01:26 < laxa> deen: there ?
01:26 < laxa> http://ddnet.tw/players/gB-46--32--124--32-LanuX/
01:26 < laxa> 502 bad gateway
01:26 < Nimda> DDNet Brazil went down!
01:27 < laxa> hum !
01:27 < Nimda> DDNet Brazil went back online!
01:29 <@deen> oops, fixed
01:30 < laxa> Just_Shotgun
01:30 < laxa> oop s:D
01:30 < laxa> http://youtu.be/a2vSsd87qiY
01:30 < Nimda> [YouTube] Title: [DDNet] Just_Shotgun - gB. | LanuX | Rating: 0.00/5.00 | Views: 0
01:31 <@deen> you still don't give me the line to add!
01:31 < laxa> hu ?
01:31 < laxa> I tried 60 fps this time, but dunno why fraps was filming only 50
01:31 < laxa> oh
01:31 < laxa> fuck
01:31 < laxa> give me the line again :D
01:32 <@deen> Kobra 3 Solo    RedFight        07:22.12        k4RCUCJAnEE
01:32 <@deen> tabs as delimiter
01:33 <@deen> i think the 60fps version isn't up yet
01:33 < laxa> Just_Shotgun	gB. | LanuX	1:22.44	a2vSsd87qiY
01:33 < laxa> and now
01:33 < laxa> fuck my IRC client
01:35 <@deen> np, thanks
01:37 <@deen> should be up: http://ddnet.tw/halloffame
01:38 <@deen> curious to see the 60fps version once it's on yt
01:43 < laxa> it's up now, but don't know where the sound is gone
01:43 <@deen> magic!
01:43 < laxa> wow, didn't knew it was written 60 fps
01:43 < laxa> so apparently it's good
01:43 < laxa> but upload and processing is twice longer :(
01:44 < laxa> ok, I guess I fucked up the sound recording
01:46 <@deen> good night
02:13 < Nimda> DDNet Brazil went down!
02:14 < Nimda> DDNet Brazil went back online!
10:02 < Nimda> DDNet South Africa went down!
10:03 < Nimda> DDNet South Africa went back online!
10:20 < Nimda> DDNet South Africa went down!
10:21 < Nimda> DDNet South Africa went back online!
11:01 < Nimda> DDNet South Africa went down!
11:02 < Nimda> DDNet South Africa went back online!
16:20 < ddnet-commits> [ddnet] def- pushed 1 new commit to DDRace64: http://git.io/g0SXnQ
16:20 < ddnet-commits> ddnet/DDRace64 a08ed87 def: Initial work on confusable utf8 name detection
16:28 <@deen> I'll do ascii confusable detection only for now
16:28 <@deen> i mean, ascii characters compared to all unicode
16:35 <@deen> pretty big list
16:39 <@deen> still necessary since the client with name faking will be made public in a few days
16:40 < coffee> Hello, What means "name faking" ?
16:41 <@deen> while someone is playing they join the same server
16:41 <@deen> and choose a name that looks very similar (the same) to the other players' name
16:41 <@deen> then you can't tell them apart in chat anymore and they write stupid stuff
16:41 <@deen> it's possible because some characters look the same, for example latin e and cyrillic e
16:42 < coffee> Ok thanks for the explanation
16:42 <@deen> This is how to properly prevent this problem: http://www.unicode.org/reports/tr39/
16:42 < coffee> But what do you plan to do ? Prevent UTF8 ?
16:42 < coffee> I mean just allowing ASCII characters :p ?
16:42 < coffee> oh okay.
16:42 <@deen> I plan to compare names by assuming confusables as identical
16:43 <@deen> so when there is a player deen (latin e) and someone tries to rename to deen (cyrillic e) the name will become (1)deen
16:43 <@deen> so people can still write their cyrillic names without problem
16:44 <@deen> The code is already there, just need to fill in the confusables now: http://git.io/g0SXnQ
16:44 < o_be_one> nice :D 
16:45 <@deen> someone is also working on flybot detection
16:45 <@deen> still a bit worried, because the bot can replicate every player's behaviour
16:48 < coffee> Do you will parse this file : http://www.unicode.org/Public/security/latest/confusables.txt  ?
16:49 <@deen> I'm converting it to C code right now^^
16:49 <@deen> at least the part i need
16:49 <@deen> something like
16:49 <@deen> grep -P ";\t00[0-7][0-9A-F] ;" confusables.txt | grep -P "\tMA\t" | sed -e "s/^\([0-9A-F]*\) ;\t\([0-9A-F]*\) ;.*/0x\2 0x\1/" | sort | ./convert.py
16:50 < coffee> Perl somewhere, perl everywhere
16:50 <@deen> I don't use perl^^
16:50 <@deen> and i wanted to do it without python, but i guess it's easier at this point^^
16:51 < coffee> Thanks for all, good luck ;à
16:56 <@deen> hm
16:56 <@deen> It's done but I'm not sure if we want that code^^
16:57 < coffee> Do you plan to associate that with the account system?
16:58 <@deen> nope
17:02 <@deen> cool, it seems to work
17:03 < o_be_one> flybot ? hookfly ??
17:04 <@deen> ?
17:04 <@deen> flybot, where you just levitate in the air
17:04 < o_be_one> its possible that oO ?
17:04 <@deen> it hooks up so quickly that you can't see it
17:04 < o_be_one> ah the carryfly !
17:04 < o_be_one> its just fast hookfly
17:05 <@deen> no
17:05 <@deen> that's not what I'm talking about
17:05 <@deen> i'm talking about a bot that makes a single tee fly, not hookfly
17:05 < o_be_one> without hook ?
17:05 <@deen> the bot hooks automatically
17:05 < o_be_one> ok its carryfly on sn0wflake
17:06 <@deen> i don't know these words
17:07 < o_be_one> and how is it a problem ? they can be precise on race ?
17:07 <@deen> they can finish very hard maps easily
17:08 < o_be_one> omg :(
17:09 < ddnet-commits> [ddnet] def- pushed 1 new commit to DDRace64: http://git.io/rVV5Og
17:09 < ddnet-commits> ddnet/DDRace64 92e4a5c def: Add ASCII confusable detection in names
17:15 < ddnet-commits> [ddnet] def- pushed 1 new commit to DDRace64: http://git.io/vD2dBg
17:15 < ddnet-commits> ddnet/DDRace64 434b757 def: Exactly wrong way around
17:17 < laxadedi> nice work deen 
17:21 <@deen> Test it please
17:21 < laxadedi> I am at work !
17:21 <@deen> ok
17:22 < laxadedi> Don't you think that server version should be different from client's version ?
17:22 < laxadedi> cause right now, you made a change, but version isn't changing
17:26 <@deen> hmmm
17:26 <@deen> Too many version numbers then, or not?
17:27 <@deen> I reload servers every few days anyway, so don't really need version
17:27 <@deen> and server for non-developers is released together with client
17:29 < laxadedi> Well I guess it's ok then
17:29 < laxadedi> I'll go back at home now :)
17:29 < laxadedi> but then version is more for package than server/client version
17:30 <@deen> exactly
17:30 <@deen> maybe i should just use git commit hash as version on server :P
17:30 < laxadedi> wow
17:30 < laxadedi> No :)
17:55 < Tobii> deen: Looks like it's working
17:55 <@deen> Tobii: =)
17:57 <@deen> sometimes, very rarely the fifo console stops working
17:57 <@deen> no idea why
18:29 < ddnet-commits> [ddnet] def- pushed 1 new commit to DDRace64: http://git.io/zT_lFQ
18:29 < ddnet-commits> ddnet/DDRace64 bfcc82f def: Fix demo saving
18:46 < Learath2> http://gyazo.com/4d4a5d16bd0c29c885a4b1d1cd359056
18:46 < Learath2> http://gyazo.com/175da8fdfa0b4eabf03c38027ec3f095
18:47 < Learath2> remaking autoupdate
19:07 <@deen> Learath2: nice
19:10 < ddnet-commits> [ddnet] def- pushed 1 new commit to DDRace64: http://git.io/2wcJwQ
19:10 < ddnet-commits> ddnet/DDRace64 d7e5a7b def: Whitespace cleanup
19:13 <@deen> ping 70 on GER, 120 on GER2
19:13 <@deen> =/
19:14 < Learath2> deen: do you update anything except the executable ?
19:14 <@deen> only EUR stays stable, EastByte ^^
19:14 <@deen> Learath2: http://update.ddnet.tw/ddnet.upd
19:14 <@deen> all files are updated
19:15 < Learath2> i have 80 ping to both GER and GER2 with my shitty 3g
19:15 < Learath2> deen: regardless of their contents ?
19:16 <@deen> Learath2: no, only the files that changed
19:16 < Learath2> and if user changes a file is it overwritten ?
19:16 <@deen> yes
19:16 <@EastByte> deen: I had another lagg on eur today... speaking of stable
19:16 <@deen> User file changes should home in config_directory
19:16 <@deen> the files there are preferred (i think)
19:17 <@deen> should happen*
19:17 <@deen> ok, so no server is stable
19:17 <@deen> except the old gerdf.ddnet.tw, which somehow still is running even though i haven't paid for a month
19:17 <@EastByte> haha :D
19:17 < Learath2> are they unstable because of constant attacks or unstable hosts ?
19:17 <@deen> he even said I can have the server back because the ddos attacks stopped..
19:18 <@deen> unstable because of Deutsche Telekom
19:18 <@deen> their network is overloaded in every diretion it seems
19:18 < Learath2> that should go away in some time
19:18 <@deen> Hah
19:18 <@EastByte> also unstable because of anti ddos solutions
19:18 <@deen> People have been reporting these problems for years
19:18 <@deen> @ Deutsche Telekom
19:18 <@EastByte> tm
19:19 < Learath2> then choose a host that doesnt go over Deutsche Telekom network
19:19 <@deen> nono, the users come from Deutsche Telekom network
19:19 < Learath2> oic
19:19 <@deen> it's the biggest internet provider in germany
19:19 < Learath2> how can they afford to have an overloaded network ????
19:19 < Learath2> monopoly ?
19:19 <@deen> they currently try to force their peering partners into paying for better peering
19:20 <@deen> so they don't upgrade their side of the peering
19:21 <@deen> well, the European Commision is currently investigating just that
19:21 <@deen> because it's probably illegal
19:21 < Learath2> deen: do little upstream providers buy from Deutsche Telekom ?
19:22 <@deen> i don't know
19:22 <@deen> oh no, they found that Deutsche Telekom isn't doing anything illegal
19:24 <@deen> EastByte: also, EUR has nice routing to some places
19:24 <@deen> ping 65 from Egypt to EUR, but 105 to GER
19:24 <@deen> (GER1 + GER2)
19:24 <@EastByte> hm yea
19:24 <@deen> Ahaha
19:25 <@deen> Cogent was the one who made the EU commision investigate Deutsche Telekom!
19:25 <@deen> Exactly the carrier that is causing problems with GER1
19:25 <@EastByte> lol
19:25 < Learath2> no idea why a new company doesnt pop up which corrects the mistakes of the other
19:26 <@deen> Learath2: because it's expensive to put new connections to millions of people
19:26 <@deen> Learath2: Deutsche Telekom was a former state company that is now private
19:26 <@deen> so they inherited the state network
19:26 < Learath2> eww the lines belong to Deutsche Telekom dont they :/
19:26 <@deen> exactly
19:26 <@deen> I'm not even at Deutsche Telekom
19:26 < Learath2> exact same situation in Turkey
19:27 <@deen> but it's their line so I still get routed through them
19:27 < Learath2> except a company invested shit ton of money in putting fiber all over turkey so we have 2 set of lines here
19:27 <@deen> nice
19:28 < Learath2> but they dont gib me fiber :(
19:28 <@deen> well, there are some other providers, but nothing as big as DT
19:28 < Learath2> deen: the other providers prolly use DT network too
19:28 <@deen> nono, cable providers and local fiber providers
19:29 <@deen> they have their own networks
19:29 <@deen> but i wouldn't even switch to a cable provider because the upload is terrible. 150 mbit/s down, 3 mbit/s up or something like that
19:29 < Learath2> monopoly laws should have prevented DT from getting that big
19:29 <@deen> they started out big
19:29 <@deen> and monopoly laws forced them to resell their network
19:30 < Learath2> deen: i have 0.6mbps upload
19:32 < Learath2> internet is a mess
19:33 < Learath2> the cables cant be shared. if everyone has their own network a mess of cables underground. if one company owns the cables monopoly
19:34 <@deen> Hm, someone just rebooted south africa
19:34 <@EastByte> they rebooted the whole continent?
19:36 <@EastByte> oh wait, it's a country
19:36 <@deen> ^^
19:36 <@deen> they rebooted the server
19:38 <@deen> If anyone wants to fix auto demo recording: http://forum.ddnet.tw/viewtopic.php?t=699&p=6200
20:44 <@EastByte> https://eastbit.net/public/videos/lagg.webm
20:45 <@EastByte> eur at its best
20:45 <@EastByte> deen: laggs look funny in demo player :D
20:45 <@deen> ouch =/
20:45 <@deen> !ddnetpeak
20:45 < Nimda> Current players on DDNet : 497
20:45 < Nimda> Current DDNet peak : 612 users online at 2014-10-26 20:15:01
20:45 <@deen> so, i think GER is the most stable server after all
20:45 <@deen> I'm now logging my pings to GER, GER2 and ddnet.tw
20:45 <@deen> can't ping EUR :P
20:45 <@deen> http://ddnet.tw/ping/
20:45 <@EastByte> :P
20:47 <@EastByte> deen: shouldn't we use some tw like protocol for pinging?
20:47 <@deen> yes we should
20:48 <@EastByte> we can't rely on icmp
20:48 <@deen> write a small python script to send SERVERBROWSE and measure time until response?
20:48 <@deen> or even c++
20:48 <@deen> in ddnet source
20:48 <@deen> I think I can do that
20:49 <@deen> more fun than fixing solo servers again
20:49 <@EastByte> do your best :D
20:49 <@EastByte> I'm busy homeworks
20:51 <@EastByte> deen: you said you are not at deutsche telekom?
20:51 <@EastByte> but you have similiar ping problems?
20:51 <@deen> I'm at 1&1, but they just resell Deutsche Telekom
20:52 <@EastByte> ah okay
20:52 <@deen> EastByte: I have ping problems on GER & GER2 all the time
20:52 <@deen> EUR is fine
20:52 <@deen> Probably just Frankfurt affected
20:52 <@deen> DECIX
20:52 <@EastByte> then I want to run such a ping statistic from home too
20:52 <@EastByte> so we can compare to ewe
20:52 <@deen> my scripts are ugly and copied together quickly^^
20:53 <@EastByte> nvm, you will make something new using the tw request/respnse right? :)
20:53 <@deen> wanted to use the same scripts and just drop in my new ping binary :P
20:53 <@EastByte> no problem for me :P just put it on github
21:15 <@deen> that was easy
21:16 < ddnet-commits> [ddnet] def- pushed 1 new commit to DDRace64: http://git.io/L-b1Nw
21:16 < ddnet-commits> ddnet/DDRace64 7ff73a2 def: Add ping tool
21:21 < ddnet-commits> [ddnet] def- pushed 1 new commit to DDRace64: http://git.io/PsoDZg
21:21 < ddnet-commits> ddnet/DDRace64 8fb0358 def: ping tool cleanup
21:23 <@EastByte> neat
21:29 <@deen> something is weird
21:29 <@deen> when I run ./ping localhost:8902 I get 0.9 ms
21:29 <@deen> when I run while true; do ./ping localhost:8902; sleep 1; done
21:29 <@deen> I get 0.05 ms
21:30 <@heinrich5991> deen: try both again
21:30 <@deen> the first one in the while loop is also slow sometimes, meh
21:33 <@deen> meh, I'll just do 2 pings and use the second one
21:33 <@deen> not like I'm trying to do scientific measurements :P
21:35 <@deen> Ah, I think it's a server issue
21:35 <@deen> When no one is on the server it just sleeps for long times
21:36 <@deen> and when it wakes up it has to do a lot of work before responding
21:36 <@deen> Yeah, issue disappears when I remove the wait
21:36 <@deen> hm, so all empty ddnet servers how a ping of 1 ms more
21:38 <@deen> have*
21:46 <@deen> ok, down to 50 µs now on localhost, looks better
21:47 <@deen> icmp ping is only 20 µs though
21:48 <@heinrich5991> deen: icmp ping is handled by the os, doesn't need to go through user space
21:48 <@EastByte> I guess a usermode process using select cannot respond as fast as the kernel speaks icmp
21:48 <@deen> nope, now it's down to 8 µs :P
21:49 <@EastByte> wut
21:49 <@deen> i think select doesn't wake up fast enough
21:49 <@deen> hmmm
21:49 <@EastByte> try poll/epoll then, never used it before
21:50 <@EastByte> but it doesn't really matter?^^
21:51 <@deen> ^^
21:52 <@deen> just a few microseconds, yeah
21:57 <@deen> still, it's always good to experiment a bit
21:58 <@deen> Now I get the pred on localhost down to 0 ms constantly
21:58 <@EastByte> do you even wait for a request? :D
22:01 <@deen> this should decrease everyone's pings by 10 ms. I'm wondering whether it causes any problems
22:01 <@EastByte> what exactly did you do?
22:02 <@deen> something weird, i think it's wrong^^
22:02 <@deen> sometimes you get small lags
22:02 <@EastByte> must be wrong if you can't explain it^^
22:17 <@deen> ok, can get the prediction down to 0 properly now
22:17 <@deen> but that's the intentional buffer
22:17 <@deen> pred always gets 10 ms added
22:31 < ddnet-commits> [ddnet] def- pushed 1 new commit to DDRace64: http://git.io/w5cgwQ
22:31 < ddnet-commits> ddnet/DDRace64 18315e8 def: Improve ping on empty server (serverbrowser)
22:32 < ddnet-commits> [ddnet] def- pushed 1 new commit to DDRace64: http://git.io/75f5Yw
22:32 < ddnet-commits> ddnet/DDRace64 e8cf3fb def: Newlines in ping tool
22:37 < ddnet-commits> [ddnet] def- pushed 1 new commit to DDRace64: http://git.io/Iuu7DA
22:37 < ddnet-commits> ddnet/DDRace64 bde77f0 def: Default port for ping tool
22:54 < Learath2> deen: you ever notice labels are misaligned in Menu buttons ?
22:55 <@deen> what menu buttons?
22:55 < Learath2> serverbrowser
22:56 < Learath2> deen: http://gyazo.com/32a50f57ac306f6a38cfbff63715f5fa
22:57 < Learath2> the pink boxes are centered on the box and are the length of text calculated by TextRender()->TextWidth()
22:57 < Learath2> I believe TextWidth returns the wrong width but no idea why
22:58 <@deen> Maybe because of the text outline?
23:00 < Learath2> but the width is calculated the exact same way the text is written
23:00 <@deen> weird
23:01 < Learath2> well have no idea how the outlines work so cant confirm that 
23:22 < ddnet-commits> [ddnet] def- pushed 1 new commit to DDRace64: http://git.io/hp6Ubw
23:22 < ddnet-commits> ddnet/DDRace64 7e89917 def: Rename ping -> twping
23:24 < laxa> !fr-en poils
23:24 < Nimda> Translation: pile
23:27 <@deen> EastByte: http://ddnet.tw/ping/
23:33 <@deen> and the sources: http://ddnet.tw/ping/ping.tar.gz
23:56 <@deen> laxa: something went wrong with your upload i think
23:57 <@deen> even at 1080p the video looks terrible
23:57 < laxa> hmmmm
23:57 <@deen> or is that because of 60fps?