01:04 < CookieMichal> deen: is it actually done? https://vgy.me/rcb7Vm.png ^^ context: map viewer in browser 01:06 < CookieMichal> https://vgy.me/7LXV60.png 09:57 < ddnet-commits> [ddnet] Learath2 opened pull request #700: Fix line width. Closes #687 (master...dd_pr_multiline) https://git.io/vyhWr 09:57 < ddnet-commits> [ddnet] def- pushed 2 new commits to master: https://git.io/vyhW6 09:57 < ddnet-commits> ddnet/master 2849e23 Learath2: Fix line width. Closes #687 09:57 < ddnet-commits> ddnet/master 70fec01 Dennis Felsing: Merge pull request #700 from Learath2/dd_pr_multiline... 10:53 < necropotame> deen, maybe a rebase next time? 10:57 <@deen> necropotame: what? 11:00 < necropotame> You did a merge in github, instead of a rebase+merge. We discuss this before with others, an I though that we agree that rebase+merge could be better. 11:00 <@deen> I don't see what's wrong with a merge 11:00 <@deen> rebase rewrites history, bad idea 11:01 <@deen> some people seem to have a fetish about keeping history linear, but i see no need 11:02 < necropotame> Since we don't use branches, it's not a problem to rewrite the history IMO. 11:03 < necropotame> In addition, ddnet.tw use git commits as changelog, so it should clean 11:05 <@deen> in your rebase you can (accidentally or intentionally) manipulate the change, and maybe no one would notice 11:05 <@deen> with a regular merge you can ensure that the original commit came in and then see what the merge diffed from that 11:06 <@deen> I'd prefer regular merges. does github even offer rebase+merge? 11:11 < necropotame> deen, yes, there is a small arrow next to the merge button where you can chose more options 11:14 <@deen> no, our git commits should not be clean, that's a totally wrong worry 11:14 <@deen> don't artificially make the git history clean 11:21 <@deen> anyway, not really an important thing to argue about, do whatever you want 11:25 < Learath2> We didn't discuss this properly with everyone around so don't change behaviour before we reach to a consensus 11:26 < Learath2> and I no longer think rebasing is a good idea either. 11:28 < necropotame> ok :) 11:53 < Learath2> we should start writing changelogs again 11:54 < Learath2> im willing to do them but need to coordinate it with releases 12:11 < ddnet-commits> [ddnet] Learath2 opened pull request #701: Properly calculate the width. Fixes #699 (master...dd_pr_hudsettings) https://git.io/vyhPR 12:14 <@heinrich5991> how does git-bisect interact with potentially unbuildable commits merged from a branch? 12:47 < ddnet-commits> [ddnet] def- pushed 2 new commits to master: https://git.io/vyhyC 12:47 < ddnet-commits> ddnet/master 55c5c3e Learath2: Properly calculate the width. Fixes #699 12:47 < ddnet-commits> ddnet/master b42ae74 Dennis Felsing: Merge pull request #701 from Learath2/dd_pr_hudsettings... 12:50 < Learath2> deen: do you happen to know which hook in a client component signals a disconnect? 12:51 <@deen> OnStateChange? 12:52 <@deen> With new state being IClient::STATE_OFFLINE 12:54 < Learath2> OnReset seems to work but idk when it gets triggered 12:55 <@heinrich5991> see gameclient.cpp 12:55 < Learath2> thank you \o/ i should work on my grep fu 12:56 < Learath2> yeah OnReset seems appropriate as its in OnStateChange 12:56 < Learath2> in case newstate is smaller then online 12:56 <@heinrich5991> can I suggest you to use any grep-like tool that ignores .gitignored files and is automatically recursive? 12:57 < Learath2> do you know such a program? :P 12:57 <@heinrich5991> I don't know what is available for your distribution. rg :) 12:58 <@deen> Really difficult to grep for something with a name like rg :D 12:58 <@heinrich5991> https://packages.gentoo.org/packages/sys-apps/ripgrep 12:58 <@heinrich5991> Learath2: ^ 12:58 < Learath2> btw would you be okay if i commit a gitignore to ignore ctags files? 12:59 <@heinrich5991> you can also .git/info/exclude them 12:59 <@heinrich5991> man gitignore 12:59 < Learath2> yeah i know thats why i asked 13:00 <@heinrich5991> looking into the gitignore, it's probably ok 13:00 <@heinrich5991> there's a lot of stuff already 13:00 <@heinrich5991> build server.bat 13:00 <@heinrich5991> [...] 13:00 <@heinrich5991> *.bat 13:06 < ddnet-commits> [ddnet] Learath2 opened pull request #702: Reset the username input on server/map change. (master...dd_pr_rconlogin) https://git.io/vyhHI 13:19 < ddnet-commits> [ddnet] Learath2 opened pull request #703: Reset input when server/map changes. Fixes #633 (master...dd_pr_rconhide) https://git.io/vyhQS 13:25 < Learath2> EastByte: not allowing dnsbl marked clients to vote seems like a good approach 13:28 < Learath2> SORBS seems to have a list specifically for proxies, we could use that and gather some statistics on false flagging for a while 13:30 < Learath2> even if we false flag its not the end of the world, they just cannot vote 13:31 <@heinrich5991> still undesirable 13:31 < Learath2> more undesirable then people getting votebotted out at the end of a map? 13:37 <@heinrich5991> that's also not nice 13:40 < Learath2> i'd say its comparatively worse 13:40 < Learath2> ofc not if the false flag rate is ridiculous but 1% we can sacrifice I guess 13:45 < Learath2> anyways as always just a suggestion :) 14:22 < CookieMichal> hey, ddnet requires python 2.x or 3.x? :s 14:30 <@deen> CookieMichal: python2 i believe 14:32 < CookieMichal> I have trouble building ddnet in visual studio again :/ so i thought it might be wrong python version 14:32 < CookieMichal> as i had such error once 14:34 <@heinrich5991> CookieMichal: how do you build it? 14:34 <@heinrich5991> bam or cmake? 14:34 < CookieMichal> cmake 14:34 <@heinrich5991> can you show an error you encountered? 14:35 < CookieMichal> right now it is crying about missing, but existing files C:\Users\MrCoo\Source\Repos\ddnet\src\engine\external\pnglite\pnglite.c(8): fatal error C1083: Cannot open include file: 'zlib.h': No such file or directory C:\Users\MrCoo\Source\Repos\ddnet\src\engine\shared\datafile.cpp(7): fatal error C1083: Cannot open include file: 'zlib.h': No such file or directory 14:36 <@heinrich5991> how did you download the source code? via git? if so, could you update to the latest version? 14:36 < CookieMichal> will do, sure 14:45 < CookieMichal> changed python to 3.6.1 from 2.x, cloned ddnet/ddnet again and tried building, but it still fails with same errors 14:45 < CookieMichal> but 14:45 < CookieMichal> i think i found the reason 1> -- Found ZLIB: D:/Program Files/PostgreSQL/9.6/lib/zlib.lib (found version "1.2.8") 14:46 < CookieMichal> https://gist.github.com/3b13ef8575d0cd67cba0dd059d713d28 14:47 < CookieMichal> whole build log: https://gist.github.com/42b834702b7d4e2a9f7083f564bb73ed 14:48 < CookieMichal> and shorter version: https://gist.github.com/a5ae25dab982365fe44a3b9e2c602c63 15:18 <@heinrich5991> CookieMichal: try replacing set(ZLIB_INCLUDEDIR) with set(ZLIB_INCLUDEDIR ${ZLIB_INCLUDE_DIR}) in CMakeLists.txt 15:18 <@heinrich5991> around line 152 15:18 < CookieMichal> guess what 15:18 <@heinrich5991> ? 15:18 < CookieMichal> I uninstalled postgres and i didn't really have need for that and it builds 15:19 <@heinrich5991> meh 15:19 < CookieMichal> shrug 15:19 <@heinrich5991> now we can't fix this bug ^^ 15:19 < CookieMichal> > installing 15:19 < CookieMichal> :DD 15:19 <@heinrich5991> :D 15:19 <@heinrich5991> will have spotty internet in the next hour 15:20 < CookieMichal> what do you mean by spotty? 15:31 < CookieMichal> heinrich5991: can u resend the message with what to change? I lost connection and it's not in logs yet 15:33 < necropotame> CookieMichal -> http://pastebin.com/ZVKRqic8 15:33 < CookieMichal> thanks 15:42 < CookieMichal> heinrich5991: if that is the line you meant it still failes to build https://vgy.me/1KioNl.png 15:43 < CookieMichal> oh wait 15:43 < CookieMichal> I forgot one underscore 16:10 < CookieMichal> fixed missing underscore, build logs: https://gist.github.com/e60ca662073d41ef1136c1b96f5352f1 :( 16:11 < Ryozuki> warning LNK4272: library machine type 'x64' conflicts with target machine type 'X86' 16:12 < Ryozuki> mb u should use the bundled lib 16:23 < Henningstone> necropotame: is the "toreproduce" label supposed to mean "reproducable" or "has to be reproduced yet"? I guess the first one is the case, judging from the issues it's assigned to. Either way, it sounds ugly. 16:26 < Learath2> Henningstone: maybe need-info? 16:26 < Henningstone> like "has to be reproduced yet"? :) 16:26 < Learath2> yeah its supposed to mean it needs to be reproduced 16:26 < necropotame> second one. What I wanted to do is to create two task list: one for issues that doesn't need more discussion to be fixed (todo label), and one for bugs the need to be reproduced to be valided (toreproduce) 16:27 < necropotame> I don't have any problem to change the name. I'm not good enough in english or grammar to have a decent opinion about this. 16:28 < Learath2> this is a hail-mary but is there anyone here who knows Windows IT? 16:29 < Henningstone> necropotame: I was just unsure about the its meaning, and apparently I actually got it wrong. But since it means the latter, I don't understand why it's assigned to issues that clearly are and have been reproduced or are simply a fact 16:30 < Henningstone> *about its meaning / *clearly are reproducable, for god sake 16:31 < necropotame> I don't know about which one you are referring. But as long are there is no answer in the issue, one can at least agree that nobody try to give us a confirmation about it (we are talking about REproducing a bug) 16:33 < necropotame> It's also a way to give a feedback for the author of the issue: your issue is readable, we are not dead, but we are waiting for a confirmation or a fix 16:33 < Henningstone> Ah I see your point, you need a second person confirming it, not only a way to reproduce it... 16:35 < necropotame> Yes. In case we don't have a description of how to reproduce it, we have to write a message to ask for it 16:35 < necropotame> It's my personal way to manage issues. Feel free to comment it :) 16:40 * Henningstone just made use of your way 16:40 < Henningstone> hopefully I provided enough information ;) 16:41 < Learath2> Henningstone: any idea why we don't use DoubleClicked on non android? 16:41 < Henningstone> maybe just noone thought about it when android support was implemented... https://github.com/ddnet/ddnet/blame/b42ae74d29d04e650ecc4651202d3cf0cd24456b/src/game/client/components/menus_browser.cpp#L471 16:42 < Learath2> i'm testing it now 16:43 < Learath2> oh doubleclicked doesn't do what you think it does 16:43 < Learath2> with the first click you mark something, if you mark something the second click counts as the double click 16:47 < Henningstone> Learath2: Ahh ok, so that's why it's android only :D Does combining both ways sound like an idea? 16:47 < Henningstone> because you can't double click an item without having selected it (since the first implicates the latter) 16:47 < Learath2> it sounds really suitable for touch devices and really weird for devices with a mouse 16:48 < Henningstone> I was talking about »if(Input()->MouseDoubleClick() && DoubleClicked)« in line 471 16:52 < Learath2> definately fixes the bug but i have no idea how it works :D 16:52 < Learath2> oh and it changes behaviour for android 16:52 <@deen> android doesn'T even build 16:52 < Henningstone> Learath2: how can that be, you shouldn't have needed to touch any android code, would you? 16:52 <@deen> and sdl2 should have android builtin 16:53 < Learath2> the doubleclicked variable is only defined in case of android 16:53 < Learath2> anyways i'll commit and push it so you can see what i did 16:54 < Henningstone> okay, thanks :) 16:57 < ddnet-commits> [ddnet] Learath2 opened pull request #704: Handle double click on server browser correctly. Fixes #597 (master...dd_pr_browserdbl) https://git.io/vyjaZ 16:57 < Learath2> Henningstone: could you give it a go, i don't have a mouse and its really hard to test it 16:58 < Henningstone> I will. What you did there is exactly what I imagined, and it really doesn't affect android 16:58 < Learath2> yeah i thought you wanted to remove the ifdef for some reason 17:00 < Learath2> windows is so so so so so inconsistant... 17:01 < Henningstone> within itself or in respect to other? 17:01 < Learath2> within itself 17:01 < Learath2> i've been trying to get a fully updated windows 7 image and its been a pain 17:03 < Learath2> even their documentation seems wrong, I update it fully, I image it exactly as they suggest, I deploy it exactly as they suggest, It hangs on setup... 17:06 < Learath2> Henningstone: how do you review with a comment btw? 17:08 < Henningstone> I clicked the 'Review changes' button in the 'Files changed' tab 17:08 < Learath2> oh yeah just saw that :P 17:24 * Henningstone killed the second 'toreproduce' 17:41 < necropotame> Henningstone, why don't you create a PR if you did a commit for ATH? It's not compatible? 17:42 < necropotame> (but thanks anyway!) 17:42 < Henningstone> what I committed is not directly compatible without modifying, and I'm too lazy to setup ddnet fork 17:42 < necropotame> ok ^^ 18:50 < Learath2> Henningstone: where is the rounding? The precision is just dropped which should result in "x1.0 , x1.0 , x2.0" no? 18:51 < Henningstone> %.0f 18:51 < Henningstone> idk 18:51 < Henningstone> maybe apparently it rounds 18:54 < Henningstone> if you switch on cl_demo_show_speed, it shows that the actual value of the variable is correct 18:55 < Learath2> Henningstone: so its set to 1.5 but it displays 2.0? 18:55 < Henningstone> Learath2: that is what we see happen, although as far as I know it should indeed drop the precision as you said 18:58 < Learath2> hmm it drops precision for me Henningstone 18:58 < Learath2> oh wait no it doesn't 18:58 < Learath2> thats weird 18:58 < Henningstone> jap just tried it once again, and it's definitely going x1, x2, x2, x4 19:01 < Henningstone> To test how %.0f behaves, I have thrown some test values at it, and it's definitely doing mathematically correct rounding. How cool. 19:02 < Learath2> yeah its rounding wtf 19:04 < Learath2> and comparing floats is not fun :/ 19:18 < ddnet-commits> [ddnet] Learath2 opened pull request #705: Fix a couple of problems with the demoviewer. (master...dd_pr_demorate) https://git.io/vSeez 19:18 < Learath2> thx Henningstone ^^ 19:18 < Learath2> you are okay with me commiting these right? 19:19 < Henningstone> Learath2: sure 19:21 < Henningstone> Learath2: some testing, research and asking people told me that rounding is desired, but depends on which compiler you ask. http://www.exploringbinary.com/inconsistent-rounding-of-printed-floating-point-numbers/ And this is what I get: https://hastebin.com/usigupigez.go 19:23 < Learath2> well .0f demonstrates once and for all rounding is unintuitive and stupid 20:03 < Ryozuki> deen, maybe you can check this https://github.com/ddnet/ddnet/issues/423