< sipa> cluelessperson: use getmempoolentry RPC
< cluelessperson> sipa: first I'd like to focus on getting the fee for every transaction in the last N blocks
< sipa> there is no way to do that without txindex
< sipa> (which is incompatible with pruning)
< cluelessperson> sipa: what about the next block mined?
< cluelessperson> sipa: If my software is just running, it can just check the transactions as they are being confirmed, before the UTXO is gone?
< sipa> cluelessperson: no
< sipa> you can for mempool transactions
< sipa> but not for blocs
< cluelessperson> sipa: is there a way to get all the UTXOs that the current block deals with? like some sort of subscriptions?
< cluelessperson> sipa: my goal is to understand fees that go through, not just those waiting
< cluelessperson> I'm concerned that I'll miss things if I just watch for mempool transactions disappearing
< cluelessperson> sipa: unless! maybe I can measure mempool transactions, and then verify they went into the last block when they disappear?
< sipa> cluelessperson: won't work for all transactions
< sipa> some transactions are not seen until they appear in a block
< Sentineo> miner not broadcasting them? or somehow they get lost during transmit to my node sipa ?
< sipa> miners have full discretion about what transactions they put in blocks
< promag> jimpo: why not use std::queue for m_queue?
< jimpo> I need an iterator where elements can be erased for the Interrupt method
< jimpo> Originally was using deque, then I changed it
< promag> initially BlockUntilSyncedToCurrentChain can take a while right?
< jimpo> Not really, it should just return false until it's gotten in sync once, then it should process blocks pretty quickly after that
< agolx> I'd like to make my bitcoind collect any valid transaction and not evict things if it gets full. I know there's a max mem limit that you can increase, but is there a minimum fee setting somewhere? and is it configurable or do I have to modify a header and rebuild.
< agolx> thx
< agolx> @here
< sipa> agolx: -minrelayfee
< agolx> thanks
< agolx> works in the config file right?
< sipa> yes
< sipa> like every setting
< agolx> there are literally zero hits in source for minrelayfee
< agolx> so no
< aj> minrelaytxfee
< agolx> <3
< aj> also, apparently patches work better when you compile them
< agolx> are when yhou don't rm -rf your desktop like i did :)
< bitcoin-git> [bitcoin] laanwj closed pull request #11740: Implement BIP159 NODE_NETWORK_LIMITED (pruned peers) *signaling only* (master...2017/11/NNL_signaling) https://github.com/bitcoin/bitcoin/pull/11740
< achow101> \o/
< mesh_> \o/
< cluelessperson> does bitcoin core support an "addrindex=1" ?
< instagibbs> cluelessperson, no, there are various efforts that add it internally or externally
< andytoshi> on regtest, under what conditions will `sendrawtransaction` silently succeed but the tx then not get confirmed by a `generate` call?
< sipa> andytoshi: what height are you at?
< andytoshi> sipa: 148
< andytoshi> my coins are 100 blocks old and i have prematurewitness on
< sipa> andytoshi: segwit only activates at block 432 iirc
< andytoshi> right, but with the -prematurewitness flag it should activate immediately, right?
< andytoshi> initially i was getting an RPC error no-witness-yet or something like that, and -prematurewitness made that error stop happening, so i think i'm in the clear
< andytoshi> i'll jump 1000 blocks forward and see what happens..
< arubi> do minrelaytxfee \ blockmintxfee behave the same on regtest? maybe the fee is too low for the defaults?
< andytoshi> there's an explicit error when the fee is too low
< andytoshi> jumping forward 1000 blocks seems to fix it, i think this may be a bug (or a misunderstanding on my part) in what prematurewitness does
< andytoshi> i'll circle back, i've got another bug to chase down in my software..
< andytoshi> ok, got that. now jumping ahead 1000 vs 100 blocks at startup is the difference between my software's integration tests passing or failing. i guess it's just the segwit activation height then.
< andytoshi> thanks. if somebody familiar with -prematurewitness can comment on whether this sounds like a bug, i'll open an issue
< sipa> andytoshi: prematurewitness is just about the mempool, iirc
< foobar_> hi
< foobar_> does bitcoin rule or what
< foobar_> doesnt wtf bye
< sipa> andytoshi: yes, the block assembly code won't include segwit transactions until the rule is enforced by consensus