=== Guest94094 is now known as RAOF === chihchun_afk is now known as chihchun === JanC is now known as Guest10980 === JanC_ is now known as JanC === JanC is now known as Guest18582 === JanC_ is now known as JanC [09:18] alan_g_ === alan_g_ is now known as alan_g [11:11] alan_g: hey [11:11] ? [11:11] so I'm once again trying to bring Mir up on Fedora [11:11] interestingly, the biggest problem now is that it can't find capnproto [11:12] apparently, there's a Debian-specific change that you guys are depending on (Debian added CMake things to the autotools build, but different ones from what upstream Provides) [11:12] the Fedora package only has pkgconfig stuff [11:12] I also found out I can't build patched libinput and mesa for Fedora 26, since the patches don't work on libinput 1.7.0 and Mesa 17 [11:13] Son_Goku: that's interesting, but I'm not really set up to investigate that [11:13] I figured, but since you were interested a few weeks ago, I figured I might as well tell you now [11:14] Son_Goku: I thought about changing mir to not require those additional symbols... should be simple change.. I think about proposing that this weekend.. [11:15] anpok_: that'd be great [11:15] Son_Goku: yes, I'd like to know and can help if (for example) you can put together a PR [11:15] alan_g: I'm already investigating how to fix gmock, since it can't find it on Fedora [11:16] Son_Goku: I just put up an MP to use cmake-extras for finding it [11:16] also I have seen Marius is working on a wayland backend platform driver.. wl client support will be way more work [11:16] (dunno if that works on fedora) [11:16] that would be very cool [11:16] but I guess more important than the other changes [11:17] alan_g: cmake-extras, as in CMake's own FindGTest.cmake? [11:17] because if so, then yes, that helps a lot [11:17] Well FindGMock [11:17] right [11:17] we offer the compiled library, since Google Mock guys lifted that silly requirement to only do sources quite a while back [11:18] at least two years ago, I think [11:18] Yeah, I just noticed a couple of days ago [11:19] it took a while to convince juliank the same thing, and then we fixed it for apt [11:19] so apt now builds on Fedora too :P [11:20] apparently, the only platform where the "only sources" thing matters is Windows :) [11:20] because compilers have vastly different ABIs there [11:21] ah, I see Android support is toast now [11:21] that will simplify things in my spec file [11:21] Yeah, even between compiler options on Windows [11:21] yeah... [11:22] * Son_Goku shudders [11:22] I'm part of a game development project, so I'm keenly aware of how broken the Windows platform is [11:22] people there have Stockholm Syndrome about the Windows development environment :/ [11:22] they don't realize it can be better [11:22] *I* made money out of Windows for years, but it was horrid to work on [11:23] yeah [11:23] the only reason people endure it is because the Linux Desktop hasn't taken off yet :( [11:23] ONE DAY!!! :D [11:23] * alan_g is hoping never to go back [11:23] * Son_Goku is hoping the same [11:24] the job I had before my current one, I had to develop for embedded systems from Windows [11:24] that's a whole new level of pain [11:24] there was no technical reason I couldn't do it from Linux, I was just forbidden to because "we already paid for the Windows licenses, so we should use them" (their words) [11:25] in a few weeks I will probably work for a linux based product.. but I fear the development systems will all run windows.. [11:25] * alan_g shuts out the bad memories [11:25] anpok_: you're leaving Canonical? [11:25] yes [11:25] anpok_: at least on Windows 10 you'll have "Bash for Windows" [11:25] alan_g: it's a very poor substitute [11:26] most things are subtly broken in horrifying ways [11:26] Better than cygwin [11:26] ironically, no [11:26] Really? [11:26] while performance is certainly better, Cygwin preserves Linux primitives better [11:26] Windows things bleed into LXSS and cause odd issues [11:27] But it could never do fork() right [11:27] well, technically, now Cygwin could do fork() correctly [11:27] since there's a brand new syscall to do it [11:28] Thanks for the warning [11:40] alan_g: one fun issue: the text encoding of paths was messed up :( [11:40] I went NUTS trying to figure out what was wrong in an app built and running in LXSS, and when it came to that, I was rather upset [11:41] I remember having problems with paths in cygwin - it worked except when it didn't [11:41] path literals were always screwy in cygwin in the old days [11:41] but latest versions don't seem to have issues === chihchun is now known as chihchun_afk === alan_g is now known as alan_g|afk === alan_g|afk is now known as alan_g