00:00 < Learath2> the second forris leaves the conversation becomes more normal :d
00:05 <@deen> a bit bad to watch with antiping on
00:06 < Learath2> i never quite liked antiping
00:06 < Learath2> always felt a tad bit stuttery
00:07 <@deen> yeah, it is. i only enable it when my ping is not what I'm used to
00:10 < Learath2> and if for some reason prediction is too off it looks weird
00:14 <@deen> antiping prediction is always way off by definition
00:14 <@deen> you can't predict when someone will jump or change direction
00:37 < eeeee> you can always improve prediction methods though to make it guess correctly more often
00:38 < eeeee> or do other crazy things like sharing input p2p
00:39 < Learath2> p2p = russian ddos mafia taking down people they hate :P
00:39 <@deen> no idea how you would improve prediction of the input of other players
00:39 < eeeee> Learath2: they are already doing that even w/o p2p just with serverbrowser :P
00:39 < eeeee> so p2p won't change much
00:44 < Learath2> actually with ping even p2p input sharing wont help much
00:45 < eeeee> would halve the latency if all peers are really close
00:46 < Learath2> oic
00:47 < Learath2> could also draw a crude line to finish from start and expect the player to be always moving toward the finish on the line
00:47 <@deen> Learath2: no idea how that would help at all
00:48 <@deen> players don't just jump into the direction of finish
00:48 < Learath2> nvm me just having trouble thinking today :P
00:49 <@deen> making a line from start to finish would be more interesting for creating a ddrace bot
00:49 < eeeee> maybe for solo, not sure how would you make bots work in a team
00:49 <@deen> i guess a gores bot would be easy if you have path finding and the bot thing where you just levitate at any position you want
00:50 <@deen> eeeee: clearly the way to go is not to teach the bots that, but instead create such an advanced AI that the bots will learn it by watching human players and then beat them
00:50 < Learath2> a learning solo bot might be a nice project :D
00:51 < eeeee> if you can watch humans then it's easy to make a bot that beats them
00:51 < Learath2> copy their moves optimize out fails i guess
00:51 <@deen> there was this bot that records the movement of other players and can replay parts of it
00:51 <@deen> yeah, something like that
00:51 < eeeee> just record the humans' demo and make minor alterations to input until you descend the runtime into some local minima
00:52 <@deen> and the bot needs some prediction for the future, like a built-in server so it can simulate a few tries and use the one that's fastest/best
00:52 < eeeee> yeah thats pretty easy
00:52 <@deen> so basically that was my idea for a fun TW project before DDNet got started instead
00:53 <@deen> would've been nice to watch ddrace servers for a few months to train the bots and then beat all ranks in a single night
00:53 < eeeee> 10/10 tsin would just delete ur ranks and lifeban ur ips ;D
00:53 <@deen> he did that anyway :P
00:54 < eeeee> yeah so would've been pointless
00:54 <@deen> pointless if my goal would've been to annoy tsin
00:55 < eeeee> oh also theres that problem where you don't actually get precise input from the demo
00:55 <@deen> right
00:55 < eeeee> movements and jumps are easy to figure out, but i'm not sure how to extract hook
00:56 <@deen> simulating some tries would help i guess
00:56 < Learath2> some experimentation of how imputs map to actual action in demo i'd guess
00:56 < Learath2> input*
00:56 <@deen> Learath2: the inputs you get sent from other players are just not exact enough to reproduce them
00:56 < eeeee> it has to be completely precise or your state would get desynced from the demo and it's usually hard to recover
00:57 <@deen> you can notice it with hook collision
00:57 < Learath2> could probably extract some decent information and optimize it until it is 100% the same with the demo
00:58 < eeeee> yeah i think it's possible, just pretty hard
00:58 < eeeee> e.g. for movement and jumps you have to kind of bruteforce 2 ticks ahead (because you get 1 snap per 2 ticks)
01:01 < eeeee> i have code for a bot lying around somewhere, that can copy your movements and jump
01:01 < Learath2> If you do try making something like this i'd love to try help 
01:01 < eeeee> was gonna add hook but it was too hard
01:02 <@deen> (or send me the patch for skin database downloads instead)
01:02 < eeeee> ^^
01:02 < Learath2> skin database downloads ?
01:04 < Learath2> if you are talking about download skins that are used I think it could be abused by people switching skins and making the others download skins
01:04 < eeeee> could hardcode some limits on that
01:05 < eeeee> like never download more than 1 skin per player
01:05 < eeeee> but yeah, kinda sux
01:06 < Learath2> could disable changing skins in server but still could be abused by leave change join repeat as you still would have to get his skin when he joins
01:09 < Learath2> well skins are quite small and there is a limit to how fast you can change your skin that could help
01:09 <@deen> could have some hard limit on how many skin downloads happen per minute and be happy
02:00 < ddnet-commits> [ddnet] timgame opened pull request #242:  added basic_freeze.rules automapper (DDRace64...DDRace64) http://git.io/vYXtw
02:04 <@deen> Oh, has anyone tested saving on Windows? Since I changed it to move the config and this: http://forum.ddnet.tw/viewtopic.php?t=1901
02:05 < ddnet-commits> [ddnet] def- pushed 3 new commits to DDRace64: http://git.io/vYXqr
02:05 < ddnet-commits> ddnet/DDRace64 b7b7b14 timgame: tiny ddmax_freeze.rules "fix"
02:05 < ddnet-commits> ddnet/DDRace64 ea03517 timgame: added basic_freeze.rules automapper
02:05 < ddnet-commits> ddnet/DDRace64 0f42598 Dennis Felsing: Merge pull request #242 from timgame/DDRace64...
02:12 < ddnet-commits> [ddnet] def- pushed 1 new commit to DDRace64: http://git.io/vYXmh
02:12 < ddnet-commits> ddnet/DDRace64 74391e6 def: Hopefully fix moving config file on save on Windows by allowing overwrite
02:13 < ddnet-commits> [ddnet] def- pushed 1 new commit to DDRace64: http://git.io/vYXY2
02:13 < ddnet-commits> ddnet/DDRace64 d2bec89 def: Version 8.0.2
02:40 <@deen> ok, that worked, in wine at least, good enough
05:05 < eeeee> and if it doesnt work for some windows user you can just recommend him to install wine in cygwin
11:51 < BeaR_> \o
11:52 <@deen> hi
11:52 <@deen> apparently quakenet doesn't work from china
11:52 < BeaR_> huh good to know
11:53 <@deen> invited someone from CHN server to this channel, he can't connect
11:53 <@deen> and he said it's a known proble
11:53 <@deen> m
11:53 < Sadale> interesting.
11:53  * Sadale from China
11:54 < Sadale> precisely, Hong Kong, China
11:54  * Sadale trololo
11:54 <@deen> Sadale: Not sure Hong Kong counts^^
11:54 <@deen> Or maybe it's just Night_L's connection
11:54 <@deen> I just watched 6 people finish Naufrage 2 on China =)
11:54 < Sadale> Well, while we're indeed a part of China, we're not behind their firewall at the moment.
11:54 < Sadale> cool!
11:56 < Sadale> Not sure whether I've told you guys. I've got my auto map update script working :D
11:56 <@deen> what'S that?
11:56 < Sadale> (by download maps from ddnet-maps repository)
11:56 < Sadale> s/download/downloading
11:56 < Sadale> deen, well, it's just a cronjob that download maps from the repo
11:57 <@deen> cool
11:57 <@deen> i know Persian Pro uses something similar to stay updated with ddnet maps
11:57 < Sadale> interesting. Isn't that official server?
11:58 <@deen> no, we have DDNet Persian, but too many ddos attacks, so it's down
11:58 < Sadale> oh :(
11:58 <@deen> Persian Pro is only in the iranian intranet, not on the internet
11:58 < Sadale> what? :o
11:58 < Sadale> Iran has an intranet? :o
11:58 <@deen> they have a separate network in Iran, yeah
11:58 < BeaR_> lol
11:58 < Sadale> interesting :o
11:58 <@deen> hosting and bandwidth is MUCH cheaper in the intranet
11:58 <@deen> and ddos attacks from the internet can't reach you^^
11:59 <@deen> normal internet hosting is so expensive, it costs as much as many people earn...
11:59 < Sadale> deen, how do you address the machines in intranet?
12:00 <@deen> Sadale: they used some private ip range for it, 10.x.x.x
12:00 < Sadale> deen, ah! Now it makes sense :>
12:00 < Sadale> I was thinking how can it address both internet and intranet address at the same time. Now I get it.
12:01 <@deen> They have a guide here to add their servers to favorites, because they're not in the internet server browser of course: http://persian-protee.ir/
12:02 < BeaR_> hm was there any decision made concerning the race servers?
12:02 <@deen> BeaR_: no, I'm really not sure what to decide
12:02 <@deen> BeaR_: what would you say?
12:02 < Sadale> deen, lol I can't read moon runes xD
12:03 < BeaR_> deen: would keep them, don't really see any drawbacks
12:03 <@deen> for me the problem is that only very few players want the race server
12:03 <@deen> and I have no one willing to test the maps of the race server
12:03 <@deen> the initial team was Welf, Index, aaa
12:03 <@deen> aaa left the team, Welf isn't interested and Index has no time
12:04 < BeaR_> oh well /:
12:04 < Sadale> What's race server? It is form the mod "Race"?
12:04 <@deen> yes
12:04 <@deen> afk
12:04 < Sadale> s/form/from/
12:24 < Sadale> speaking of intranet, my home does have a simple intranet hosting a few services :p
12:25 < Sadale> It includes print/scan service, copy and paste service and a few more less useful services.
14:18 < ddnet-commits> [ddnet] timgame opened pull request #243: generic_clear.rules automapper by hi_leute_gll (DDRace64...DDRace64) http://git.io/vYDLO
14:29 < ddnet-commits> [ddnet] def- closed pull request #243: generic_clear.rules automapper by hi_leute_gll (DDRace64...DDRace64) http://git.io/vYDLO
16:14 < ddnet-commits> [ddnet] def- pushed 2 new commits to DDRace64: http://git.io/vYDx9
16:14 < ddnet-commits> ddnet/DDRace64 eb29a21 def: Improve friends and foes feature...
16:14 < ddnet-commits> ddnet/DDRace64 8ba387e def: Merge branch 'DDRace64' of github.com:def-/ddnet into DDRace64
16:53 < ddnet-commits> [ddnet] timgame opened pull request #244: added basic_freeze.rules automappers (DDRace64...DDRace64) http://git.io/vYy3x
17:04 < ddnet-commits> [ddnet] def- pushed 3 new commits to DDRace64: http://git.io/vYyWl
17:04 < ddnet-commits> ddnet/DDRace64 af2b43c timgame: added round corners automapper by hi_leute_gll
17:04 < ddnet-commits> ddnet/DDRace64 5162bd0 timgame: added diagonal corners automapper
17:04 < ddnet-commits> ddnet/DDRace64 2c5d6ed Dennis Felsing: Merge pull request #244 from timgame/DDRace64...
17:06 < Chairn> i dont get something, compiler crashes, but the output is completely weird
17:06 < Chairn> it doesnt seem to like this function
17:06 <@deen> Chairn: huh?
17:06 < Chairn> static char* Tile(int Index)
17:07 < Chairn> it's just a function to help, that translate an index number into a char* to be more readable
17:07 < Chairn> 0 == TILE_AIR, etc..., i know there are some index that appears twice
17:09 <@deen> you mean const char*?
17:09 <@deen> can't tell what's going wrong without code or error
17:09 < Chairn> src/game/mapitems.h:484:16: warning: dIenp rfeiclaet eidn cclounvdeersdi ofnr of mr osmr cs/tgraimneg/ ccoolnlsitsainotn .tcop p':c1h1a:r0*:' [s-rWcw/rgiatme-ster/imanpitems.h: In function 'char* Tile(ignt)': src/games/ma]p t  e  m s . h : 4r8e2t:u1r6n:  "wTaIrLnE_iSnOgL:I Dd"e;p ec a t e d   c o n v e r s i o n  ^f om strinsgr cc/ognasmtea/nmta ptiot e'mcsh.ahr:*4'8 6[:-1W3w:r iwtaer-ns
17:09 < Chairn> here is what the compiler says :D
17:10 <@deen> that means you're getting two error messages overlapping
17:10 <@deen> compile with bam -j1
17:11 <@deen> it's a bug with bam on windows i guess
17:11 < Chairn> https://github.com/Chairn/ddnet/commit/18da43471a72ea953e2fc1f4d0266c78b6d2f717
17:11 < Chairn> Unknown switch -j1
17:11 <@deen> -j 1
17:12 < Chairn> okay, so warning + error due to duplicate case
17:14 < Chairn> switchtimedopen == solo_end
17:14 <@deen> right, i know
17:14 <@deen> but on different layers
17:15 < Chairn> now it compiles, ty ;)
17:16 < Nimda> Uzmutulunculunbé by Fňokurka oo7 just released on Brutal at 2015-07-28 17:08
18:39 < o_be_one> hi all :)
19:07 < laxadedi> Chairn: Tu compiles avec quoi ?
19:07 < Chairn> mingw
19:08 < laxadedi> Tu te mets a developper ou quoi ?
20:37 < o_be_one> hey guys just wanted to know
20:37 < o_be_one> is there really a trojan in baumalein ? have you some proofs ?
20:38 < o_be_one> btw just to let you know that pikotee and meskalin can do clones/bots attack (when they fill the server with tee dummy)
20:40 < Learath2> o_be_one: i believe EastByte actually could listen to the packets they use to control the clients from the outside
20:40 <@EastByte> o_be_one: yes we know about the client
20:41 < o_be_one> ok
20:41 <@EastByte> he can connect votebot dummys using the client's clients
20:41 <@EastByte> that's all
20:41 < o_be_one> ok thank you :)
23:30 < ddnet-commits> [ddnet] def- pushed 2 new commits to DDRace64: http://git.io/vYHWu
23:30 < ddnet-commits> ddnet/DDRace64 6a2cc86 def: Sort spectators by name
23:30 < ddnet-commits> ddnet/DDRace64 30f0b9c def: Merge branch 'DDRace64' of github.com:def-/ddnet into DDRace64
23:54 <@deen> any thoughts on the flag showing here?: http://ddnet.tw/ranks/
23:59 < WolfAlex> deen: why don't you show the default flag instead of these points?
23:59 <@deen> WolfAlex: those points are supposed to mean "same flag as last" =/
23:59 <@deen> but if you ask that question i guess that isn't as obvious as I hoped
23:59 < WolfAlex> ah xd