00:39 < bridge> [teeworlds] What is a `CNetBan::CNetHash::MakeHashArray`? 02:34 < bridge> [teeworlds] c++20 gonna be liiiiit 08:39 < bridge> [teeworlds] Lit like a dumpster fire 10:02 < bridge> [teeworlds] Don't be like that, c++20 is bringing some great features 10:04 < bridge> [teeworlds] It is, yeah 11:00 < bridge> [teeworlds] header files, fuck header files 😄 12:11 < bridge> [teeworlds] Modules, concepts, std::format are more than enough for me 12:11 < bridge> [teeworlds] ❤️ 17:09 < bridge> [teeworlds] is tw cpu or gpu "bottlenecked"? 17:11 < bridge> [teeworlds] okay after quickly testing it's cpu bottlenecked for a single core 17:14 < bridge> [teeworlds] try `gfx_finish 0` 17:15 < bridge> [teeworlds] but at least if you have a dedicated gpu it's really limited by the cpu 17:15 < bridge> [teeworlds] mostly because it uploads all vertex data to the gpu each frame 17:17 < bridge> [teeworlds] by storing the text and map data on the gpu you can easily get ~90% gpu utilization while keeping the cpu load kinda low 17:20 < bridge> [teeworlds] what does this command do? I have not quite but close to doubled my fps with it 17:20 < bridge> [teeworlds] it's still "just" 70-80% gpu utilization but much closer at least 17:42 < bridge> [teeworlds] `gfx_finish 1` makes sure that the gpu finished rendering the last frame before submitting commands for the new frame 17:42 < bridge> [teeworlds] in certain scenarios this can reduce input latency, but most of the time it just wastes gpu performance, since the gpu is idling while the cpu is preparing the new frame 17:42 < bridge> [teeworlds] in tw it usually does not matter. for me the game runs with >1000 fps anyway 😄 18:49 < bridge> [teeworlds] @redix I think you kinda broke net_addr_comp, it no longer returns the correct order 18:55 < Oy> what order? 18:56 < bridge> [teeworlds] Atleast the comment in system.h needs to be updated it never returns >0 anymore 18:57 < bridge> [teeworlds] If something like a sorted array of NETADDRs was relying on net_addr_comp > 0 to order it, it would have broken 18:57 < bridge> [teeworlds] yeah okay. but i'd rather just change the docs 19:06 < bridge> [teeworlds] Sure, but that won't leave any way to have an order relationship between NETADDRs just so you know 19:11 < bridge> [teeworlds] true, but one can add this if it's really needed 19:15 < bridge> [teeworlds] since we have prediction for other players now... what about projectiles? no real physics prediction, but rather just displaying them at the position for the predicted tick, so all objects on the screen are in sync 19:16 < bridge> [teeworlds] for high ping the projectiles may seem to spawn late and fly through walls then, but the experience with prediction for other players is also awful with high ping 19:17 < bridge> [teeworlds] just as a note: that would only require changing a few lines 19:21 < Oy> yeah, sure sounds good 19:23 < Oy> but should probably make a config for it, so people can turn it off. just in case 19:24 < bridge> [teeworlds] yeah sure, it should also be disabled by default, like `cl_predict_players` 19:26 < Oy> yeah 19:27 < bridge> [teeworlds] @Dune if oy appears again, I started a versionsrv running on current master (2020-09-05, dccfaf9354af1dcfbdf60fc517f669346aa8b578) on version.teeworlds.com 19:28 < Oy> good thx, anyone tested if it works? 19:28 < bridge> [teeworlds] @heinrich5991 19:28 < bridge> [teeworlds] iirc it doesn't 19:29 < Oy> hm ok, maybe it's the server? 19:30 < Oy> can't reproduce it and don't see any changes that could cause that 19:42 < bridge> [teeworlds] redix said "but still no response =\" 19:44 < bridge> [teeworlds] yeah henrich5991.de works so it's the server somehow 19:47 < bridge> [teeworlds] Anyone ever experienced the problem that you can see all servers in the browser, but cannot connect to any? After restarting the game everything is fine =\ 19:48 < bridge> [teeworlds] Maybe some uninitialized stuff... 19:48 < bridge> [teeworlds] Cannot connect as in stuck on the connecting screen? 19:48 < bridge> [teeworlds] yeah. times out after a few seconds 19:48 < bridge> [teeworlds] Next time it happens, launch wireshark 😄 19:49 < bridge> [teeworlds] yeah should do that ^^ 19:49 < bridge> [teeworlds] It happened to me once or twice before, but I chalked it up to the very broken 0.7 bridge stuff I was messing around with 19:49 < bridge> [teeworlds] experienced it several times, but it's kinda seldom