--- Log opened Sat Jun 05 00:00:29 2021
03:36 -!- undeath [~undeath@user/undeath] has joined #joinmarket
03:39 < JoinMarketRelay> [hackint/nikolai] waxwing: should i encrypt it with your key?
03:55 < waxwing> nikolai no worries, i've pinged belcher_ and he encrypted it to mine too. taking a quick look now.
03:56 < JoinMarketRelay> [hackint/nikolai] it also happened once more during the night
03:57 < waxwing> did you strip out the deadbeef markers? because i'm not seeing it in the deserialized tx here.
03:57 < JoinMarketRelay> [hackint/nikolai] no, i have not stripped any data, all the lines i've sent are the same as in the log file
03:58 < JoinMarketRelay> [hackint/nikolai] actually, i've only changed timestamp but that's unrelevant
03:58 < waxwing> ok, maybe a different tx? this one does not have those markers (which you would see in pre-0.8.2 and post 0.7.0 JM)
04:00 < waxwing> also 'it happened once more' - do you mean, a tx with an absurdly high fee, which didn't go through, or something else?
04:02 < JoinMarketRelay> [hackint/nikolai] the second one had deadbeefs, there were 3 txes in total with absurd high fee - since reporting the first one until today morning. i've sent two
04:03 < waxwing> oh you've sent two? i'll ping belcher_ again i guess.
04:03 < waxwing> to send me the second.
04:03 < belcher_> wheres the second?
04:03 < JoinMarketRelay> [hackint/nikolai] lets make this easier. what's your pgp key?
04:04 < JoinMarketRelay> [hackint/nikolai] it was in the same file
04:04 < belcher_> ah i see, yes the file contained two txes
04:04 < waxwing> oh did it?
04:04 < belcher_> his key is https://github.com/JoinMarket-Org/joinmarket-clientserver/blob/master/pubkeys/AdamGibson.asc fwiw
04:05 -!- belcher_ is now known as belcher
04:05 < waxwing> yes thanks but i already have the decryption. i just didn't see two transactions.
04:05 < waxwing> i see it now, thanks.
04:05 < JoinMarketRelay> [hackint/nikolai] good :)
04:06 < waxwing> yeah the second one has them. interesting.
04:07 < waxwing> i can see all details and everything looks normal, except the most important one of course - the (expected) tx fee, i'll need to do some shenanigans to read the input values off the blockchain.
04:07 < JoinMarketRelay> [hackint/nikolai] i tried to provide less data at first, but if you need some additional let me know
04:07 < waxwing> with our psbt code we can call 'get_fee' but of course that's because it already did the hard work of finding the prevout value.
04:08 < waxwing> all i can think of doing right now is verifying that these txs have absurd fees. after that it'll be a bit hard to know. i'm still not seeing an obvious way you could somehow configure JM and end up with stupid fees like 0.25
04:10 < waxwing> so do i take it, the reason you know about the absurd fee is specifically because you  tried to rebroadcast and Core refused?
04:10 < waxwing> oh no wait. these are the unsigned versions. you couldnt' broadcast them.
04:12 < waxwing> nikolai :  if you could explain that point ^ it'd help. if you just saw they were not broadcast, but never had the full signed version .. wouldn't you need to programmatically check all the input values yourself, to figure out the fee?
04:16 < JoinMarketRelay> [hackint/nikolai] correct, the reason i know about absurd fee is that i tried to broadcast it manually. i thought it's a signed tx as it replied with 'goodtx'. and below that we had a timeout with info that it was not broadcasted - so it tried manually to see what happens. i tried. sorry but i do not have much experience with handling raw transactions, i've used
04:16 < JoinMarketRelay> a public tx broadcast tool with tor+curl.
04:16 < waxwing> but wait, you couldn't just send that serialized tx in and try to broadcast it. it doesn't have any signatures!
04:17 < waxwing> if the tool you used replied with 'absurd fee' or whatever it might just be that it fails to read all the inputs or something .. not sure, but clearly it can't be broadcast without any sigs.
04:18 < waxwing> do you see the idea? what the taker is sending you there is a tx *template*; you reply by sending the sigs on *your* inputs, but you never see the full transaction with sigs, until the taker broadcasts it. if they don't, it just times out.
04:18 < waxwing> that happens quite a lot for all kinds of reasons.
04:18 < belcher> i think "goodtx" just refers to some check that the maker is earning money from the coinjoin
04:18 < waxwing> yes, from `verify_unsigned_tx`
04:20 < JoinMarketRelay> [hackint/nikolai] you made it a bit more clear. i'll have to finish some other work first, sit and look through the code as i'm interested more how it works. sorry for the false positive and wasting your time
04:20 < waxwing> no worries at all. it's good that people are checking things out.
04:21 < waxwing> but it's curious/interesting if a public broadcast tool is returning "big fee" in case a user submits tx with no signatures.
04:23 < waxwing> on blockstream.info i got, for the second one:
04:23 < waxwing> sendrawtransaction RPC error: {"code":-26,"message":"scriptsig-not-pushonly"}
04:23 < waxwing> which makes sense because of the "deadbeef"s in there.
04:23 < waxwing> for the first i got: sendrawtransaction RPC error: {"code":-25,"message":"bad-txns-inputs-missingorspent"}
04:24 < waxwing> but that might be misleading, because it might immediately fail that check if some of the inputs are now spent, since that tx first happened (overwhelmingly likely)
04:24 < waxwing> so i won't see what error it *would* have been.
04:25 < JoinMarketRelay> [hackint/nikolai] i think either for the first one or the second one also on blockstream.info i received the absurd fee.
04:26 < JoinMarketRelay> [hackint/nikolai] i use it like this - txhex="$1"; curl --proxy socks5h://127.0.0.1:9050 -X POST --data "${txhex}" https://blockstream.info/api/tx ... probably because some of the inputs have been spent as you said already
04:27 < JoinMarketRelay> [hackint/nikolai] i was wondering if something like this could exist in JM, where a taker wouldn't have to broadcast himself but use a proxy - for example such sertvice over tor - giving more privacy
04:27 < undeath> there is an option to have makers broadcast a tx. but is that actually implemented?
04:29 < waxwing> yes
04:29 < waxwing> for a long time it wasn't but it was re-done i think early last year
04:30 < waxwing> [POLICY].tx_broadcast = random_peer etc.
04:30 < JoinMarketRelay> [hackint/nikolai] what happens if all peers have 'not-self'
04:31 < waxwing> not-self only refers to *this* tx, it doesn't mean they won't broadcast anything
04:31 < waxwing> but .. it's only a thing that takers are choosing, anyway
04:33 < JoinMarketRelay> [hackint/nikolai] thanks for the explanation. will be around. need to do some work, forgot pwd/pim/key for container decryption that contains a btc seed. hasn't been opened for years. wrote a cracking script but so far i couldn't get the combinations. wish me luck ://
04:34 < belcher> the sendrawtransaction error messages are incredibly opaque in my experience
04:35 < belcher> PR to set the default tx_broadcast = random_peer ?
04:35 < belcher> hold on ill just code it myself and open a PR, sec
05:16 < belcher> waxwing iv rebased the fidelity bonds PR onto master as you requested
06:13 < waxwing> belcher, thanks 
06:53 -!- undeath [~undeath@user/undeath] has quit [Quit: WeeChat 3.1]
09:26 < waxwing> i have pushed this direct to master: https://github.com/JoinMarket-Org/joinmarket-clientserver/commit/b25d03c8d4b78df4bdf3b1ffbd6315c7f00599f8
09:27 < waxwing> 4 days ago the custom change address feature addition introduced a bug such that change addresses would get reused in tumbler runs.
09:27 < waxwing> so that fixes that.
09:28 -!- ghost43 [~ghost43@gateway/tor-sasl/ghost43] has quit [Remote host closed the connection]
09:28 < waxwing> i'm kind of annoyed with myself for not being more adamant to the PR author that changing the interpretation of `my_change_addr` like that was dubious, but i suppose more to the point i should have noticed this specific very naughty behaviour, in tumbler.
09:28 < waxwing> of course i would usually PR but having that out there at all is bad enough, hence immediate push. still, review it "retroactively" please.
09:28 -!- ghost43 [~ghost43@gateway/tor-sasl/ghost43] has joined #joinmarket
13:11 -!- stortz [~stortz@2804:30:7f5d:d000:9c3:753d:1d13:f563] has joined #joinmarket
13:13 -!- stortz [~stortz@2804:30:7f5d:d000:9c3:753d:1d13:f563] has quit [Changing host]
13:13 -!- stortz [~stortz@user/stortz] has joined #joinmarket
13:32 -!- stortz [~stortz@user/stortz] has quit [Quit: Client closed]
20:32 -!- xyy [~xyy@2001:470:69fc:105::f2d] has quit [Remote host closed the connection]
20:32 -!- openoms[m] [~openomsma@2001:470:69fc:105::c2f] has quit [Write error: Connection reset by peer]
20:32 -!- Evanito[m] [~evanito@2001:470:69fc:105::1ec] has quit [Remote host closed the connection]
20:33 -!- Evanito[m] [~evanito@2001:470:69fc:105::1ec] has joined #joinmarket
20:37 -!- xyy [~xyy@2001:470:69fc:105::f2d] has joined #joinmarket
20:37 -!- openoms[m] [~openomsma@2001:470:69fc:105::c2f] has joined #joinmarket
21:33 -!- belcher_ [~belcher@user/belcher] has joined #joinmarket
21:35 -!- michaelfolkson2 [~michaelfo@138.68.143.20] has joined #joinmarket
21:36 -!- belcher [~belcher@user/belcher] has quit [Ping timeout: 268 seconds]
21:37 -!- pigeons_ [~pigeons@androzani.sysevolve.com] has joined #joinmarket
21:37 -!- michaelfolkson [~michaelfo@138.68.143.20] has quit [Ping timeout: 264 seconds]
21:37 -!- pigeons [~pigeons@androzani.sysevolve.com] has quit [Ping timeout: 264 seconds]
23:36 -!- Netsplit *.net <-> *.split quits: dr-orlovsky, emzy, berndj, stoner19, openoms[m]
23:37 -!- Netsplit over, joins: dr-orlovsky, berndj, stoner19, emzy, openoms[m]
--- Log closed Sun Jun 06 00:00:30 2021