00:23 <+bridge> [ddnet] I’ll cry 00:29 <+bridge> [ddnet] https://tenor.com/view/big-hero6-baymax-feel-better-hug-hugging-gif-4782499 00:30 <+bridge> [ddnet] @Learath2 00:30 <+bridge> [ddnet] Thanks 00:33 <+bridge> [ddnet] im here now 00:34 <+bridge> [ddnet] so what should i do build d7 server and test chat commands with latest vanilla client? 00:35 <+bridge> [ddnet] Yep 00:37 <+bridge> [ddnet] what am i looking for exactly? 00:39 <+bridge> [ddnet] https://cdn.discordapp.com/attachments/293493549758939136/676935389230923776/unknown.png 00:39 <+bridge> [ddnet] oh cmn 00:39 <+bridge> [ddnet] can we pls keep same binary names for master and d7 00:41 <+bridge> [ddnet] @Learath2 d7 is very unstable i can hardly get it runnign longer than 10 sec 00:41 <+bridge> [ddnet] https://paste.zillyhuhn.com/rg 00:43 <+bridge> [ddnet] did i mess something up? Because for me it only shows vanilla commands ._. 00:51 <+bridge> [ddnet] Its not d7 00:51 <+bridge> [ddnet] Its a new branch 00:51 <+bridge> [ddnet] GitHub - ddnet/ddnet7 at d7_pr_chatcom 00:51 <+bridge> [ddnet] 00:51 <+bridge> [ddnet] https://github.com/ddnet/ddnet7/tree/d7_pr_chatcom 00:55 <+bridge> [ddnet] @ChillerDragon ^^ 00:59 <+bridge> [ddnet] oh ah 01:00 <+bridge> [ddnet] but did you know that selfkill caused a crash on d7? 01:00 <+bridge> [ddnet] i thought d7 is the super clean super stable branch? 01:01 <+bridge> [ddnet] D7 is a constant work in progress 01:01 <+bridge> [ddnet] https://cdn.discordapp.com/attachments/293493549758939136/676940802353791007/unknown.png 01:01 <+bridge> [ddnet] men why cant branch switching be clean 01:01 <+bridge> [ddnet] ever 01:01 <+bridge> [ddnet] It currently requires a huge vanilla restructuring to get working, so I’m doing that 01:01 <+bridge> [ddnet] That submodule pissed me off so much too 01:01 <+bridge> [ddnet] ye 01:01 <+bridge> [ddnet] same here 01:02 <+bridge> [ddnet] sorry for that 01:02 <+bridge> [ddnet] :/ 01:03 <+bridge> [ddnet] The problem with kill in d7 is that the gameworld pointer is invalid at that point 😛 01:05 <+bridge> [ddnet] I’m rewriting gamecontroller in a way that we no longer need to inherit from the vanilla controller when writing our own controllers 01:08 <+bridge> [ddnet] looks good to me @Learath2 01:08 <+bridge> [ddnet] so what is this branch now? 01:08 <+bridge> [ddnet] a fork of d7? 01:08 <+bridge> [ddnet] looks more like master 01:08 <+bridge> [ddnet] No its a branch off of master 01:08 <+bridge> [ddnet] why is it called d7 then 01:08 <+bridge> [ddnet] clickbait 01:08 <+bridge> [ddnet] Wanted to give fokkonaut a hand with the new chat commands thing 01:09 <+bridge> [ddnet] i wasnt thinking while creating the branch 01:09 <+bridge> [ddnet] why /me and not /me ? is this some irc thing i do not get? 01:10 <+bridge> [ddnet] Hmm, its probably a bug in the code that reads the argument list 01:10 <+bridge> [ddnet] idk how the new system works but it looks fine 01:10 <+bridge> [ddnet] i guess most important is that binds like say /pause still work 01:10 <+bridge> [ddnet] seems functional to me 01:10 <+bridge> [ddnet] Oooh, didnt think about the binds 01:11 <+bridge> [ddnet] We should add a way to bind chat commands 01:11 <+bridge> [ddnet] it is working tho 01:11 <+bridge> [ddnet] so all fine i guess 01:11 <+bridge> [ddnet] Well its a janky way to keep them working 01:12 <+bridge> [ddnet] well so before you remove it 01:12 <+bridge> [ddnet] Master supports both old and new style of sending commands 01:12 <+bridge> [ddnet] keep in mind its a important thing 01:12 <+bridge> [ddnet] Would it be fine if you had to change binds? 01:12 <+bridge> [ddnet] its a good idea to support old clients anways 01:12 <+bridge> [ddnet] yikes @Learath2 01:12 <+bridge> [ddnet] like how? 01:13 <+bridge> [ddnet] yet another reason not to update to 0.7? 01:13 <+bridge> [ddnet] loosing all binds 01:13 <+bridge> [ddnet] bind q ccom pause 01:13 <+bridge> [ddnet] Oh I know a less janky way to keep old style chat commands 01:13 <+bridge> [ddnet] ❤️ 01:14 <+bridge> [ddnet] Strip the /, get the command name, find the command, go through OnCommand 05:56 <+bridge> [ddnet] yup 05:56 <+bridge> [ddnet] but binding is probably better 05:56 <+bridge> [ddnet] and redoing 05:56 <+bridge> [ddnet] binds 05:56 <+bridge> [ddnet] otherwise we will never detach from the chat commands... 09:27 <+bridge> [ddnet] @Jupstar ✪ https://github.com/ddnet/ddnet/issues/1997 can you take another look? 12:37 <+bridge> [ddnet] @deen nothing i can do 12:37 <+bridge> [ddnet] i see a huge difference between 2.0.10 and 2.0.8 from 12:37 <+bridge> [ddnet] https://libsdl.org/release/ 12:37 <+bridge> [ddnet] 12:37 <+bridge> [ddnet] so even if its not on software side, they seem to compile without optimazations or something 12:37 <+bridge> [ddnet] 12:37 <+bridge> [ddnet] also, if they gpu is their limiting factor they might not notice the difference so much(obviously) 12:39 <+bridge> [ddnet] their tmp release also has bad performance 12:47 <+bridge> [ddnet] @Jupstar ✪ no chance to build sdl2 with visual studio? then we'd get the debugging info 12:47 <+bridge> [ddnet] I see the same difference, but apparently they can't repro it 13:02 <+bridge> [ddnet] i did 13:02 <+bridge> [ddnet] but as i said sdl has dynamic api and vs cant debug it @deen 13:03 <+bridge> [ddnet] https://cdn.discordapp.com/attachments/293493549758939136/677122500475355139/unknown.png 13:03 <+bridge> [ddnet] dunno if u can disable that or smth 13:03 <+bridge> [ddnet] probs need a static sdl build 13:23 <+bridge> [ddnet] @deen see github 13:23 <+bridge> [ddnet] cant do more, bcs i am not really into sdl 13:37 <+bridge> [ddnet] if you have the problem too the guys in the mailing list lie for some reason 13:50 <+bridge> [ddnet] ok, thanks a lot. too bad there isn't anymore information 13:51 <+bridge> [ddnet] @deen see github 13:51 <+bridge> [ddnet] u can see the line numbers in the last screens 14:35 <+bridge> [ddnet] Is it only @Jupstar ✪ that has the issue? 15:00 <+bridge> [ddnet] @Learath2 well deen said he noticed it too on wine, but these guys from sdl are all like "YO I GET 700FPS, SO NO BUG" instead of actually checking a difference 15:23 <+bridge> [ddnet] http://lists.libsdl.org/pipermail/commits-libsdl.org/2018-August/002913.html 15:23 <+bridge> [ddnet] this was the commit probably, and it also feels wrong 15:23 <+bridge> [ddnet] it updates something every frame, instead of only if "it has been stolen" 15:23 <+bridge> [ddnet] but hopefully a real sdl coder sees it, else it never gets fixed xD 15:43 <+bridge> [ddnet] @Jupstar ✪ you could try with and without that commit and report the difference 15:49 <+bridge> [ddnet] i should, but don't want to switch to windows all the time, but if no SDL coder answers i'll do it probably 😄 16:17 <+bridge> [ddnet] Is it only @Jupstar ✪ that has the issue? 16:18 <+bridge> [ddnet] i see it in wine, 800 instead of 900 fps 18:27 <+bridge> [ddnet] hope we get some 0.7 servers this time o.O 18:27 <+bridge> [ddnet] no 18:27 <+bridge> [ddnet] ouch 18:27 <+bridge> [ddnet] they crash, so i can't update 18:27 <+bridge> [ddnet] and no players, so not motivated much 18:28 <+bridge> [ddnet] i dont get the thing about the crashes 18:28 <+bridge> [ddnet] could you tell me how to reproduce it? 18:28 <+bridge> [ddnet] nope 18:28 <+bridge> [ddnet] :/ 18:28 <+bridge> [ddnet] just seen it on all official servers 18:28 <+bridge> [ddnet] so i guess id have to import official db and runs some servers? 18:29 <+bridge> [ddnet] do you have a simple script that sets the servers up so i can try to reproduce? 18:29 <+bridge> [ddnet] nope 18:29 <+bridge> [ddnet] k 18:29 <+bridge> [ddnet] it seems to happen on map load 18:29 <+bridge> [ddnet] so probably trying to load every official map once should trigger it 18:29 <+bridge> [ddnet] https://github.com/ddnet/ddnet7/issues/62 18:29 <+bridge> [ddnet] this one? 18:29 <+bridge> [ddnet] could do that in a loop 18:30 <+bridge> [ddnet] yeah 18:30 <+bridge> [ddnet] yea sounds like a good idea 18:30 <+bridge> [ddnet] thanks i will try 18:30 <+bridge> [ddnet] whats the best way to get official maps? downloading the maps github repo i guess? 18:30 <+bridge> [ddnet] for i in *.map; do ./DDNet-Server "sv_map $i; quit"; done or sth like that 18:30 <+bridge> [ddnet] yy i know some basic bash 🙂 18:30 <+bridge> [ddnet] https://maps.ddnet.tw/compilations/maps7.zip 18:31 <+bridge> [ddnet] oh thats nice 18:31 <+bridge> [ddnet] thanks a lot 18:48 <+bridge> [ddnet] [2020-02-12 17:48:20][server]: failed to load map. mapname='blmapDT-UPTU' 18:48 <+bridge> [ddnet] but i guess that is not the issue since main servers do not have blockmaps huh? 18:49 <+bridge> [ddnet] but still it seems to be broken and make trouble on block servers 18:58 <+bridge> [ddnet] weird that your gdb output shows optimized variables 18:58 <+bridge> [ddnet] https://cdn.discordapp.com/attachments/293493549758939136/677211985691279361/unknown.png 18:58 <+bridge> [ddnet] shouldnt that not happen in debug builds? 19:38 <+bridge> [ddnet] Hmm i iterated over all maps 2 times and never got a non zero exit code. Also i iterated over Bullets 1000 times and even joined manually because it seemed to be crazy shotgun killing it. 19:38 <+bridge> [ddnet] :/ 20:00 <@deen> too bad 20:18 <+bridge> [ddnet] i don't have debug builds 20:18 <+bridge> [ddnet] ddnet uses relwithdebinfo 20:21 <+bridge> [ddnet] ah 21:20 <+bridge> [ddnet] @deen does it happen to you, when a tourn map os loaded it errors and you have to reconnect? It happened me all the time i has the 4 client stream, might be becausd of that? I stream on windows btw 21:21 <+bridge> [ddnet] is* had* 21:24 <+bridge> [ddnet] I think its cuz my 4 clients download at the same time 21:24 <+bridge> [ddnet] Idk 21:25 <+bridge> [ddnet] Maybe you dont get the error cuz u have the map beforehand if thats the case? 22:19 <+bridge> [ddnet] @Ryozuki that happens because the map download happens multiple times at once and they all write to the same file 22:19 <+bridge> [ddnet] multiple clients don't synchronize at all 22:20 <+bridge> [ddnet] i guessed that 22:20 <+bridge> [ddnet] do u have a workaround? 22:20 <+bridge> [ddnet] @Jupstar ✪ looks like they got a fix: https://bugzilla.libsdl.org/show_bug.cgi?id=4897 22:20 <+bridge> [ddnet] @Ryozuki like sending you the tournament map beforehand? 😄 22:20 <+bridge> [ddnet] :pepeH: 22:21 <+bridge> [ddnet] every process should write the map to map.$PID.tmp and then move it into the correct location atomically 22:21 <+bridge> [ddnet] at least that 's what i do with my shell/python scripts when I have this problem 22:21 <+bridge> [ddnet] then the map is always in a correct state, either it doesn't exist or it's finished 22:22 <+bridge> [ddnet] hmm 22:22 <+bridge> [ddnet] let me see if I can implement that 22:23 <+bridge> [ddnet] hey, the server does it that way already, see pid() in gamecontext.cpp 22:24 <+bridge> [ddnet] ill check 22:25 <+bridge> [ddnet] str_format(aTemp, sizeof(aTemp), "%s.temp.%d", pNewMapName, pid()); 22:25 <+bridge> [ddnet] ye 22:28 <+bridge> [ddnet] do you want to implement it? 22:30 <+bridge> [ddnet] If i do it will be tomorrow 22:31 <+bridge> [ddnet] hm, actually it's super simple, I'll just do it now 22:40 <+bridge> [ddnet] btw, is the cursor supposed to spin like crazy? 😄