| *** Cobra_Fast_ is now away: vacant | 03:09 | |
| *** Cobra_Fast_ is back | 03:09 | |
| *** bryjen <bryjen!~bryjen@cpe-74-129-51-30.kya.res.rr.com> has joined #bzflag | 03:35 | |
| *** bryjen <bryjen!~bryjen@cpe-74-129-51-30.kya.res.rr.com> has left #bzflag | 03:37 | |
| *** bryjen <bryjen!~bryjen@cpe-74-129-51-30.kya.res.rr.com> has joined #bzflag | 03:46 | |
| *** Cobra_Fast_ is now away: vacant | 03:46 | |
| *** Cobra_Fast_ is back | 03:46 | |
| JunkoAkemi[m] | Will the API offer some version for server IPs? | 04:01 |
|---|---|---|
| *** I_Died_Once <I_Died_Once!~I_Died_On@c-73-184-170-223.hsd1.ga.comcast.net> has quit IRC (Ping timeout: 268 seconds) | 04:09 | |
| *** FastLizard4 is back | 05:10 | |
| *** FastLizard4 is now away: AWAY from keyboard | 05:40 | |
| TimRiker | what do you mean by that? you mean other than the bz protocol version? | 05:41 |
| BZNotify | bzflag: timriker synchronized pull request #323 "IPv6" (https://github.com/BZFlag-Dev/bzflag/pull/323) | 06:21 |
| BZNotify | bzflag: timriker synchronized pull request #323 "IPv6" (https://github.com/BZFlag-Dev/bzflag/pull/323) | 07:11 |
| *** bboles <bboles!~bboles@64.94.212.195> has quit IRC (Quit: ZNC 1.8.2 - https://znc.in) | 07:15 | |
| *** bboles <bboles!~bboles@64.94.212.195> has joined #bzflag | 07:15 | |
| BZNotify | bzflag: timriker synchronized pull request #323 "IPv6" (https://github.com/BZFlag-Dev/bzflag/pull/323) | 07:19 |
| BZNotify | bzflag: timriker synchronized pull request #323 "IPv6" (https://github.com/BZFlag-Dev/bzflag/pull/323) | 07:27 |
| *** Sgeo_ <Sgeo_!~Sgeo@user/sgeo> has quit IRC (Read error: Connection reset by peer) | 07:46 | |
| *** Cobra_Fast_ is now away: vacant | 08:44 | |
| *** Cobra_Fast_ is back | 09:01 | |
| *** FastLizard4 is back | 09:25 | |
| *** FastLizard4 is now away: AWAY from keyboard | 09:58 | |
| blast007 | JunkoAkemi[m]: the API I'm talking about is an HTTP REST API replacement for the server list (plus extra features) | 10:13 |
| *** FastLizard4 is now away: GONE - Screen Detached and Disconnected from IRC (I'm probably asleep, at work, or doing something in real life) | 10:16 | |
| *** Cobra_Fast_ is now away: vacant | 10:17 | |
| *** Cobra_Fast_ is back | 10:32 | |
| *** Cobra_Fast_ is now away: vacant | 10:54 | |
| BZNotify | master @ bzflag-bzfls: blast007 pushed 5 commits (https://github.com/BZFlag-Dev/bzflag-bzfls/compare/40795a6f0087...a342fbff1700): | 11:05 |
| BZNotify | master @ bzflag-bzfls: blast007 93271e: Better CLI detection (https://github.com/BZFlag-Dev/bzflag-bzfls/commit/93271ee861d22084896babb90debb61eb9558053) | 11:05 |
| BZNotify | master @ bzflag-bzfls: blast007 24dd30: Normalize the Content-Type headers (https://github.com/BZFlag-Dev/bzflag-bzfls/commit/24dd30d606431da203c5109c1c3ba57543c757e4) | 11:05 |
| BZNotify | master @ bzflag-bzfls: blast007 aa2906: Remove unused JSON code (https://github.com/BZFlag-Dev/bzflag-bzfls/commit/aa29064a2662e216e8b206bb66958f3644a67e1e) | 11:05 |
| BZNotify | master @ bzflag-bzfls: blast007 e2b714: Better detection for HTTPS enabled web servers not behind a load balancer (https://github.com/BZFlag-Dev/bzflag-bzfls/commit/e2b714c8375054f013441ddf2876c1fd321b5e01) | 11:05 |
| BZNotify | master @ bzflag-bzfls: blast007 a342fb: Directly set the $config['ourdomain'] value (https://github.com/BZFlag-Dev/bzflag-bzfls/commit/a342fbff170041635343b24ba94aa78ef0f78856) | 11:05 |
| *** I_Died_Once <I_Died_Once!~I_Died_On@c-73-184-170-223.hsd1.ga.comcast.net> has joined #bzflag | 11:32 | |
| *** Cobra_Fast_ is back | 12:01 | |
| *** Cobra_Fast_ is now away: vacant | 13:49 | |
| *** Cobra_Fast_ is back | 13:53 | |
| *** Cobra_Fast_ is now away: vacant | 13:53 | |
| *** Cobra_Fast_ is back | 14:04 | |
| *** Cobra_Fast_ is now away: vacant | 14:12 | |
| JunkoAkemi[m] | @blast007 It's related to that, I recall someone did years back some sort of patch/hack which sorted servers by lag via ping by IP. | 14:17 |
| JunkoAkemi[m] | Or something around those lines, IIRC. | 14:19 |
| blast007 | that was in 2.99.x's enhanced server list | 14:30 |
| blast007 | The server list doesn't need to send the server IP for that to work, though. The client should resolve the IP of the server's hostname. | 14:31 |
| JunkoAkemi[m] | Makes sense, thought it might be a minor optimization if sort by lag becomes an option. | 14:41 |
| blast007 | with our player count, sorting by lag isn't going to be too helpful | 14:43 |
| blast007 | *showing* the latency on the list still might be interesting | 14:44 |
| blast007 | if we did that, would we use some UDP message instead of TCP? | 14:49 |
| blast007 | that might not even work that well without some changes to the game loop on the server, since I think it sleeps longer when there are no players connected | 15:03 |
| JunkoAkemi[m] | ^I did actually find out the tick event is much quicker when more players are connected. | 15:04 |
| blast007 | a plugin can reduce the maximum sleep duration if it needs more frequent ticks, but yeah, a normal server will vary the loop delay | 15:06 |
| *** Sgeo <Sgeo!~Sgeo@user/sgeo> has joined #bzflag | 15:12 | |
| *** Cobra_Fast_ is back | 15:12 | |
| JunkoAkemi[m] | Latency on the list. Hmm, it's tricky since a quick ping could be really useful, but if we want something more like a "live" viewing connection, I'm not sure which is the most optimal solution. | 15:15 |
| blast007 | and if we do it with UDP, we need to make sure we're not creating a DDoS system that can amplify attacks :) | 15:16 |
| JunkoAkemi[m] | It depends if it is centralized or not which opens the door in that degree. ;) | 15:21 |
| blast007 | you can't centralize it | 15:21 |
| blast007 | the clients have to communicate directly with the game server if you want to show latency | 15:22 |
| JunkoAkemi[m] | TCP initial connection to UDP connection. (Like normal connections.) Could work for latency. | 15:26 |
| JunkoAkemi[m] | True, latency could only be shown by actual connection, but at the same time, an approximation would work fine in most cases. | 15:28 |
| JunkoAkemi[m] | I wonder if some form of "regions" + latency check would work okay. If player is in NA, check NA servers first, then check SA servers next, then check EU servers. | 15:31 |
| *** Cobra_Fast_ is now away: vacant | 15:56 | |
| *** Cobra_Fast_ is back | 15:56 | |
| *** Cobra_Fast_ is now away: vacant | 16:02 | |
| *** Cobra_Fast_ is back | 16:02 | |
| blast007 | https://www.bzexcess.com/geolist/ | 16:07 |
| blast007 | a problem with doing a TCP connection is that takes up one of the player slots | 16:12 |
| blast007 | so my thought was some sort of UDP package that would not be associated with a player | 16:13 |
| blast007 | MsgNoticeMeSenpai | 16:13 |
| TimRiker | A udp message that is larger than the udp reply should avoid it being used as an amplification attack. | 16:36 |
| TimRiker | we open a udp socket with the same port number as the tcp server connection. if we're pinging servers over udp, that's probably another random port and not the same one used for tcp, unless we do the reverse and open server connections from the same port as the ping udp random port. | 16:38 |
| TimRiker | You saw I added udp+ support in bzadmin on my fork, right? using it mostly for testing the udp code. I know why udp is not working on some connections. | 16:39 |
| TimRiker | Address comparisons are comparing everything at the moment. I plan to change that to just look at the ip and not the port or family. This will help fix ban lists etc. | 16:40 |
| TimRiker | when I get past that ::ffff:8.8.8.8 will equal 8.8.8.8 so ban lists can be ipv4 or ipv6 and a start/end address check should work. I suppose I should add common nat64 space as well as IPv4-Mapped space (RFC-5156). I don't currently have a v6 only network to test that on. | 16:42 |
| TimRiker | ie: ::ffff:8.8.8.8 == 64:ff9b::8.8.8.8 == 8.8.8.8 | 16:47 |
| blast007 | shouldn't the family be the same? | 17:01 |
| blast007 | (for the TCP and the UDP connection) | 17:02 |
| TimRiker | yes. the banlist will have things like 192.0.2.0/24 as a ban. is ::ffff:192.168.2.5 in that range? | 17:10 |
| TimRiker | yes. the banlist will have things like 192.0.2.0/24 as a ban. is ::ffff:192.0.2.5 in that range? | 17:10 |
| TimRiker | the incoming udp check is only checking ipv6 addresses at the moment. I will update it to check the family and then compare v4 address or v6 addresses. As you mention, the tcp and udp connection will be the same family. | 17:12 |
| blast007 | oh, was "I know why udp is not working on some connections" referring only the IPv6 fork? | 17:20 |
| TimRiker | yes. just in my fork. Sorry for the confusion. | 17:27 |
| blast007 | ok :) | 17:39 |
| *** FastLizard4 is back | 19:46 | |
| *** FastLizard4 is now away: AWAY from keyboard | 20:32 | |
Generated by irclog2html.py 2.17.3 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!