02:27 < rymer> ey xxltomate
11:39 < f4c3> anyone online?
11:40 < f4c3> my question is: where to write the command "git clone path" ??
12:22 < f4c3> Can anyone help me?
12:23 < f4c3> When I type the command "git clone path" it everytime says fatal: cant connect blabla
12:33 <@heinrich5991> f4c3: can you give the full command and the full error?
12:41 < f4c3> yes wait
12:41 < f4c3> command: git clone https://github.com/ddnet/ddnet-maps
12:44 < f4c3> error: unable to access 'https://github.com/ddnet/ddnet-maps': error setting certificate verify locations:   :CAfile: C:Programme/Git/mingw32/libexec/ssl/certs/ca-bundle.crt   ... CApath: none
12:51 <@deen> Oh, git clone on Windows :/
12:51 <@deen> f4c3: what are you  trying to do? If you just need the maps once, you can download the zip file
12:52 <@deen> I don't know how to solve it properly, but as an insecure workaround this might work: git config --global http.sslverify "false"
12:56 < f4c3> I try the second method at the mainpage from ddnet.tw .... i want to easy update ir
12:56 <@deen> ok
12:57 <@deen> If you know of a good tutorial for getting git running on Windows, I could add that on ddnet.tw
12:58 < Gabee_> deen: I'm really sorry
12:58 < Gabee_> git in the ubuntu bash on w10?
12:59 < f4c3> Shall i use bash?
12:59 < f4c3> because there it dont work
13:00 <@deen> No idea, I don't know all this windows stuff
13:01 < f4c3> you use linux?
13:01 <@deen> yeah
13:02 < f4c3> ah nice ...:)
13:02 < f4c3> i search for a tutorial
18:42 < vali> huhu *_*
19:21 < ddnet-commits> [ddnet] def- pushed 1 new commit to master: https://git.io/vrWnh
19:21 < ddnet-commits> ddnet/master bb634d7 def: Reconnect to last server address, not the one selected in browser
21:29 <@deen> I thought about using recvmmsg to get a bit more performance, but it's broken: "The timeout argument does  not  work  as  intended.   The  timeout  is checked  only  after  the  receipt  of each datagram, so that if up to vlen-1 datagrams are received before the timeout expires, but then  no further datagrams are received, the call will block forever."
21:30 <@deen> Should be fine with non-blocking reads only I guess
21:32 <@EastByte> Okay
21:37 <@deen> A third of the sys calls of a DDNet server are recfrom and sendto each
21:37 <@deen> they always occur in large chunks, so ideal for sendmmsg/recvmmsg
21:38 <@deen> the last third are mostly strange time and gettimeofday calls, I thought I got rid of most of them
21:40 <@deen> server with 30 players causes 3000 syscalls a second
22:00 <@deen> PACKET_MMAP is very cool, reading packets with 0 syscalls, too bad it's not for regular applications (requires root, doesn't work with normal udp socket): https://www.kernel.org/doc/Documentation/networking/packet_mmap.txt
22:04 <@deen> and then the glibc version matters a lot for syscalls. new versions of glibc don't syscall at all for gettimeofday
22:05 <@deen> (or something else is causing that, not sure yet)
22:13 <@deen> Reading the glibc implementation, that's done by vDSO, interesting: http://man7.org/linux/man-pages/man7/vdso.7.html
22:19 <@heinrich5991> what's recvmmsg?
22:19 <@heinrich5991> ah
22:19 <@heinrich5991> receive multiple packets
22:32 <@deen> totally confused why some of our servers use vdso gettimeofday, others not even though they have more recent kernel and glibc
22:41 <@deen> ok, probably depends on the underlying clock that the vps uses. pvclock is used with kvm and doesn't support vdso. but looks like there's some progress being made: https://lkml.org/lkml/2015/12/9/914
22:45 < Gabee_> deen: you're on vpses?
22:46 < Gabee_> wouldn't it cheaper to get dedis?
22:46 <@deen> how are dedicated servers cheaper?
22:46 < Gabee_> it depends on how much vpses you have
22:47 < Gabee_> so my bad, I thought you had a lot of vpses :P
22:47 <@deen> 1 per location
22:47 <@deen> usually 1 weak cpu core is enough as well
22:47 < Gabee_> Oh
22:47 < Gabee_> So all the ddnet ger.. are on one VPS?
22:47 <@deen> just GER is struggling a bit since we got switched to an overloaded server, so I'm optimizing the server a bit more
22:47 <@deen> yeah
22:48 <@deen> (and GER only struggled in tournament, not usually)
22:55 < Gabee_> Oh interesting
22:55 < Gabee_> I thought it would require more ressources to run the server
22:55 < Gabee_> deen: if you want to check I'm running an iwar server
22:56 < Gabee_> it's currently in dev, iwar is a ddnet-based copy of ddwar
22:56 < Gabee_> with up to date stuff
22:56 <@deen> what's ddwar?
22:57 <@deen> and resource usage is pretty low. the biggest problem are big maps for RAM. CPU-wise 200-500 players work fine with a core depending on its performance
23:07 < Gabee_> you don't know what is ddwar?
23:07 < Gabee_> eeeee: kill deen pls :P
23:16 <@heinrich5991> ddwar is a blocker mod by eeeee AFAIK
23:16 <@deen> ok
23:16 <@deen> right, i remember reading that in his git repo
23:17 <@deen> when I stole the 64 player code
23:17 < Gabee_> deen: lol
23:17 < Gabee_> deen: so it's ddnet
23:18 < Gabee_> with points system for block
23:18 < Gabee_> + anti chatblock
23:40 <@deen> I believe this can be abused to DoS a server easily, can someone confirm? https://github.com/ddnet/ddnet/blob/master/src/engine/shared/network_server.cpp#L599
23:41 <@deen> Basically if you send an invalid ctrl packet to a server, the server returns its Recv function, does skipping all other packets for this tick.
23:41 <@deen> So if you spam the server with invalid ctrl packets regular packets will get read much slower than they could be
23:42 <@deen> I believe it should just be a "continue" instead of "return 0"
23:44 <@deen> oh, that's new code from EastByte, thought that's something old. Vanilla teeworlds 0.7 has something very similar though
23:45 < ddnet-commits> [ddnet] def- pushed 1 new commit to master: https://git.io/vrlqa
23:45 < ddnet-commits> ddnet/master 1645310 def: Keep reading packets after invalid control packet