| 00:19.32 | *** join/#bzflag infobot (~infobot@rikers.org) |
| 00:19.32 | *** topic/#bzflag is https://BZFlag.org || http://ohloh.net/p/bzflag || https://bzflag.org/help || Channel Logs: http://infobot.rikers.org/%23bzflag/ || 2.4.10 released! https://forums.bzflag.org/viewtopic.php?f=8&t=19401 |
| 00:19.32 | *** mode/#bzflag [+v infobot] by ChanServ |
| 00:35.23 | alfa1 | maybe because Iuse ident? |
| 00:37.28 | alfa1 | https://pastebin.com/CqeKQzYZ |
| 00:37.45 | alfa1 | I wonder it too |
| 00:38.26 | alfa1 | or all here are registered which can help |
| 00:39.20 | alfa1 | hand up who is not registered :P |
| 00:41.34 | alfa1 | (a lot of registered) |
| 01:17.55 | *** join/#bzflag bertman (~bert@unaffiliated/bertman) |
| 01:59.07 | *** join/#bzflag bertman (~bert@unaffiliated/bertman) |
| 02:32.07 | *** join/#bzflag alfa1 (~alfa1@181.91.61.164) |
| 02:49.51 | *** join/#bzflag nadir (uid134094@gateway/web/irccloud.com/x-ngwwxbmafllmrrjo) |
| 04:02.38 | alfa1 | converting a double/float to const char* is making me crazy, any help? |
| 04:03.35 | alfa1 | casting is not working it seems |
| 04:04.00 | allejo | the heck are you trying to do? |
| 04:04.48 | alfa1 | in example to print server messages to know what tick value is the current one (double to const char*) |
| 04:05.08 | allejo | use bz_sendTextMessagef() |
| 04:05.11 | allejo | http://www.cplusplus.com/reference/cstdio/printf/ |
| 04:05.25 | alfa1 | TY :) |
| 04:05.41 | allejo | it respects those formats |
| 04:06.53 | allejo | bz_sendTextMessagef(..., "Value: %.2f", myFloat); for exampel that'll print out a float with 2 decimals |
| 04:07.12 | alfa1 | ok |
| 04:07.15 | alfa1 | ty |
| 04:07.30 | alfa1 | my second experience with plugins :) |
| 04:14.45 | alfa1 | nice, a tick in my machine is about a second |
| 04:15.10 | alfa1 | does that mean my machine is normal and has good health?:P |
| 04:15.19 | alfa1 | (joke) |
| 04:15.53 | alfa1 | oh no, not a tick but the integer part was a second |
| 04:16.14 | alfa1 | so my calculations will be fine |
| 04:33.08 | alfa1 | well finally got turning tanks off in a slight way with Jeff method :) (now not commenting more to not be annoying) |
| 04:33.59 | alfa1 | (didn't copy it directly, just had a look; also allej-o's one) |
| 05:02.20 | JeffM | The tick is not constant |
| 05:02.43 | JeffM | Other tragic and plugins can adjust it |
| 05:02.49 | JeffM | Tragic |
| 05:02.57 | JeffM | Baby traffic |
| 05:03.11 | alfa1 | depends on every machine |
| 05:03.32 | alfa1 | but I guess it tries to approximate to a second right? |
| 05:03.37 | JeffM | No, it's not CPU bound |
| 05:03.39 | JeffM | Nope |
| 05:03.48 | alfa1 | not every tick but the integer part |
| 05:04.28 | alfa1 | not sure why there is an integer part tho |
| 05:04.34 | JeffM | You mean the clock? |
| 05:04.46 | JeffM | The clock is in seconds iirc |
| 05:05.12 | alfa1 | I used bz_sendTextMessagef(BZ_SERVER, BZ_ALLUSERS, "Tick: %.2f", tick); |
| 05:05.33 | JeffM | The tick happens when there is nothing else happening, and as often as the fastest plugin |
| 05:05.45 | alfa1 | double tick=data->eventTime; |
| 05:05.49 | JeffM | It's once per main loop of bzfs |
| 05:06.13 | JeffM | Yeah that's the clock time. |
| 05:06.37 | alfa1 | it is the only available variable into the tick event |
| 05:06.58 | alfa1 | or given by it to use |
| 05:07.32 | alfa1 | / Data (double) eventTime - Local Server time of the event (in seconds) |
| 05:07.37 | alfa1 | /* |
| 05:07.37 | JeffM | Sure, but that isn't the only clock you can use, you can call other time functions |
| 05:07.51 | alfa1 | double / |
| 05:08.47 | alfa1 | ah ok, then that was the clock |
| 05:09.25 | alfa1 | the ticks were very fast when I printed them |
| 05:10.35 | JeffM | Sure , a plugin needs cycles to be able to do useful work |
| 05:11.10 | alfa1 | your method was not so easy to make though :) |
| 05:11.40 | JeffM | It's pretty standard event driven programming |
| 05:11.46 | JeffM | Let me see your code |
| 05:12.15 | alfa1 | I have to find a way to convert double to const char* since now I am adjusting variables statically |
| 05:12.32 | alfa1 | all variables seem to be const char* |
| 05:12.41 | alfa1 | but calculations are numbers |
| 05:12.47 | JeffM | Prints |
| 05:12.59 | JeffM | That's basic c |
| 05:13.05 | alfa1 | yes, I found several print methods |
| 05:13.24 | alfa1 | I also forgot what we used on university somewhat |
| 05:13.48 | JeffM | API string has a format method too |
| 05:14.51 | alfa1 | what do you mean? |
| 05:15.44 | alfa1 | I have both h and cxx opened |
| 05:18.50 | JeffM | void bz_ApiString::format(const char* fmt, ...) |
| 05:20.20 | JeffM | So if you have an API that takes a string you can just call format on it. |
| 13:08.29 | macsforme | I'm going to let teamchillax.com expire unless someone wants it |
| 14:30.05 | JeffM | I let the server expire |
| 15:24.04 | JeffM | no clue if it works or not, but this is how I'd write alfa1's plug-in https://usercontent.irccloud-cdn.com/file/dEsn3Coj/StopAll.cpp |
| 16:01.56 | *** join/#bzflag Leppp (~Leppp@cpe-24-25-206-147.san.res.rr.com) |
| 16:22.07 | *** join/#bzflag Leppp (~Leppp@cpe-24-25-206-147.san.res.rr.com) |
| 16:49.25 | *** join/#bzflag alfa1 (~alfa1@181.92.200.72) |
| 16:52.54 | alfa1 | going to a general situation, cycles are not totally forbidden into plugins right? (I see them into those into source); also, plugins should be ideally as short and as best logically as possible (to be faster)? |
| 16:57.20 | alfa1 | question #2: is it posible to include something like bz_eAllowAutoPilotChangeEvent or bz_eAutoPilotChangeEvent from 2.99x to 2.4.x (and maybe some functions and related data like bz_isPlayerAutopilot) to help mainly handling autopilot bots? |
| 17:08.04 | alfa1 | question #3: under Player Lists and Records (https://wiki.bzflag.org/Functions_(API)) I see 2 groups; the second one talking of API functions: why were those API new functions created? to replace the first group ones? (JeffM mentioned yesterday something similar about string management) |
| 17:09.46 | JeffM | 1)yes, 2) no that has network messages associated with it, 3) just different ways of doing it |
| 17:15.05 | alfa1 | 2) I see Bz_eAutoPilotChangeEvent ideal in the sense to detect who is using autopilot and then treating it as a bot (killing it when necessary, etc) |
| 17:15.37 | JeffM | it may be, but it's not available in 2.4.x |
| 17:16.04 | alfa1 | I know, I am just asking if it would be possible to include it into 2.4.x |
| 17:16.13 | JeffM | no, because part of it is a message that the client sends |
| 17:16.20 | JeffM | and 2.4 clients don't sent that |
| 17:16.39 | JeffM | it's not JUST an api event, it's the network messages that trigger the event. |
| 17:17.20 | alfa1 | is it related to the fact that you all are moving player handling to server side? |
| 17:17.27 | JeffM | no |
| 17:17.35 | JeffM | it's a network message the client sends |
| 17:17.45 | JeffM | "I_am_now_a_robot" |
| 17:17.56 | alfa1 | has to do with the network protocol? |
| 17:18.23 | alfa1 | (seems something serious, not trivial) |
| 17:18.39 | alfa1 | (I thought it was trivial) |
| 17:19.08 | JeffM | yes |
| 17:19.17 | JeffM | it is not trival |
| 17:19.46 | JeffM | if it was it probalby would have been backported long ago |
| 17:20.11 | alfa1 | ok, and it is related to the fact you all want to move bots to server side in the future right? |
| 17:20.18 | JeffM | no |
| 17:24.22 | alfa1 | then nowadays it is not convenient to implement it or not so easy? |
| 17:24.47 | alfa1 | either* |
| 17:25.04 | JeffM | it would require a client change |
| 17:25.25 | alfa1 | ok |
| 17:26.28 | alfa1 | well, I am thinking on making a list of human players then (which usually are the less amount) and then do a search; but now I will think about question #1 |
| 17:27.04 | JeffM | loops are not bad, just don't sleep |
| 17:27.55 | alfa1 | ah ok |
| 17:28.32 | alfa1 | 3) do you recomend API ones better or pretty similar? |
| 17:28.46 | JeffM | it doesn't mattert |
| 17:29.02 | alfa1 | okay, TY |
| 18:12.59 | blast007 | I'd think the autopilot event would already be possible |
| 18:13.19 | blast007 | there is a network message sent when a client enables/disables autopilot |
| 18:13.27 | JeffM | if it's sent then yah |
| 18:13.56 | alfa1 | oh nice :) |
| 18:14.04 | JeffM | patches welcome |
| 18:14.09 | alfa1 | (working on workarounds) |
| 18:14.24 | alfa1 | "patches welcome" :) I know hehe |
| 18:14.45 | blast007 | 'course, that requires a pull request on github |
| 18:15.18 | JeffM | why not just make bzfs use github logins? |
| 18:16.09 | blast007 | I'm working on a new bzauthd :P |
| 18:17.28 | alfa1 | blast007: using what method/s if I may know? |
| 18:17.37 | blast007 | code |
| 18:17.48 | JeffM | probably some math too |
| 18:17.58 | alfa1 | ah 'd' for daemon |
| 18:18.24 | alfa1 | top secret :) no problem |
| 18:19.55 | blast007 | no, it's not top secret |
| 18:20.31 | blast007 | basically, I'm wanting to use public key cryptography to pass the sort of info that our 'CHECKTOKEN' response generates currently |
| 18:21.22 | blast007 | bzfs starts up, generates a keypair. Sends the public key to an auth daemon (of which there will be multiple) and also adds itself to the server list. |
| 18:22.09 | blast007 | (also sends groups and host/port to the auth daemon, that sort of info) |
| 18:22.18 | JeffM | see.. math :) |
| 18:22.36 | alfa1 | ah I can see the keys TOKENs in my log yes |
| 18:22.48 | JeffM | that's just a random number |
| 18:22.54 | JeffM | it can be so much more as a payload |
| 18:22.58 | alfa1 | (not encripted I meant) |
| 18:22.59 | blast007 | client wants to join a server, so it passes credentials and the host/port to an auth daemon. auth daemon checks credentials, and if valid, it encrypts a response using that game servers public key. gives that payload to the client, which the client can't read or modify. |
| 18:23.19 | JeffM | by the power of prime numbers! |
| 18:23.31 | blast007 | client passes that to the game server, game server decrypts it with its private key, and it then trusts the response since it can read it. |
| 18:23.54 | blast007 | so it elminates the need to tie tokens to IP addresses, since that payload is only valid for that server anyway. |
| 18:23.54 | JeffM | it's like a mini ssl connection |
| 18:24.12 | JeffM | and eliminates the double hit to the auth |
| 18:24.17 | blast007 | yup |
| 18:24.27 | alfa1 | wow, seems good :P |
| 18:24.35 | blast007 | and reduces the need for an immediate sync across all the daemons |
| 18:24.49 | JeffM | only reall mod you'd need to make the client is sending the desired server with the auth request |
| 18:24.56 | alfa1 | well, I know you were working on TLS lately, then all goes on that direction |
| 18:25.30 | blast007 | it'll have a proper format as well, like JSON, protobuf, or capnproto |
| 18:25.52 | JeffM | whahahahat..... an actual industry standard format?! |
| 18:25.54 | blast007 | auth daemon will be written in Rust |
| 18:26.42 | blast007 | :) |
| 19:10.55 | *** join/#bzflag nadir (uid134094@gateway/web/irccloud.com/x-twktwqirvsgtjamz) |
| 19:11.12 | *** join/#bzflag Zehra (~Zehra@unaffiliated/zehra) |
| 19:48.47 | JeffM | right language for the right job |
| 23:44.14 | *** join/#bzflag Kanone (~Zehra@unaffiliated/zehra) |
| 23:53.11 | *** join/#bzflag alfa1 (~alfa1@host251.200-45-200.telecom.net.ar) |