14:33 < ddnet-commits> [ddnet] heinrich5991 pushed 2 new commits to master: https://git.io/vg5kJ 14:33 < ddnet-commits> ddnet/master e183a17 Tim Schumacher: removed other gametypes files and gametype setting 14:33 < ddnet-commits> ddnet/master c1df605 heinrich5991: Merge pull request #422 from timakro/master... 14:55 < o_be_one> hello 14:55 < o_be_one> i've to ping again deen EastByte 14:56 <@deen> yes? 14:56 <@deen> hi o_be_one 14:56 < o_be_one> hello :) 14:57 < o_be_one> well deen as ive said some days ago, i've to move can.ddnet 14:57 <@deen> you only pinged EastByte last time, so I conveniently ignored it :P 14:58 <@deen> I guess I can move CAN to the new server if you think we should keep it running 14:58 < o_be_one> ahah yes, yesterday, but not 5 days ago :p no problem ^^ 14:58 <@deen> it's the least popular DDNet server I think 14:58 < o_be_one> there's not mutch players, US works well, but if you think it's useful for DDNet (like telling that we have mutch servers) i've no problem to give you a new can host 14:59 <@deen> sure, it doesn't hurt. the only thing I'm worried about is the cost for you 15:00 < o_be_one> yes i understand, as you like 15:01 <@deen> if you send me the info about the new server, I'll set it up 15:14 < o_be_one> done :). Ok more news about ger2 ? 15:14 < o_be_one> since ive changed arbor settings 15:17 <@deen> the only thing I notice is that ssh still takes 3 min 15:18 <@deen> didn't check anything else, you should ask EastByte 15:19 <@heinrich5991> 3min to connect? 15:20 <@deen> yes 15:20 <@heinrich5991> if it's the tcp that is behaving poorly, you could try using mosh (uses UDP, except for initial session establishment) 15:20 <@deen> and rewrite all other programs to use udp as well?^^ 15:20 <@heinrich5991> no ^^ 15:21 <@heinrich5991> just to fight that particular problem :) 15:21 <@deen> it's just a symptom I'd say 15:32 < fstd> deen: i've had similar problems in the past (although not as bad as 3 min) and it turned out to be a problem of name resolving on the remote site. might wanna check /etc/nsswitch.conf whether it attempts any methods that are not really available 15:33 <@deen> fstd: interesting. I believe it's the firewall doing funny stuff, also dropping my tcp connections after a few hours 15:34 <@deen> nsswitch.conf looks totally normal, same as on other servers 15:34 <@deen> but hey, name resolution doesn't work^^ 15:35 < fstd> heh 15:35 <@deen> wait no, it works. I thought you have to restart some service to reread nsswitch.conf 15:35 <@deen> I just disabled dns lookup to test 15:35 < o_be_one> fstd: you right, it's often an issue with name resolve 15:36 < fstd> nsswitch.conf is consulted directly by glibc i think 15:36 < o_be_one> but i get dropped on my own vps on the same host after a while 15:36 < o_be_one> hello fstd btw, do you have some news from XXLTomate ? 15:36 < o_be_one> maybe i'll mail him 15:37 <@deen> fstd: strace confirms :) 15:37 < fstd> :) 15:37 < fstd> o_be_one: no, haven't talked to him in a while 15:38 <@heinrich5991> why do you ask? :) 15:53 < o_be_one> heinrich5991: your question is for me ? 15:53 <@heinrich5991> yes 15:53 < o_be_one> cause XXLTomate is a friend, and he helped me a lot when i've started with blockers :) 15:53 < o_be_one> long time i didnt get news from him 16:09 < Paszczak> hey guys I got a problem. When I compile the lastest ddnet on windows 7 64 bit, this shows 16:09 < Paszczak> LINK : fatal error LNK1181: cannot open input file 'libopusfile-0.lib' bam.exe: 'link DDNet.exe' error 1181 bam.exe: error: a build step failed 16:10 <@deen> you're using the bundled opus libs I guess? 16:10 < Paszczak> yes 16:10 <@deen> I see that the libopusfile-0.lib is missing for windows x86_64, oops 16:10 <@deen> I thought that wasn't necessary, hmm 16:11 <@deen> Or I didn'T know how to get that file because I built it myself? 16:11 <@deen> with win32 it should work 16:11 < Paszczak> yea it works 16:12 <@deen> Ah, you're building with Visual studio? 16:12 < Paszczak> you mean do I use cl or mingw 16:12 <@deen> yes 16:12 < Paszczak> I use cl 16:12 <@deen> ok, that's the reason I don't see that error 16:12 <@deen> Do you know how to get a .lib file? 16:12 < Paszczak> on 16:13 < Paszczak> no* 16:14 <@deen> someone with the same problem: http://www.enchantedage.com/opusfile-0.6-for-visual-studio 16:15 < Paszczak> visual-studio-compatile-opusfile-0.6-win32.zip the name of the attachement suggests thats not what we are looking for 16:15 < o_be_one> hey Paszczak whatsup ! 16:15 <@deen> yeah right 16:15 < Paszczak> hey dude, whats up 16:16 < o_be_one> not mutch, it seems that r0x.TW is dying, i'm waiting for a fix from EastByte and have to change server bin to latest DDNet 16:16 < Paszczak> deen, I can stay on win32 but just asked if there's any fast fix or if you even know about it 16:17 < Paszczak> obeone what about baam? 16:17 <@deen> Paszczak: I'm just checking my logs, I built that 64bit opus myself, so I think no .lib file was created. My guess is that someone would have to compile these with cl and then get the .lib file 16:18 < o_be_one> Someone just doesnt care, he stole the map and hosted it. Maybe he didn't know that this map was made for r0x.TW ... originally, from Bip. 16:18 < Paszczak> deen: okay, i'll stay on 32 bit version 16:18 < o_be_one> but when i'll have the time and motivation to get back r0x.TW to the top, i'll do it, actually i'm just a little too busy ^^ 16:19 < Paszczak> o_be_one: yea that sucks,i don't know why it's more popular than original baam 16:19 < o_be_one> oh, it's just cause i have some bugs on the ddr0x bin i think so they exploit it to drop players 16:20 < o_be_one> no mutch people is really fair on TW x) they dont care, they just want to think that they are so pro and popular with their servers 16:20 <@deen> Paszczak: I made an issue in case someone wants to fix it: https://github.com/ddnet/ddnet/issues/423 16:21 < Paszczak> deen: nice, thanks 16:24 < laxadedi> deen: all servers are on debian ? 16:24 <@deen> laxadedi: no, a few on CentOS 16:24 < laxadedi> 0day pretty darn violent got releaed, you should update asap 16:24 < laxadedi> https://googleonlinesecurity.blogspot.fr/2016/02/cve-2015-7547-glibc-getaddrinfo-stack.html 16:42 <@deen> ok, thanks 16:42 <@deen> reading that reminds me that I removed a few coredumps from ssh client randomly crashing a few days ago. Maybe I should look closer at crashes, oops 16:53 <@deen> CAN running on new server 16:59 < o_be_one> nice :) 16:59 < o_be_one> so i can kill the old one ? 17:00 <@deen> yes 17:00 < o_be_one> ty 17:02 <@deen> nah, thanks to you for offering the server 17:05 < ddnet-commits> [ddnet] timakro opened pull request #424: made fading border in menu more customizeable (master...master) https://git.io/vg5iv 17:15 <@deen> Installing GTK on Windows is becoming weirder and weirder with time 17:16 <@deen> To install GTK+3 on Windows you need msys2 which contains pacman, the package manager from Archlinux 17:29 <@heinrich5991> gtk3's windows support is lacking 17:29 <@heinrich5991> should probably use something different if you want to target windows 17:31 <@heinrich5991> on another note: I like msys2 a lot saner than cygwin 17:31 <@heinrich5991> s/saner/better 17:31 <@deen> heinrich5991: so, what to use? 17:32 <@heinrich5991> qt is good, if you use python or c++ 17:32 <@deen> I'm trying to get ghc-vis updated again. Back then I saw no good alternative 17:32 <@deen> Well, it's Haskell 17:33 <@heinrich5991> also used wxwidgets for a small application (from python) 17:33 <@deen> I tried in SDL2 but then you lose all the GUI support and just have a raw window 17:33 <@deen> otherwise html/css/js :P 17:38 <@heinrich5991> deen: https://wiki.haskell.org/Applications_and_libraries/GUI_libraries#HsQML 17:39 <@heinrich5991> also wxWidgets on that page 17:39 <@heinrich5991> however, I have never done any UI in haskell, so I can't comment on how usable they are 17:40 <@deen> too late anyway, don't want to work on that anymore. I'll just submit all the fixes I can find for Windows and OSX gtk and be done with it^^ 17:41 <@heinrich5991> yea, porting an existing application isn't worth it probably 17:41 <@deen> I just hope never to write a GUI thingy again 17:43 <@deen> Also, damn compile-times in Haskell! 17:52 <@heinrich5991> go has fast compile times :P 17:52 <@heinrich5991> (i.e. that's what you get if you wish for fast compile times ^^) 18:04 <@deen> Nim has faster compile times than go 18:04 <@deen> :P 18:05 <@deen> so you get some nice compile-time stuff (not rust or haskell level) and still have fast compilation 18:06 <@heinrich5991> nim is faster than go? wow :) 18:06 <@heinrich5991> (in compile times I mean) 18:07 <@deen> doing hindley-milner type inference or borrow checking is probably not so fast 18:07 <@deen> well, depends on what you compile of course, I only ported one Go application to Nim, so my sample size is 1: http://hookrace.net/blog/porting-nes-go-nim/#comparison-of-go-and-nim 18:07 <@heinrich5991> well, C typically has slower compile times than go 18:07 <@heinrich5991> because of the whole header stuff 18:07 <@heinrich5991> afk