" in it (the original question), just a would-be-nice thing
<laanwj>
sipa: hah
<Randolf>
Boost created so many problems for me that I avoid it as if it were the plague -- they keep breaking backward compatibility, and then on some systems it's not possible to run different software that each depend on versions of the Boost libraries that aren't compatible.
<sipa>
i think our current effort is reducing boost usage to the extent possible to headers-only libraries
<sipa>
which at least avoid all the runtime build/link/compatibility issues
<Randolf>
I remember -- I was part of the meetings where that was decided.
<Randolf>
As I recall, it wasn't a difficult decision.
<laanwj>
right, the impulse has been to move away from boost, long, long ago it used boost::asio for the RPC socket it was one of the first things
<laanwj>
there were a lot of stability issues and from what i remember it had a high overhead, there's zero chance of going back to that
<laanwj>
libevent is good performance-wise but the HTTP library is very, very limited
<laanwj>
but we deemed it'd be enough for our use for RPC (which is also quite limited)
bitcoin-git has left #bitcoin-core-dev [#bitcoin-core-dev]
<laanwj>
which, everything considered, it mostly also is-we have run into some limitations around streaming data (without buffering it all in memory) and UNIX sockets
Henrik has quit [Quit: My MacBook Air has gone to sleep. ZZZzzz…]
Guest65 has joined #bitcoin-core-dev
Randolf has quit [Quit: Leaving]
Guest65 has quit [Client Quit]
sipsorcery has quit [Ping timeout: 240 seconds]
sipsorcery has joined #bitcoin-core-dev
lkqwejhhgasdjhgn has quit [Quit: Konversation terminated!]
<bitcoin-git>
bitcoin/master 75290ae Pieter Wuille: Drop us=... message in net debug for sending version message
<bitcoin-git>
bitcoin/master dbcb574 MarcoFalke: Merge bitcoin/bitcoin#20541: Move special CAddress-without-nTime logic to ...
bitcoin-git has left #bitcoin-core-dev [#bitcoin-core-dev]
bitcoin-git has joined #bitcoin-core-dev
<bitcoin-git>
[bitcoin] MarcoFalke merged pull request #20541: Move special CAddress-without-nTime logic to net_processing (master...202012_addr_without_time_is_no_addr) https://github.com/bitcoin/bitcoin/pull/20541
bitcoin-git has left #bitcoin-core-dev [#bitcoin-core-dev]
<theStack>
could one of the maintainers take a look at #22565? with seven ACKs (including one by kallewoof, the original author of the script), this PR should be ripe for merging
<bitcoin-git>
[bitcoin] Saibato opened pull request #22781: wallet: fix the behavior of IsHDEnabled, return false in case of a blank hd wallet. (master...fix_ishdenabled) https://github.com/bitcoin/bitcoin/pull/22781
bitcoin-git has left #bitcoin-core-dev [#bitcoin-core-dev]
<jonatack>
maybe it's just my peers.dat or a coincidence
bitcoin-git has left #bitcoin-core-dev [#bitcoin-core-dev]
<jonatack>
hm, i haven't been able to reproduce the issue on the other chain, only on mainnet. but it's the first time the addrman check failed for me (five times out of five restarts) after a couple of months of running it
<jonatack>
*chains
<jonatack>
"2021-08-23T17:25:48Z [init] addrman lost 22813 new and 2 tried addresses due to collisions or invalid addresses"
<jonatack>
it had been about a week since the previous restart
<sipa>
and it's actually a cleanly created peers.dat, or something from a possibly older version of the codebase?
<jonatack>
sipa: first with a well-used peers.dat. a couple days ago i noticed that it strangely lost about 30k addresses. then today the check failed. i then renamed the corrupt peers.dat and restarted. all well. i then restarted a few minutes later. re-fail.
<jonatack>
on current master dbcb5742c48f
<MarcoFalke>
So it doesn't happen on previous master commits?
<jonatack>
i was also testing #21879 rebased to master just before. let me reverify that it's not that.
<jonatack>
2021-08-23T19:18:27Z [init] Loaded 5021 addresses from peers.dat 201m
<jonatack>
-> the first check passes
nathanael has joined #bitcoin-core-dev
<MarcoFalke>
Oh I found the issue. It is asmap related.
<MarcoFalke>
jonatack: Are you using asmap?
<jonatack>
yes
<jonatack>
the second check fails after addcon thread start / opencon thread start
<MarcoFalke>
The asmap must be loaded before addrman
<jonatack>
ah. init order as laanwj suggested
<MarcoFalke>
181a1207ba made it load in the wrong order
<MarcoFalke>
I won't be working on a fix (bed time)
<MarcoFalke>
Unless it is still unfixed tomorrow, heh
<jonatack>
makes sense. i'll have a look as i'm now used to reproducing it :D
sipsorcery has joined #bitcoin-core-dev
<jonatack>
that explains why i was only seeing it on mainnet, too
grettke has joined #bitcoin-core-dev
<sipa>
jonatack: nice catch
<sipa>
loading asmap after loading addrman would explain it
<sipa>
both the lost addresses and the consistency check
<jnewbery>
jonatack: MarcoFalke: good catch. Thank you - I'll work on a fix tomorrow.
<jnewbery>
I didn't realise that connman was reaching into addrman to modify its internal data after construction. I think the fix is to pass asmap as a parameter to addrman's ctor, set it in the intializer list and make it const.
<sipa>
that sounds cleaner
<jnewbery>
Eventually asmap probably shouldn't be part of addrman at all, but that can wait until later
<laanwj>
while reviewing i noticed the ordering change but didn't realize this interaction was there
TheCharlatan has quit [Ping timeout: 240 seconds]
<jonatack>
jnewbery: agree, improving separation and less reaching into other class internals will avoid future hard-to-spot issues
TheCharlatan has joined #bitcoin-core-dev
<jonatack>
a test wouldn't have caught this unless it ran both -asmap and -checkaddrman...best to make the code cleaner