--- Day changed Mon Dec 10 2018
00:31 -!- undeath [~undeath@hashcat/team/undeath] has joined #joinmarket     
03:07 -!- belcher_ [~user@unaffiliated/belcher] has joined #joinmarket     
03:08 -!- beIcher [~user@unaffiliated/belcher] has quit [Ping timeout: 250 seconds]     
05:25 -!- belcher [~belcher@unaffiliated/belcher] has quit [Ping timeout: 252 seconds]     
05:39 -!- belcher [~belcher@unaffiliated/belcher] has joined #joinmarket     
06:26 < waxwing> uhh GitHub-bot are you OK?
08:33 -!- stoner19_ [~stoner19@vmi221374.contaboserver.net] has joined #joinmarket     
08:33 -!- stoner19_ [~stoner19@vmi221374.contaboserver.net] has quit [Changing host]     
08:33 -!- stoner19_ [~stoner19@unaffiliated/stoner19] has joined #joinmarket     
08:33 -!- stoner19 [~stoner19@unaffiliated/stoner19] has quit [Read error: Connection reset by peer]     
08:34 -!- stoner19_ is now known as stoner19     
09:24 -!- mr_paz [~mr_paz@84.39.112.85] has joined #joinmarket     
09:38 < waxwing> been spending a lot of time trying to figure this out. no success really yet. gdb is giving me output very similar to this: https://pastebin.com/5bJ41ri9
09:38 < waxwing> which is coming out of the above thread.
09:39 < waxwing> the QSocketNotifier constructor seems to not be the only place it fails, i think when the 'activated' signal is triggered (of that notifier object), that may also be calling a segfault, but it's not so easy to figure out. what i can't seem to do is get to debug what is triggering/activating the signal, so, because its just segfaulting on receiving it, i'm in the dark.
09:40 < waxwing> the pattern of behaviour is: python2 only, pyside2 using qt5reactor, segfaults at some point after asknig for the orderbook on mainnet.
09:41 < waxwing> there seem to be lots of events happening at that point as you'd expect :) but at this point basically no clue which events are triggering the crash, or if that even matters, since from the above bug report discussion, it may well be the case that the underlying pyside2 code is just wrong (albeit, if i knew more i might be able to work round it)
09:45 < undeath> what sometimes works very well for figuring out where it goes wrong is compiling with asan
09:45 < waxwing> what's asan?
09:45 < undeath> a gcc plugin, address sanitizer
09:46 < undeath> a crash may look like this https://gist.github.com/undeath/d13ee81b664a2400fb8da9567b65b1ce
09:46 < waxwing> ok. so what should i be compiling though?
09:46 < undeath> probably qt5reactor/pyside2 and/or qt
09:47 < waxwing> ok, i guess that would still work if the python exe was not compiled with it, right?
09:47 < undeath> yes     
09:50 -!- Giszmo [~leo@190.162.241.129] has quit [Ping timeout: 250 seconds]     
09:50 < undeath> but it would be far more sane to just deprecate py2 for jm :)
09:53 < waxwing> well, here it'd be deprecate for joinmarketqt, which is obviously more reasonable
09:53 < waxwing> i guess there's a correlation between people using py2 and people who are very techy, since they're using old setups ... but not sure how strong that is, might be other cases.
09:54 < waxwing> is there anybody in the channel for whom switching to Python3 is a problem?
10:01 < pigeons> i'm still transitioning myself many places, but it wouldnt be a problem
10:11 -!- Giszmo [~leo@ip-100-233.219.201.nextelmovil.cl] has joined #joinmarket     
10:47 -!- GitHub176 [GitHub176@gateway/service/github.com/x-wapgobsjyvvultwa] has joined #joinmarket     
10:47 < GitHub176> [joinmarket-clientserver] kristapsk opened pull request #249: Write pid to wallet lockfile, display it when lock already in place (master...walletlockpid) https://git.io/fpHFK
10:47 -!- GitHub176 [GitHub176@gateway/service/github.com/x-wapgobsjyvvultwa] has left #joinmarket []     
10:54 < waxwing> pigeons, yeah, in a sense i realise it's kind of a weird question, because it's not like we'd be making breaking changes. you coudl still run an old version anyway.
10:54 < waxwing> i mean obviously it'd be a big signal to upgrade though, we'd deprecate it at some point, albeit with a long warning period hopefully.
10:54 < waxwing> and anyway i'm only talking about Qt, you can run master (and next release etc.) on command line with py2
10:55 < undeath> https://travis-ci.org/JoinMarket-Org/joinmarket-clientserver/requests
10:55 < undeath> ffs travis
10:56 < arubi> :(     
10:56 < arubi> maybe waxwing can restart it manually
10:57 < arubi> oh no it didn't even get a job slot
10:59 < waxwing> he has to email them i think
11:00 < waxwing> in this case it's not important
11:00 < arubi> yep.  good thing they do respond quickly
11:03 -!- belcher_ [~user@unaffiliated/belcher] has quit [Ping timeout: 272 seconds]     
11:04 -!- belcher_ [~user@unaffiliated/belcher] has joined #joinmarket     
11:05 -!- mr_paz [~mr_paz@84.39.112.85] has quit [Ping timeout: 252 seconds]     
11:07 -!- mr_paz [~mr_paz@31.7.57.246] has joined #joinmarket     
11:28 -!- belcher_ [~user@unaffiliated/belcher] has quit [Ping timeout: 250 seconds]     
11:29 -!- belcher_ [~user@unaffiliated/belcher] has joined #joinmarket     
11:45 -!- belcher_ [~user@unaffiliated/belcher] has quit [Ping timeout: 244 seconds]     
11:45 -!- belcher_ [~user@unaffiliated/belcher] has joined #joinmarket     
12:32 -!- mr_paz [~mr_paz@31.7.57.246] has quit [Remote host closed the connection]     
13:09 -!- Giszmo [~leo@ip-100-233.219.201.nextelmovil.cl] has quit [Ping timeout: 268 seconds]     
13:29 -!- Giszmo [~leo@190.162.241.129] has joined #joinmarket     
14:16 -!- midnightmagic [~midnightm@unaffiliated/midnightmagic] has quit [Ping timeout: 264 seconds]     
14:20 -!- midnightmagic [~midnightm@unaffiliated/midnightmagic] has joined #joinmarket     
15:00 -!- GitHub23 [GitHub23@gateway/service/github.com/x-phryroakyhvwyjwt] has joined #joinmarket     
15:00 < GitHub23> [joinmarket-clientserver] undeath opened pull request #250: fix crash on IRC notice with non-ascii characters (master...fix-irc-control-chars) https://git.io/fpQ3R
15:00 -!- GitHub23 [GitHub23@gateway/service/github.com/x-phryroakyhvwyjwt] has left #joinmarket []     
15:01 < waxwing> undeath, hmm, interesting, - i'm guessing that's from the new py3 compat changes? because i'm sure they use non-ascii characters all the time.
15:01 < waxwing> but wait, that can't be right either, i've been using that code live for a week+ now
15:02 < undeath> has a notice ever been sent to your client that went through wlog?
15:02 < waxwing> i saw that notice about not clicking things recently
15:02 < waxwing> was that was caused the crash?
15:02 < undeath> no idea, I never got the log message
15:03 < undeath> I'm not running the latest commit, maybe it's not happening on the latest master?
15:05 < waxwing> oh, hmm
15:06 < waxwing> but i get notices like this every time i log on
15:06 < waxwing> https://pastebin.com/FviFdVX3
15:06 < undeath> no, the one causing this crash was certainly no regular one, that bot was running for days
15:07 < waxwing> oh damn, that doesn't show up the non-ascii, but they're there
15:10 < waxwing> sure was easier when it was just python2 :)
15:10 < undeath> well, there errors='ignore' sure is nice
15:10 < undeath> but py2 doesn't have it
15:11 < waxwing> but i really don't get how this happened
15:11 < waxwing> as  you can see in my pastebin, there's "(unhandled) noticed: " and etc
15:11 < undeath> yeah, not really sure myself
15:11 < waxwing> those notices are sent in every log in  and they almost always have non-ascii chars in them
15:11 < waxwing> wait i should actually read the stack trace :)
15:12 < undeath> the error is not caused by the notice send on login
15:12 < waxwing> right
15:12 < waxwing> what commit are you running on out of interest?
15:12 < undeath> a rather old one, right after the coincurve switch
15:15 < waxwing> hmph. looking at blame i can't see how the latest changes are really doing anything that could change things
15:15 < waxwing> and i'm especially bewildered because of the notices on log in
15:17 < waxwing> it might be useful to know the exact commit undeath 
15:17 < waxwing> since it's doubtless to do with str() behaviour right.
15:18 < undeath> abad5977fe890456511ea894d8d68258e81cbfc2
15:24 < waxwing> i dunno exactly. have received many noticed calls, and they contain non-ascii characters, both before and after the change. seems like it should be to do with the change to using future/builtins, but .. hmm i guess i can just test what the current wlog does with the current code.
15:24 < waxwing> sorry if i'm overanalysing but it bothers me not knowing, this has always been part of the deal (having non-ascii) and somehow it's never been an issue.
15:26 < undeath> on the console I can easily reproduce it
15:26 < undeath> from builtins import *
15:26 < undeath> str('\xc3')
15:27 < undeath> will raise the exception
15:29 < waxwing> yes i see now, it's my error. they are not using non-ascii, just stuff like '\x01'
15:30 < undeath> my fix breaks py3 though, as travis tells
15:30 < waxwing> so, the picture is: in python2, after the compatibility change, if a notice contains non ascii, we have a crash condition
15:30 < undeath> i guess that's the gist
15:30 < waxwing> i'm having trouble finding the aberrant notice in the cgan log, but .. ofc it may be agora
15:30 < undeath> no idea what server it came from
15:31 < waxwing> yeah doesn't matter
15:34 < undeath> force-pushed, removed the errors='strict'
15:34 < undeath> py3 seems to have a str there already, so not needed
15:37 < undeath> might actually be some weirdness introduced by the future lib
15:47 < undeath> oh, sure, the buittins import effectively makes the str() a unicode() in py2. in py3 we already receive a str for some reason, while in py2 we receive a bytes obj
15:49 < undeath> this might be fixed on latest master because we make sure to import the future lib before everything else, so twisted may behave differently, not sure
15:52 -!- undeath [~undeath@hashcat/team/undeath] has quit [Quit: WeeChat 2.3]     
15:58 -!- bsm117532 [~mcelrath@c-24-61-184-150.hsd1.ma.comcast.net] has joined #joinmarket     
17:49 -!- AgoraRelay [~jmrelayfn@p5DE4AAA8.dip0.t-ipconnect.de] has quit [Ping timeout: 244 seconds]     
18:02 -!- AgoraRelay [~jmrelayfn@p5DE4AAFA.dip0.t-ipconnect.de] has joined #joinmarket     
18:21 -!- bsm117532 [~mcelrath@c-24-61-184-150.hsd1.ma.comcast.net] has quit [Ping timeout: 272 seconds]     
21:59 -!- v_unimportant_pe [~user@46.243.206.135] has quit [Ping timeout: 268 seconds]     
22:00 -!- v_unimportant_pe [~user@46.243.206.131] has joined #joinmarket     
23:02 -!- felix34_ [~muh@5.2.76.19] has quit [Quit: leaving]