--- Log opened Tue Jul 23 00:00:18 2019 01:11 -!- jb55 [~jb55@gateway/tor-sasl/jb55] has quit [Ping timeout: 260 seconds] 01:21 -!- jb55 [~jb55@gateway/tor-sasl/jb55] has joined #bitcoin-builds 04:28 -!- hebasto [~hebasto@95.164.65.194] has joined #bitcoin-builds 04:46 -!- jonatack [8613b3a3@134.19.179.163] has joined #bitcoin-builds 05:59 -!- jonatack [8613b3a3@134.19.179.163] has quit [Remote host closed the connection] 06:07 -!- jb55 [~jb55@gateway/tor-sasl/jb55] has quit [Ping timeout: 260 seconds] 06:14 -!- jonatack [d598a259@213.152.162.89] has joined #bitcoin-builds 10:17 -!- jb55 [~jb55@gateway/tor-sasl/jb55] has joined #bitcoin-builds 10:55 < dongcarl> cfields: Just did a review of the configure flag changes in #16370, I have all the `configure --help` output here if you want: https://gist.github.com/dongcarl/892d7deda8da3ff72317009ecbb07ffb 10:55 < dongcarl> cfields: Would love to know your thoughts here: https://github.com/bitcoin/bitcoin/pull/16370#issuecomment-514315845 11:18 < cfields> dongcarl: commented (and edited) 11:18 < cfields> dongcarl: That's cool and useful output, btw :) 11:19 < fanquake> 🚀 11:19 < fanquake> Cheers for some review guys. 11:19 < cfields> fanquake: Great initiative. 11:20 < dongcarl> cfields: Haha yeah... I actually purchased the "configure.help" domain and was gunna put up a site where you can see the configure output of the latest depends tree... 11:20 < dongcarl> One of these days... 11:20 < cfields> dongcarl: lol! 11:20 < cfields> fanquake: you might want to think about merging the depends PRs in a clump though, to minimize ci disruption. 11:21 < cfields> dongcarl: maybe a makefile target in depends would help? "make confighelp" would just recurse through all packages and output their helps. 11:22 < cfields> dongcarl: I'm not convinced that's really worth doing, though. Qt, for example, has a shitload of hidden/undocumented config switches. 11:23 < fanquake> cfields: yep sure, will merge a bunch together when possible. 11:23 < dongcarl> cfields: Yeah, we'll see. One day I'll get annoyed enough to build out configure.help 11:24 < cfields> haha 11:26 < cfields> dongcarl: before I forget, pretty sure I remember seeing "vnc" as part of our qt enabled features. 11:26 < cfields> If so, that can and definitely should be removed. 11:26 < dongcarl> wtf 11:26 < fanquake> 👀 11:27 < cfields> IIRC it's just a possible output module, likely unreachable, but still :) 11:27 < dongcarl> cfields: It's one of those hidden flags for qt... 11:27 < fanquake> I’ve been hanging out for someone to comment on me adding the -no-ico feature flag 11:27 < dongcarl> fanquake: hahaha I saw that 11:28 < cfields> fanquake: lol 11:28 < fanquake> hah. I can throw -no-vnc, or whatever the flag is into the same PR if you want. 11:29 < dongcarl> cfields: Is it possible to get a list of hidden flags? 11:29 < cfields> fanquake: please do. Let's hope that's all it takes. 11:29 < cfields> dongcarl: sure, if you pay. 11:29 < dongcarl> `-list-features`? 11:29 < cfields> dongcarl: sadly that's not a joke :( 11:29 < dongcarl> wait what 11:30 < fanquake> also -list-libraries 11:30 < dongcarl> the commercial version has better tooling? 11:31 < cfields> dongcarl: There's a gui tool that lets you specify what you want, scans and creates the necessary config, and spits out a .yaml or .json or something. 11:31 < fanquake> I wonder if that Qt Configurator program contains every option 🧐 11:31 < cfields> Yup. 11:31 < cfields> https://doc.qt.io/QtForDeviceCreation/qt-configuration-tool.html 11:32 < cfields> Configuring without that tool is more obfuscated now as a result. It's a bit hard to assume pure intentions there. 11:34 < dongcarl> :-/ 11:35 < fanquake> Is a commercial license worth some rude amount of $? 11:36 < cfields> fanquake: I assume :\ 11:38 < dongcarl> fanquake: all features and libs: https://gist.github.com/dongcarl/f15946dff501dea85977b2c2fb9f00a0 11:39 < fanquake> It’s also incredibly unintuitive what you can and can’t turn off from that list. 11:40 < fanquake> Can turn of one thing, say like a font picker box, that we definitely don’t use, and it turns out that it’s actually needed by windows, or the test module or something. 11:40 < cfields> fanquake: right... 11:41 < cfields> fanquake: at one point I did a "disable everything and turn things back on until I have a functional build" run, but it was so broken at every step that it wasn't useful. 11:41 < cfields> I understand it's supposed to be more tolerant to weird build configs though, so maybe it's worth trying that again. 11:41 < fanquake> Ah yea. That was a bit of the approach in #16386. 11:42 < cfields> Ah, great, haven't gotten to that one yet. 11:43 < cfields> Mmm, IIRC OSX requires sessionmanager. 11:44 < fanquake> hmm ok. I thought it was going to be used on Windows if anywhere. Can have a look at it again. 11:44 < cfields> Maybe it's windows? I definitely remember making that platform-specific on purpose, though. 11:44 < cfields> I remember that being a runtime issue, though. So we just need to make sure that Win/Osx/Linux all still actually run ok after. 11:45 < fanquake> Ok. Will add a note to re-check that 👍 11:45 < cfields> It could just be that things are less tangled now and it's really no longer necessary. 11:58 < cfields> fanquake: yep, you're right, looks like 16386 is exactly what I meant. Thanks for doing that :) 12:02 < fanquake> cfields: no worries 18:03 < fanquake> cfields: had a quick look. You are correct in that we are building the VNC QPA backend. --- Log closed Wed Jul 24 00:00:19 2019