01:04 <@deen> Didn't see much of the announced attacks. No big ddos or any server crashes
01:14 < DMTee> hey jemand da?
01:14 <@deen> hi DMTee 
01:14 < DMTee> hi
01:14 <@deen> this is an english chat
01:14 < DMTee> tok bot
01:14 < DMTee> whats up deen?
01:15 <@deen> nothing
01:15 < DMTee> i finished a map in 528 minutes
01:15 <@deen> nice, which?
01:15 < DMTee> i have chances to get a video of that on ddnet ? :D
01:15 <@deen> not in the hall of fame, don't know who else makes videos
01:16 < DMTee> difficult 2.1
01:16 < DMTee> just thought that maybe the highest time ever have chances :D
01:17 <@deen> there are higher times
01:17 < DMTee> ok
01:17 < DMTee> but u are german right?
01:18 <@deen> yes
01:18 <@deen> for example milk on "At Sundown": 612 minutes
01:18 <@deen> darky ragequit after about 700 minutes iirc
01:19 < DMTee> is doch niemand da grad können wir nich in deutsch reden mal kurz?
01:20 <@deen> only if it's really necessary
01:20 < DMTee> no it isnt
01:20 < DMTee> but its more easy if u talk in your mother language espacially if there is now reason why you dont do so
01:21 <@deen> for me English is more comfortable
01:21 <@deen> also others might want to read up on the chat later
01:21 <@deen> just checked, your time is the 4th highest ever
01:21 < DMTee> ok sry
01:21 < DMTee> nice
01:21 <@deen> except for milk on "At Sundown" there were also 2 guys on run_guy_25
01:22 < DMTee> are u the "boss" of ddnet?
01:22 <@deen> 12:345 with 543 minutes and Kobr(CCCP) with 550 minutes
01:23 <@deen> I run the servers and have developed some of the stuff, yes
01:23 < DMTee> how many hours of work u spend every day on this?
01:23 <@deen> is this an interview?
01:24 < DMTee> i mean ddnet is some great stuff
01:24 < DMTee> u can see that things are done with love and passion
01:24 < DMTee> without getting any money from it
01:24 <@deen> yeah, it actually costs quite a bit
01:25 <@deen> I don't have any exact numbers
01:25 < DMTee> so u play since houre 0?
01:25 <@deen> some days I spend just 1-2 hours on ddnet, other days more
01:25 <@deen> play teeworlds? or what?
01:25 < DMTee> tyes
01:26 < DMTee> tsince 4.9 or what was the latest version?
01:26 <@deen> You've spent too much time in Teeworlds obviously :P
01:26 <@deen> All the "t"s at the start, haha
01:26 < DMTee> thehe why u ask?
01:26 < DMTee> :D :D:D:
01:26 <@deen> teeworlds 4.9?
01:26 < DMTee> yeah maybe
01:26 <@deen> i thought it's at 0.6.3 right now
01:26 < Nimda> DDNet Brazil went down!
01:27 < DMTee> 6.04
01:27 <@deen> I don't understand
01:27 < Nimda> DDNet Brazil went back online!
01:28 < DMTee> i mean how long u play this game ?
01:28 < DMTee> since waht version
01:28 <@deen> No idea^^
01:28 <@deen> I played teeworlds a bit when it became available in the gentoo repository, would have to research when that was
01:28 <@deen> but definitely some time before it was renamed from Teewars
01:28 < DMTee> u played it before it become famous on giga?
01:29 <@deen> no idea, never watched giga
01:29 <@heinrich5991> so long, wow :D
01:29 <@deen> i didn't really play much, maybe a game and break for a month
01:30 <@heinrich5991> I learned about it when we had a project in school and were in the computer room for that
01:30 <@heinrich5991> so we needed network games t hat don't need installation ;)
01:31 < DMTee> u established some really great stuff with ddnet thx 4 all the time and engagement u spend on it
01:31 <@deen> heinrich5991: haha
01:31 <@deen> DMTee: you're welcome
01:32 < DMTee> how ald are u deen?
01:32 <@deen> still feels like an interview :P
01:33 <@deen> 24 i think
01:33 < DMTee> yeah it also is somehow :D
01:33 < DMTee> just wanna know who is that guy behind ddnet
01:34 < DMTee> doeas halloweed stream anymore or doesnt he want to anymore on sundays tournaments?
01:35 <@deen> i don't think he'll stream anymore unfortunately
01:36 < DMTee> u got some alternatives?
01:36 <@deen> nope
01:36 < DMTee> the streaming was nice damn
01:36 <@deen> I know, he's great at that. No idea why he stopped now
01:37 < DMTee> ok i dont wanna interupt u anymore :) 
01:38 < DMTee> have a good night
01:38 <@deen> If you want to help you can try to find great streamers
01:38 <@deen> good night
01:45 < Nimda> DDNet Brazil went down!
01:46 < Nimda> DDNet Brazil went back online!
01:57 < Nimda> DDNet Brazil went down!
01:58 < Nimda> DDNet Brazil went back online!
02:28 <@deen> Channel about ddnet on Russian 2ch/4chan clone: http://m2-ch.ru/vg/res/9206126.html
02:28 <@deen> s/Channel/Thread/
02:55 < eeeee> nothing interesting there
02:55 <@deen> Yeah, I've tried to understand the English translation
02:56 <@deen> Hm, how to implement this best?: http://forum.ddnet.tw/viewtopic.php?t=764&p=6860#p6860
02:56 <@deen> Could change the gametype of block servers
02:57 <@deen> or add an exclude field to the search
02:57 <@deen> or allow for regexps in the search field :P
02:57 < eeeee> doesn't it make sense to have actually different gamemode for block anyway?
02:58 <@deen> i don't know, it's actually 100% the same mod right now
02:58 <@deen> but the gameplay is different
02:58 <@deen> And just when we're so close to becoming the most played mod!!!: http://teeworlds-stats.info/general
02:59 < eeeee> does that thing even account for >16 players?
02:59 <@deen> nope
02:59 < eeeee> :P
02:59 <@deen> anyway, yes, just changing the gametype would be easiest
03:00 <@deen> any suggestions for the name?
03:02 <@deen> and there's a problem with the name because some features are rather hardcoded with the server type =/
03:03 <@deen> DDraceNetworkBL or BlDDraceNetwork or BlockDDraceNetw would work
03:39 < ddnet-commits> [ddnet] def- pushed 1 new commit to DDRace64: http://git.io/im3VgQ
03:39 < ddnet-commits> ddnet/DDRace64 8b108a2 def: Clean up gametype determination
04:50 < ddnet-commits> [ddnet] def- pushed 1 new commit to DDRace64: http://git.io/qjtUpw
04:50 < ddnet-commits> ddnet/DDRace64 76b6a99 def: Add Exclusion filter strings and clean up menu a bit
11:21 < ddnet-commits> [ddnet] def- pushed 1 new commit to DDRace64: http://git.io/x1KdEQ
11:21 < ddnet-commits> ddnet/DDRace64 dfa804f def: Add missing file
15:49 < o_be_one> morning !
16:37 < Nimda> Rayquaza by Zerodin just released on Moderate at 2014-12-10 16:36
16:58 < Savander> hi
17:10 <@deen> hi Savander 
18:17 < BeaR> sup
18:30 <@deen> hi BeaR 
18:30 <@deen> nice smoke trails
18:30 < BeaR> ^^
18:31 < BeaR> just needed to cleanup the code, smoke came by itself :D
19:29 < BeaR> deen: haskell has some sort of default constructor?
19:29 < Nimda> DDNet CHN went down!
19:30 <@deen> what do you mean, BeaR ?
19:30 <@deen> what kind of constructor?
19:30 < Nimda> DDNet CHN went back online!
19:31 < BeaR> sth like I create a standard object of a type
19:31 < BeaR> I need to return sth but not an exception /:
19:31 < BeaR> fiddling around with this for some time but couldn't  get it working ._.
19:32 <@deen> which exercise is it?^
19:32 < BeaR> H9.4
19:32 < BeaR> h ::
19:32 <@deen> you could add something to your datatype maybe?
19:32 <@deen> like data X = A | B | Exception
19:34 <@deen> that looks like more fun than our exercises
19:34 < BeaR> xd
19:34 <@deen> (maybe because I've seen ours for 4 years now and they didn't change much)
19:34 < BeaR> looks more like magic
19:35 < BeaR> well I could let it loop forever but I doubt thats the solution ..
19:35 <@deen> nope, it says you're not allowed to do that
19:35 <@deen> so, g is easy, right?
19:35 <@deen> g x _ (Left a) = Left (x a)
19:35 <@deen> g _ y (Right b) = Right (y b)
19:35 < BeaR> y
19:36 < BeaR> "soweit möglich" :>
19:36 <@deen> ok, I get it
19:36 <@deen> h :: (a -> Either a b) -> a -> b
19:37 <@deen> oh wait
19:37 <@deen> guess you have to loop or something
19:37 < BeaR> yeah dunno
19:37 <@deen> from the first function you can get a (Left x) or (Right y)
19:37 <@deen> if it's a (Right y) you can just return y
19:37 < BeaR> y (Right x) is the problem
19:37 < BeaR> Left *
19:38 <@deen> but if it's a (Left x) you can't determine a value for b
19:38 <@deen> looks impossible to me
19:38 <@deen> ah
19:38 <@deen> you could pass x back to the function
19:38 <@deen> until it returns Right y
19:39 < BeaR> hm true
19:39 <@deen> probably the only useful solution
19:39 < BeaR> yeah that's probably the pont
19:39 < BeaR> thanks gonna try it (:
19:41 <@deen> That's an interesting area of research btw, determining what a function can possibly do given only its type
19:41 <@deen> There were some PhD positions open for that in the UK iirc
19:46 <@deen> BeaR: have it?
19:46 < BeaR> testing (:
19:46 < BeaR> it's correct, nice (:
19:47 < BeaR> thanks again
19:47 <@deen> Show me your solution!
19:47 < BeaR> h foo x = h' foo (foo x) 
19:47 < BeaR> h' foo (Left x) = h foo x
19:47 < BeaR> h' foo (Right x) = x
19:48 <@deen> Hm, should write a "where" before h'
19:48 < BeaR> (probably can be written nicer)
19:48 <@deen> or with a case
19:48 <@deen> h f x = case f x of Left x'  -> h f x' Right y' -> y'
19:48 <@deen> h f x = case f x of
19:48 <@deen>   Left x'  -> h f x'
19:48 <@deen>   Right y' -> y'
19:49 < BeaR> y looks nicer
19:49 < BeaR> but im not used to that atm (:
19:49 <@deen> needs more crazy stuff
19:49 < BeaR> (well never used case at all)
19:49 < BeaR> need more dots
19:49 <@deen> h f = (\e -> case e of {Left x -> h f x; Right y -> y}) . f
19:49 < BeaR> (.).(.).(.)
19:49 <@deen> one liner
19:50 < BeaR> black magic
19:50 <@deen> nah, it's exactly the same as the case
19:50 <@deen> and case is like switch-case in C
19:52 < BeaR> how will your solution evaluate?
19:52 < BeaR> dont really understand it completly
19:53 <@deen> hm?
19:53 <@deen> first or second?
19:53 < BeaR> h f = (\e -> case e of {Left x -> h f x; Right y -> y}) . f
19:53 < BeaR> why .f
19:53 <@deen> first you want to run f on the Either value
19:53 < BeaR> ah
19:53 <@deen> afterwards the result of that is passed to the case
19:54 < BeaR> now xd
19:55 < BeaR> dunno can'T get used to the dot operator
19:55 < BeaR> it's too small!
19:55 < BeaR> (and I'm serious about this)
19:55 <@deen> I don't like to use it normally
19:55 <@deen> it's just fun to annoy students :P
19:55 < BeaR> true :<
19:56 < BeaR> so many people are considering to stop going to the course (:
19:56 <@deen> if you use a language extension it gets even nicer!
19:56 <@deen> {-#LANGUAGE LambdaCase #-}
19:56 <@deen> h f = (\case {Left x -> h f x; Right y -> y}) . f
19:57 < BeaR> eh
19:57 < BeaR> that moment when I saw {- #LANGUAGE C++ -}
19:57 <@deen> I use C pragmas in my haskell code, haha
19:57 < BeaR> huh why
19:58 <@deen> some code works with some GHC versions, but not with others..
19:58 <@deen> never write code that peaks into the GHC internals
19:58 <@deen> ehm, not C pragmas, C preprocessor*
19:59 <@deen> lots of ifdefs
19:59 < BeaR> (never write c++11 code when you working on windows and it's 2014)
20:00 <@heinrich5991> things might change
20:00 <@deen> Ah, don't need the brackets!
20:00 <@deen> h f = \case {Left x -> h f x; Right y -> y} . f
20:00 <@heinrich5991> BeaR: did you know clang tries to produce a drop-in replacement vor msvc++'s compiler?
20:00 < BeaR> heinrich5991: yeah already using it :D
20:00 <@heinrich5991> great!
20:00 <@heinrich5991> do you know whether teeworlds compiles with that?
20:00 < BeaR> still not supports everything though
20:00 <@heinrich5991> does it have downsides?
20:00 <@heinrich5991> e.g.?
20:00 < BeaR> hm not sure
20:01 < BeaR> uhm can't remember tbh
20:01 < BeaR> just tried it with my engine
20:01 < BeaR> but got some errors
20:01 <@deen> I compiled DDNet with clang recently, about same performance
20:02 < BeaR> I think there were some errors with the windows headers iirc
20:02 < BeaR> would love to change to clang, the compiler output is so nice
20:03 < BeaR> visual studio 2015 uses clang for android and ios support
20:04 <@heinrich5991> wow
20:04 <@heinrich5991> that's an admittance of failure :D
20:05 < BeaR> well still not fully supported constexpr in the latest cl /:
20:06 <@deen> You guys should be happy you don't have to use Intel's C++ compiler!
20:06 <@heinrich5991> constexpr is a hard problem
20:06 < BeaR> writing RenderCommand::get_index<SetPipeline>:type is not funny (:
20:06 < BeaR> deen: :D
20:07 <@deen> BeaR: I had to work around multiple intel compiler bugs to get a big graph algorithms library ported...
20:07 < BeaR> y constexpr is probably pretty much work to implement but still /:
20:08 < BeaR> RenderCommand::get_index<SetPipeline>:type, being able to use this requires latest visual studio though xd
20:08 < BeaR> is intel compiler up to date?
20:09 <@deen> yeah, icpc15 actually is
20:09 <@deen> but it's no longer available for free..
20:09 < BeaR> mh..
20:09 <@deen> https://software.intel.com/en-us/articles/c0x-features-supported-by-intel-c-compiler
20:09 <@deen> C++11 nearly fully supported!!
20:09 <@deen> C++14 not really: https://software.intel.com/en-us/articles/c14-features-supported-by-intel-c-compiler
20:11 < BeaR> huh, vs2013 has an implementation of <filesystem>
20:12 < BeaR> really hoping c++ wil support reflection in the next years
20:13 < BeaR> (so I dont have to use clang to extract stuff from header files)
20:14 <@heinrich5991> BeaR: do you want run-time reflection or compile-time reflection?
20:14 < BeaR> compile time
20:15 < BeaR> I generate the meta data via clang and build header files
20:15 < BeaR> useful for GUI and such stuff (:
20:16 <@deen> You need to impress your McHaskell guy
20:17 < BeaR> huh?
20:17 <@deen> I have a fun solution to the exercise :P
20:17 < BeaR> btw did you finish this hex game competition ?
20:18 <@deen> nope
20:19 <@deen> h f = (\(Right x) -> x) . foldr1 (\case {Left _ -> id; c -> const c}) . iterate (\(Left x) -> f x) . Left
20:19 <@deen> (non-recursive)
20:21 < BeaR> eh
20:21 < BeaR> Left _ -> id, this will return the _ ?
20:21 <@deen> nope
20:21 <@deen> it returns the identity function
20:22 <@deen> which will take the next value and return that
20:23 < BeaR> (too much for me ^^)
20:28 < BeaR> deen: you alread did the fuzzing testing for the network as I've seen some related comments
20:28 <@deen> well, it hasn't progressed far
20:28 <@deen> and I'm just testing the basic connection process
20:28 <@deen> I have stopped the fuzzing test in favor of freetype :P
20:29 <@deen> (and a few other libraries)
20:36 < BeaR> what are you doing with freetype :?
20:36 <@deen> fuzzing it
20:36 <@deen> found 2 segfaults in it
20:37 < BeaR> xd
20:39 <@deen> They just made a security release, I'm clearly not the only one fuzzing
20:39 <@deen> I never know how to properly report security bugs though
20:39 <@deen> Wrote a mail to the 2 main people, but no response
20:39 <@deen> So I opened a private bug report, but I can't see it myself^^
20:39 < BeaR> :)
20:42 <@deen> some people just openly report them on the bugtrackers, but I think that's a bad idea, to just publicate possible exploits like that
20:50 < eeeee> otoh if you don't publicate it they might get ignored for longer
21:06 <@deen> yay, 30 times speedup on the Xeon Phi with a single line change!
21:06 < BeaR> huh lol
21:06 <@deen> the scheduling was terrible, i fixed it
21:07 <@deen> nowit scales totally linearly, it's beautiful
21:08 < BeaR> ^^
23:58 < xRoThx> deen: 
23:58 <@deen> hi xRoThx 
23:58 < xRoThx> Everything alright? 
23:59 < xRoThx> I heard about a hack of Candys server? 
23:59 <@deen> Actually I wanted to ask you something
23:59 < xRoThx> Sure
23:59 <@deen> Didn't hear of any problems with DDNet
23:59 <@deen> There were many ways to steal rcon password, if that's what happened to Candys