09:04 < rand> Dune: from server side at best ?
11:02 < Oy> Dune: what do u think about https://github.com/teeworlds/teeworlds/pull/2396/commits/a601865314766f7e3baffabf32afc422e21f83f3#diff-dab5fcb897efd8b09195abaaa043e621R82 ?
11:02 < Oy> old address will be gone because of the name change
11:06 < bridge> [teeworlds] <fokkonaut> that doesnt matter
11:06 < bridge> [teeworlds] <fokkonaut> in this case
11:38 < bridge> [teeworlds] <Dune> Oy: as long as it doesn't touch connect, I think it's unimpactful
11:39 < bridge> [teeworlds] <Dune> I would have used ui_server_address and ui_server_address_lan though, just in case
11:42 < bridge> [teeworlds] <Dune> rand: you cant change the code for a CTF server
11:42 < rand> then, that's set, no bots on vanilla
11:45 < Oy> yeah, would prefer that name sheme as well
11:46 < Oy> rand: why?
11:48 < rand> because Dune said that ctf code cant be change to this extent
11:49 < rand> I did not say my point though
11:49 < bridge> [teeworlds] <Dune> client bots tho
11:50 < rand> I feel like bots in vanilla is fine as long as they are server sided
11:50 < bridge> [teeworlds] <Dune> why?
11:51 < Oy> well you can't protect your server from client bots anyway
11:51 < rand> because I dont expect a bot to join a server without bot while I am playing
11:51 < bridge> [teeworlds] <Dune> it would make the "pure server only" rule less strict
11:51 < rand> Oy: yes
11:52 < rand> Dune: allow official client bots also is
11:52 < bridge> [teeworlds] <Dune> then clients should report a bot flag, but can't check that either:/
11:52 < rand> pure server are official game types
11:52 < Oy> doubt a server will be banned from master because there are client bots on that server
11:53 < bridge> [teeworlds] <Dune> yeah you can't enforce that, but it is abuseable
11:53 < rand> well, that case occurs
11:53 < Oy> could do the client bot flag though
11:53 < rand> *occured
11:53 < Oy> huh really?
11:54 < Oy> people can just ban servers that way then, really bad
11:54 < rand> someone complained
11:54 < bridge> [teeworlds] <Dune> can't expect the server to know which client is a bot
11:55 < rand> that is also a reason not to bring client side bot in the loop
11:55 < Oy> we could tie that client bot flag to a specific ip whitelist. then people can'T abused when u run the bot clients on the same mashine
11:56 < rand> why a client would run a bot on a server that's not his own ?
11:58 < Oy> dunno, maybe hardware restrictions
12:13 < Dune> we can only engage the responsibility of servers, not clients though
12:14 < Dune> any idea where to put a prediction margin stability indicator in the UI? https://streamable.com/2y9jx
12:14 < Dune> widening the scoreboard just for that sucks
12:14 < Dune> and it is only visible for the self player
12:17 < Oy> below fps counter, top right corner
12:17 < Dune> only shows with scoreboard open?
12:18 < Oy> ah thought you can enable it like showfps
12:18 < Dune> it could be done like that
12:18 < Dune> there are killmessages there though
12:20 < Dune> https://puu.sh/F3SFK/2c20e3316e.png
12:21 < Dune> something like the rec popup?
12:23 < Oy> yeah
12:31 < bridge> [teeworlds] <fokkonaut> oy: what happens if you set max clients < amount of players?
12:38 < Oy> all players stay on the server, nobody gets removed. new players can only join when max_clients > players
12:40 < Dune> rand, >why a client would run a bot on a server that's not his own ?
12:40 < Dune> another usecase is dummies
12:46 < bridge> [teeworlds] <fokkonaut> oy, so the browser shows for example 20/16 players?
12:47 < bridge> [teeworlds] <fokkonaut> shouldnt it show max(g_config.m_svmaxclients, clientcount)?
13:02 < rand> are dummies both controlled ? could be this handled server side too ?
14:23 < Oy> @fokkonaut fixed that
14:23 < Oy> rand: dummies aren't broadcasted to non-ingame clients
14:26 < rand> I dont understand what I should understand
14:27 < Dune> (I was talking of client dummies)
14:59 < rand> I was asking if a dummy can/will receive input or is brain dead the whole game
15:00 < rand> and I was wondering if this could be handled server-side
15:01 < rand> but, does a client dummy behave like a new client ?
15:01 < rand> what is about not broadcasting dummies ?
15:02 < rand> in server info (only) ?
15:02 < Dune> yeah
15:02 < Dune> but that's probably not a big concern atm 
15:02 < Dune> just inflates the human count a bit
15:13 < rand> i think that a client might ask a server for its dummy in a nice future
15:17 < rand> but, if ddnet client allows the player to exchange the control between the dummy and the original character, this idea might lead to higher latency for control swapping
15:56 < bridge> [teeworlds] <Dune> windows taskbar notifications upon ping/whispers
15:56 < bridge> [teeworlds] <Dune> https://cdn.discordapp.com/attachments/490150878934990850/671005654080552960/unknown.png
15:57 < bridge> [teeworlds] <Dune> yeay
15:57 < bridge> [teeworlds] <fokkonaut> ddnet has this too
15:57 < bridge> [teeworlds] <Dune> does it work well?
15:57 < bridge> [teeworlds] <fokkonaut> yes, very
15:58 < bridge> [teeworlds] <Dune> does it work with linux/osx too?
15:58 < bridge> [teeworlds] <fokkonaut> no idea
15:58 < bridge> [teeworlds] <fokkonaut> guess you can take a look at the code
16:01 < bridge> [teeworlds] <Dune> "doing it platform-dependent would be too much of a hassle regarding linux and OS X." (Oy)
16:01 < bridge> [teeworlds] <Dune> apparently it's not trivial for other OS
16:18 < bridge> [teeworlds] <Assa> hope it's only the taskbar and not the other annoying windows notification thing
16:22 < bridge> [teeworlds] <ChillerDragon> As far as I remember it worked fine on Linux and even macOS
16:22 < bridge> [teeworlds] <ChillerDragon> But I had it deactivated most of the time :p
16:25 < bridge> [teeworlds] <ChillerDragon> Well in my case if I remember correctly it used the normal gnome popups and the macOS popups so not a blinking icon in the taskbar
21:21 < bridge> [teeworlds] <heinrich5991> @jxsl13 https://discordapp.com/channels/407308363031117832/490150878934990850/670610317037797396 I asked matricks about that once, and he showed me that it compiles to different code
21:21 < bridge> [teeworlds] <heinrich5991> one without branches and one with
21:27 < bridge> [teeworlds] <ChillerDragon> @Dune i reached 800 lines for a single automapper rule with comments tho
21:27 < bridge> [teeworlds] <ChillerDragon> but it doesnt look like im done any soon xd
22:05 < Dune> Oy: build broke?
22:12 < Oy> oh didn't build fake server
23:24 < bridge> [teeworlds] <jxsl13> does it change how that compression is executed @heinrich5991 ? or what exactly changes there?
23:25 < bridge> [teeworlds] <jxsl13> did I just see a new commit about shutting down econ connections when the server is shutdown properly?
23:25 < bridge> [teeworlds] <heinrich5991> the cpu instructions at various optimization levels
23:25 < bridge> [teeworlds] <jxsl13> ah, you mean code is getting less optimized in my version?
23:25 < bridge> [teeworlds] <heinrich5991> I believe only `clang` with `-O3` was able to generate the same instructions as matricks' code there
23:25 < bridge> [teeworlds] <heinrich5991> yes
23:25 < bridge> [teeworlds] <jxsl13> hm, interesting
23:26 < bridge> [teeworlds] <jxsl13> write bad looking code for good looking optimization 😄
23:32 < bridge> [teeworlds] <heinrich5991> https://godbolt.org/z/5qHAKe
23:32 < bridge> [teeworlds] <heinrich5991> I haven't looked further into it
23:32 < bridge> [teeworlds] <jxsl13> there is some weird server running under74.91.114.132:9303 that returns an incorrect token for browsers ._.
23:33 < bridge> [teeworlds] <heinrich5991> but the assembly output of the current version is shorter than your version
23:33 < bridge> [teeworlds] <heinrich5991> on gcc, clang with `-O3`
23:42 < bridge> [teeworlds] <jxsl13> cannot tell much about that.
23:43 < bridge> [teeworlds] <jxsl13> well, codewise yes, but does shorter necessarily mean faster Oo?
23:44 < bridge> [teeworlds] <heinrich5991> no
23:44 < bridge> [teeworlds] <heinrich5991> not in general
23:45 < bridge> [teeworlds] <jxsl13> for example,  `|= (!= 0) << 7` might be performed in any case with the current version
23:45 < bridge> [teeworlds] <jxsl13> imagine there being an i in front of the `!=`
23:45 < bridge> [teeworlds] <jxsl13> as I cannot read assembly straight like I can with c++ x)
23:46 < bridge> [teeworlds] <jxsl13> I'm not 100% sure
23:48 < bridge> [teeworlds] <jxsl13> is it possible to keep a line highlighted without having to hover with ones cursor above it
23:51 < bridge> [teeworlds] <heinrich5991> dunno
23:51 < bridge> [teeworlds] <heinrich5991> @ keeping it highlighted