< bitcoin-git>
[bitcoin] practicalswift opened pull request #20581: Don't make "in" parameters look like "out"/"in-out" parameters: pass by ref to const instead of ref to non-const (master...const-ref-parameters) https://github.com/bitcoin/bitcoin/pull/20581
< hebasto>
promag: it was useful (at least for me) when I was playing with wallet stuff
< promag>
hebasto: how so? what you used it for?
< hebasto>
promag: opening multiple copies of sqlite wallet
< bitcoin-git>
[bitcoin] wodry opened pull request #20582: [doc] Fix documenation of options onlynet and listenonion (master...patch-1) https://github.com/bitcoin/bitcoin/pull/20582
< theStack>
i'm just running my first gitian build. is there any reason why the use of Docker is not more encouraged (it's not even mentioned in gitian-building.md)? are there potential problems/drawbacks? setting up an extra VM seems to be massive overkill for this task
< luke-jr>
Docker would be the extra ;)
< theStack>
installing docker to the distro of your choice and simply passing "--docker" to the gitian build script (if it's not that easy i may missed somethig :)) seems less of an effort to me than to setting up a virtual machine
< luke-jr>
just wait until you see the macOS hoops
< theStack>
grated, i was focussing on linux here
< theStack>
*granted
< hebasto>
theStack: conventional lxc works fine with less overhead
< bitcoin-git>
bitcoin/master 12dcdaa practicalswift: Don't make "in" parameters look like "out"/"in-out" parameters: pass by re...
< bitcoin-git>
bitcoin/master 9385549 MarcoFalke: Merge #20581: Don't make "in" parameters look like "out"/"in-out" paramete...
< bitcoin-git>
[bitcoin] MarcoFalke merged pull request #20581: Don't make "in" parameters look like "out"/"in-out" parameters: pass by ref to const instead of ref to non-const (master...const-ref-parameters) https://github.com/bitcoin/bitcoin/pull/20581
< phantomcircuit>
there's no reason to drop peers sending you commands you don't understand
< phantomcircuit>
it's literally not a DoS
< sipa>
you could argue it is one if there were no other way to waste nodes' bandwidth
< sipa>
given that that is (a) trivial and (b) solutions for that (e.g. deprioritizing peers that send too much) aren't orthogonal to permitting unknown messages or not, it isn't
< phantomcircuit>
sipa, no matter what you do i can always waste your inbound bandwidth
< phantomcircuit>
there's just no way to prevent that
< luke-jr>
pls dont udp flood me phantomcircuit
< phantomcircuit>
what luke-jr said
< bitcoin-git>
[bitcoin] MarcoFalke opened pull request #20583: rpc: Add missing BlockUntilSyncedToCurrentChain to wallet RPCs (master...2012-walletSync) https://github.com/bitcoin/bitcoin/pull/20583
< sipa>
phantomcircuit: well there are CPU costs involved with handling messages you're ignoring (making its way to the network stack into process, checksum computation, ...) that just spamming UDP messages doesn't have
< sipa>
my point is those costs are negligable compared to say spamming with ping or tx messages
< phantomcircuit>
sipa, sure, but the worst thing you can do is cause the cpu to spin doing nonsense for a brief instant, in between doing real work
< sipa>
phantomcircuit: i'm not saying you're wrong - i obviously agree that there is no harm in just ignoring unknown messages
< sipa>
but i think it needs a slightly more nuanced explanation than "literally not a DoS"
< phantomcircuit>
the amount of cpu time used is roughly on par with the cpu time used to resolve udp packet fragmentation
< sipa>
yes, that sort of expalanation :)
< phantomcircuit>
sipa, it's certainly less of a DoS than just spamming someone with ip fragments
< bitcoin-git>
[bitcoin] practicalswift opened pull request #20584: Don't declare de facto const member functions as non-const. Don't declare de facto const reference variables as non-const. (master...const-member-fn-and-const-ref-var) https://github.com/bitcoin/bitcoin/pull/20584
< bitcoin-git>
[bitcoin] hebasto closed pull request #20571: refactor: Use deduction guide for std::array<uint8_t, N> (master...201204-u8) https://github.com/bitcoin/bitcoin/pull/20571