02:04 < bridge> [teeworlds] @redix 02:04 < bridge> [teeworlds] it worked for 2 minutes :D then the master failed again 02:04 < bridge> [teeworlds] https://cdn.discordapp.com/attachments/490150878934990850/738547654371442788/unknown.png 02:04 < bridge> [teeworlds] :( 02:05 < bridge> [teeworlds] usually its 1 minute, idk what happened 02:05 < bridge> [teeworlds] 2 minutes and 10 seconds 02:06 < bridge> [teeworlds] only register7 of course again, 6 works 02:07 < bridge> [teeworlds] Okay. I decreased the heartbeat to 3 sec for testing and only tested with < 20 seconds. Maybe there is more that breaks =\ 02:08 < bridge> [teeworlds] Mhh 02:08 < bridge> [teeworlds] but once again, thanks for looking into it 15:23 < bridge> [teeworlds] @fokkonaut found the next issue. Guess what... uninitialized padding bytes in `NETADDR` when generating the tokens :D 15:23 < bridge> [teeworlds] Pushed the fix to the same branch 15:23 < bridge> [teeworlds] shall I apply both? 15:24 < bridge> [teeworlds] yes 16:46 < bridge> [teeworlds] Ill try it later 16:46 < bridge> [teeworlds] the stuff above is about the matter with the server losing contact to the masterservers right? 16:47 < bridge> [teeworlds] yes 16:47 < bridge> [teeworlds] when I compile with gcc 6.3 and logged into rcon I lost immediately contact to the masters 16:48 < bridge> [teeworlds] seems like it wasn't only me with the same problem 16:56 < bridge> [teeworlds] clang seems to work for a long time but also loses contact to masterserver after a while on debian9 19:30 < bridge> [teeworlds] can you try out the patch by redix? 19:30 < bridge> [teeworlds] i cant try it right now 19:30 < bridge> [teeworlds] but i would like to know if it works with it implemented and NOT without it