< da2ce7> kanzure, is it possible for to fix the headers of the emails so that regular email clients show the original author? - Or could it be policy that people can put their name/nym at the end of the email. I find it a frustration to work out the author of many of the mails now.
< kanzure> has this changed recently?
< achow101> da2ce7: the original author is usually in the reply-to field
< achow101> both gmail and protonmail show me "from: X via bitcoin-dev"
< da2ce7> achow101, kanzure, I worked it out, my computer had a contact card for the mailing list and then just showing than instead of the real-name. :)
< da2ce7> thankyou anyway.
< kanzure> now that we have solved the email protocol, let's move on to coinjoin
< provoostenator> Ok, this is new... I'm getting a 500 server error for #13937
< gribble> https://github.com/bitcoin/bitcoin/issues/13937 | HTTP Error 500: Internal Server Error
< provoostenator> (it's my rebase of #12138)
< gribble> https://github.com/bitcoin/bitcoin/issues/12138 | Track best-possible-headers by TheBlueMatt · Pull Request #12138 · bitcoin/bitcoin · GitHub
< fanquake> wumpus 14140, 14135, 14129, 14128, 14094, 14013 probably all mergeable
< wumpus> fanquake: thanks!
< fanquake> wumpus thanks. 14133 could also go in
< as1nc> hello folks, beautiful day here in south of france, bitcoin is doing great, what else can we ask ? :) oh yeah, anyone has any tips on how I could guess what type of a addresses is used (p2sh, p2pkh..) given an xpub, ypub or zpub key ? is that even possible ?
< as1nc> given an xpubkey, i would like to know the good derivation path i should use for that key, without having the user to specify it upfront
< BlueMatt> anyone else get a 500 trying to open #13937?
< gribble> https://github.com/bitcoin/bitcoin/issues/13937 | HTTP Error 500: Internal Server Error
< BlueMatt> heh, I guess gribble does too
< as1nc> same here
< michagogo> Hrm, weird, looks like I disconnected a week ago and half my channels didn't rejoin
< echeveria> michagogo: +r
< michagogo> Hm?
< harding> michagogo: since the recent spam attacks about a month ago, a lot of channels now require being a registered user (+r). If you were disconnected and don't use SASL auth, you probably weren't registered with nickserv when you reconnected.
< michagogo> Was there a problem with services a week ago? Because I ID with SASL...
< michagogo> Anyway
< echeveria> nickserv has also been funky and was rebooted a few times.
< michagogo> I may be able to get back to gitian again soon - working on migrating my VM over to my new laptop 🎉
< phantomcircuit> michagogo, services were broken and one of the servers decided to stop doing anything at all
< phantomcircuit> (my connection to it got pongs but that's it)
< michagogo> Heh
< michagogo> In other news: It seems on Windows 10 the OS itself runs in a VM, so Virtualbox doesn't work because Hyper-V is already using the CPU's virtualization capabilities as a bare-metal hypervisor
< michagogo> So I migrated the VM over to Hyper-V, converting from legacy to UEFI Secure Boot along the way
< michagogo> And it seems it worked!
< michagogo> (Apparently boot-repair doesn't like it when apt on the target system is configured to use localhost as a proxy)
< provoostenator> 13937 can be found here, though without PR comments: https://github.com/Sjors/bitcoin/commits/2018/08/best-header-tracking
< phantomcircuit> michagogo, windows uses hyper-v for some compatibility stuff
< phantomcircuit> iirc literally runs win xp
< michagogo> phantomcircuit: Huh? For what exactly?
< michagogo> I thought it was for the virtualization-based security
< phantomcircuit> michagogo, possibly they dont do this anymore but vista ran an entire copy of xp for binaries with the xp compatibility flag set
< sipa> phantomcircuit: that was an optional component, and it was pretty much just a VM
< sipa> by default it ran in a window, where had the xp desktop etc
< sipa> though there was a mode that would forward the windows in the VM to the outside video
< sipa> hmm, googling tells me it still exists and is more integrated now
< sipa> my experience with it is from win7 iirc
< phantomcircuit> sipa, i remember it being able to pull the window out of the vm and integrate it with the main ui
< sipa> yeah
< phantomcircuit> iirc it also had a shim to bring the local fs into the vm or something
< michagogo> Anything new I should know about building?
< phantomcircuit> that reminds me, it would be nice if the integration testing ran for each commit in a pull
< phantomcircuit> as it stands it's not easy to tell from a glance whether commits are themselves passing the tests
< gmaxwell> phantomcircuit: is your reason for asking because of the preference that every commit passes the test in order to aid bisecting? or?
< phantomcircuit> gmaxwell, yes
< sipa> phantomcircuit: there was some discussion about that lately; i don't remember where though
< sipa> unfortunately, it doesn't seem easy
< gmaxwell> it's straight forward with jenkins and I assume other traditional ci stuff.
< jamesob> phantomcircuit gmaxwell: that's going to amount to *a lot* of compute time given the amount of rebasing we do on large PRs
< jamesob> we could start with `make check` for each commit, probably
< phantomcircuit> in the (not to distant) past i've run into commits that simply dont build while bisecting
< andytoshi> i believe `git rebase master -x make` will simply try to build ... you could disable tests and gui on this
< gmaxwell> Passing make check for each commit is likely sufficient.
< phantomcircuit> indeed
< sipa> andytoshi: ooh, i didn't know about rebase -x
< sipa> jamesob: make check on every commit would be cool
< sipa> jamesob: it could even be a separate phase, between the lint checks and the full checks
< sipa> actually, that'd probably be a bad idea as we don't expect many failures in those, and they're probably even slower than the other runs
< gmaxwell> make check runs a lot of weird stuff.
< gmaxwell> maybe just one of the other CI tests that does the most trivial thing that involves actually starting the daemon.
< sipa> but for large PRs, this will likely introduce an order of magnitude increase in delay until results are available
< sipa> unless we can have one travis job for each
< michagogo> cfields: The depends source fallback is missing qt
< michagogo> 0.17.0's qt, at least
< jamesob> sipa: maybe we can have another service (e.g. circleci) run checks like that out of band with travis
< sipa> right
< polydin> herro?
< sipa> wercome
< polydin> lol
< polydin> got a question about pruning
< sipa> if you don't ask it, you'll never get an a swer
< polydin> i was running low on laptop space so i decided to run bitcoind prune=100000 to restrict disk space used to 100 GB
< polydin> but it is incredibly slow
< polydin> its been days and it still hasn't loaded
< polydin> is this normal?
< sipa> with pruning, disk flushing is much more frequent which can hurt performance, especially on systems with sloe i/o
< echeveria> you should ask in #bitcoin.
< sipa> i believe this is improved in 0.17
< polydin> sipa, thanks yeah i was wondering if there was a project in there i could work on
< polydin> if there was something i could work on to improve it
< polydin> but i guess you guys already did it!
< sipa> contributions on other issues are of course always welcome
< polydin> are there any other issues you can think of that need working other than the github list?
< polydin> like maybe a longer term thing to work on
< echeveria> polydin: there's a tag for low hanging, get started fruit.
< polydin> echeveria, ok thanks i guess i can just check that out
< echeveria> polydin: just hanging around here, there's often small tasks commented on
< polydin> echeveria, sorry could you rephrase that last sentence, not clear to me?
< sipa> polydin: it's useful to just hang out in this irc channel
< echeveria> polydin: many times in this channel, people will have suggestions about things that don't make it into issues. there's frequent opportunities to get your teeth into the code.
< sipa> also, getting familiar with the project, the source code, and its workflows takes time
< sipa> having a specific project is a good way to get going, but getting proficient takes time
< polydin> sipa, i know, i've honestly spent so much time reading the code but it seems like i'm better off just contributing to get into hte workflow
< polydin> i will start off with the low hanging fruit and build up to a project, thank you so much for your help!
< echeveria> polydin: you're invited to message me privately, or ask me on #bitcoin any questions you have about getting started.
< echeveria> or here, but this channel isn't great for in depth discussion.
< polydin> echeveria, thanks so much!
< polydin> is this channel more for just updates about specific PRs and such?
< sipa> polydin: no, this channel is about development of bitcoin core
< polydin> ok got it, again thanks guys
< echeveria> it's just meant to be somewhat low volume so people can kept up with the logs (I've made a point of reading the logs, even if I haven't been here), but not restricted to PR discussion.
< sipa> if you have questions about the code, specific PRs, workflow, ... feel free to ask here
< sipa> but not general discussion on bitcoin or usage
< echeveria> or how to pronounce sipa's name.
< polydin> lulz ok
< michagogo> Ooh, 0.17 builds on bionic?
< michagogo> Cool
< * michagogo> runs make-base-vm and heads towards bed
< achow101> michagogo: no guarantee that that's gonna work
< achow101> supposedly kvm stil doesn't work, even with the new vm-builder
< michagogo> Hmm?
< michagogo> That doesn’t sound promising
< michagogo> Though I’m using local
< michagogo> lxc
< michagogo> Wait, does gitian lxc still use vmbuilder and convert, or did it get changed to make it better?
< achow101> I think it uses vmbuilder
< michagogo> Crap
< achow101> I could be wrong
< achow101> you should keep an eye on your build though
< michagogo> I’ll see later (tomorrow or something) if it worked
< michagogo> It’s almost 2 am here
< michagogo> achow101: nope, looks like I remembered correctly
< michagogo> It uses debootstrap directly
< achow101> ah ok
< michagogo> Looks like it went in as a birthday gift for me :D
< luke-jr> michagogo: good luck. I had to install by hand
< michagogo> luke-jr: perhaps the image itself is broken, but `bin/make-base-vm —lxc —suite bionic` appears to have completed without errors
< leishman> is there a reason logs don't show up on https://botbot.me/freenode/bitcoin-core-dev/ anymore? Is there a bot user in this channel?
< achow101> leishman: perhaps the bot wasn't able to login after freenode's recent server downtime/maintenance/whatever happened a few days ago
< michagogo> luke-jr: hrm, make-clean-vm fails, some error about lxc-init and a container name or something
< achow101> michagogo: are you using lxc 3?
< michagogo> My host is xenial
< achow101> what do you get from lxc-start --version