| 02:33:12 | vogon_jeltz | is back (gone 05:32:00) |
| 02:39:02 | Zig | hey |
| 02:47:37 | HEMI | moves Zig |
| 03:16:45 | Zig | someone talk |
| 03:16:48 | Zig | someone go on ducati |
| 06:06:50 | vogon_jeltz | is busying himself by wrapping a lot of the SDL functions into the class SDLWindow |
| 06:07:07 | vogon_jeltz | ... which, in turn, will make it assloads easier to finish the port :) |
| 07:41:48 | vogon_jeltz | is away: taking a break from coding |
| 08:19:42 | captain_proton | success! |
| 08:21:08 | captain_proton | is happy |
| 08:26:06 | captain_proton | linux without a distro |
| 08:28:28 | captain_proton | time to sacrifice my uptime and play a little alice :) |
| 09:01:59 | BZFlag | wants to get bzflag running on the 6bone ;-) |
| 12:37:10 | Asms | help |
| 12:37:50 | I_KalamBa | ? |
| 12:38:21 | Asms | can I play bzflag behing a isa server? |
| 16:44:58 | vogon_jeltz | is back (gone 09:03:06) |
| 20:34:54 | roadrunner | hey BZFlag! |
| 20:50:40 | BZFlag | roadrunner: heya. |
| 20:50:47 | BZFlag | roadrunner: got the server running yet? |
| 20:51:21 | BZFlag | I don't have a redhat box to build on. I need one with standard mesa libs to build and release the redhat rpm on. |
| 20:51:49 | vogon_jeltz | hehe |
| 20:52:31 | vogon_jeltz | is away: restarting X yet again |
| 20:55:12 | | i haven't a clue, bzflag |
| 20:55:12 | BZFlag | ibot: build a redhat rpm for me k? |
| 21:00:27 | roadrunner | doh |
| 21:03:49 | roadrunner | isp problems vogon_jeltz? |
| 21:03:58 | vogon_jeltz | no |
| 21:04:05 | vogon_jeltz | video driver problems |
| 21:04:08 | roadrunner | ahh |
| 21:04:09 | roadrunner | fun |
| 22:32:34 | Chestal | BZFlag: alive? |
| 22:33:58 | roadrunner | hrm |
| 22:34:41 | roadrunner | looks like someone else saw the bug i'm seeing |
| 22:35:15 | Chestal | fg |
| 22:35:16 | Chestal | oops |
| 22:35:45 | Chestal | I haven't tried E4 yet |
| 22:35:57 | roadrunner | i've been seeing server/client comm issues |
| 22:36:13 | roadrunner | someone jsut posted to the bugtrack about it |
| 22:36:46 | Chestal | you can pick up flags and everythign, just donot see other tanks? |
| 22:36:53 | Chestal | you get 'not responding' vor them? |
| 22:36:55 | Chestal | for, even |
| 22:37:06 | roadrunner | yep |
| 22:37:17 | roadrunner | any ideas? |
| 22:37:21 | Chestal | you're usign UDP I suppose? |
| 22:37:25 | roadrunner | tried both |
| 22:37:31 | roadrunner | and its on an internal LAN |
| 22:37:35 | roadrunner | switched 100BT |
| 22:37:45 | Chestal | that's strange. Looks like the client<->client messages get messed up |
| 22:38:05 | roadrunner | do the clients talk to each other too? |
| 22:38:12 | Chestal | is there any E4 server otu there which is working for you? |
| 22:38:25 | roadrunner | i haven't seen one...but i haven't tried very hard either |
| 22:38:42 | Chestal | I'm still at E3+some personal patches |
| 22:38:49 | roadrunner | most of the servers that have people on them require udp, and that doesn't work w/my firewall |
| 22:38:54 | Chestal | will probably upgrade once I'll find the time |
| 22:38:59 | roadrunner | hehe |
| 22:39:50 | Chestal | well, the server acts as an realy for the client<->client communication |
| 22:39:56 | Chestal | relay |
| 22:40:07 | roadrunner | right...ok |
| 22:40:24 | Chestal | the game can use dierct client<->client group communication (UDP multicasts) in theory |
| 22:40:43 | roadrunner | i wonder if it's trying to do that, but failing |
| 22:41:11 | Chestal | when you connect to the server, what is it saying? |
| 22:41:15 | roadrunner | erm... |
| 22:41:16 | roadrunner | jas |
| 22:41:41 | BZFlag | Chestal: now, yeah |
| 22:42:00 | roadrunner | using multcast relay |
| 22:42:06 | roadrunner | multicast even |
| 22:42:11 | roadrunner | maybe that's the problem |
| 22:42:17 | Chestal | BZFlag: did you ever see a player going to 'not responding' state, but coudl still see its shots? |
| 22:42:20 | BZFlag | roadrunner: hmm... multicast must be broken. |
| 22:42:28 | roadrunner | how can i shut it off? |
| 22:43:07 | BZFlag | what happens if you connect to localhost with some bots? |
| 22:43:22 | BZFlag | I don't think localhost can do multicast. |
| 22:44:00 | roadrunner | hrm..i can try... |
| 22:44:00 | Chestal | BZFlag: I think this is strange. 'not responding' seems to mean the I do not receive any UDP packets fromt he player. But then I shouldnt see his shots either? |
| 22:45:19 | roadrunner | hrmm...how do i start solo game? |
| 22:45:30 | BZFlag | roadrunner: -solo <n> |
| 22:45:36 | BZFlag | on the client. |
| 22:45:40 | roadrunner | and it should just start? |
| 22:45:51 | BZFlag | that will add bots when you join. |
| 22:46:01 | roadrunner | so i still need a server running? |
| 22:46:05 | Chestal | roadrunner: what operatign system are you on anyway? |
| 22:46:07 | BZFlag | you can start the server beforehand in another window or from inside the game. |
| 22:46:20 | BZFlag | redhat linux I presume |
| 22:46:22 | roadrunner | Chestal: have seen this on FreeBSD, Linux (RH7.2 and Gentoo) and Windows |
| 22:46:50 | Chestal | BZFlag: what do you think about the scenario I described? I have seen this several times |
| 22:47:05 | roadrunner | k...trying |
| 22:47:16 | BZFlag | Chestal: I've seen it while developing e5 but that was due to my bugs. |
| 22:47:31 | Chestal | BZFlag: I see it with e3 server and e2 clients |
| 22:47:47 | roadrunner | i get that with solo as well |
| 22:47:51 | roadrunner | i can see where the bots start |
| 22:47:53 | Chestal | BZFlag: I guess I have seen it soem months ago already |
| 22:47:55 | roadrunner | but they look paused |
| 22:48:01 | BZFlag | hmmm. we need to track that down then. |
| 22:48:03 | roadrunner | and i can definitely see their shots |
| 22:48:27 | Chestal | hmm, so MsgUpdate gets lost, but MsgShotBegin doesn't |
| 22:48:37 | roadrunner | BZFlag: someone submitted a bug with it whilst using teh precompiled windows binaries |
| 22:48:55 | BZFlag | yep. this is probably an issue with the playerid mangling. |
| 22:49:10 | BZFlag | in other words it's my fault. =( |
| 22:49:13 | roadrunner | :( |
| 22:49:25 | roadrunner | let me know if i can help |
| 22:49:43 | BZFlag | when I get a minute I think I'll hack out all the multicast stuff. |
| 22:49:44 | Chestal | BZFlag: but it worked when we tested it on your server soem time ago, hmm. That ws pre-release-code though |
| 22:50:08 | BZFlag | Chestal: yes, but that would have disabled multicast. |
| 22:50:10 | roadrunner | it may be that the ether-cards i'm using don't support multicast? |
| 22:50:32 | BZFlag | no, I think without multicast should be fine. |
| 22:50:42 | BZFlag | it's with multicast that's the problem. |
| 22:50:43 | Chestal | roadrunner: in most cases, multicast cannot be used successfully |
| 22:50:55 | roadrunner | so do we have a flag to disable that from the server or client side? |
| 22:50:56 | Chestal | roadrunner: most routers are too dumb or not configured correctly |
| 22:51:18 | BZFlag | I'm getting fed up with this stuff. I think it's time to break the old clients and servers and just ship a working 1.8 |
| 22:51:35 | roadrunner | lol |
| 22:51:46 | roadrunner | fine by me... :) |
| 22:51:48 | BZFlag | rip out the multicast, fix the broken portions of the protocol add ipv6 support etc. |
| 22:52:03 | roadrunner | fix colision detection, add ramps |
| 22:52:16 | Chestal | I also think removing multicast is best. In theory multicast is fine, but it breaks too often |
| 22:52:17 | roadrunner | add tank dameage... ;) |
| 22:52:21 | BZFlag | heh. at least add protocol support for it so we can do it later. |
| 22:52:35 | roadrunner | wants tank damage |
| 22:52:54 | BZFlag | hears roadrunner but is not sure that would be good for gameplay. |
| 22:53:18 | BZFlag | the one shot = kill makes the game simple to learn/understand etc. |
| 22:53:22 | Chestal | sadly, lag has becoem worse on my server durign the last weeks. It's not my fault, obviously the backbone interconnections to other ISPs have becoem worse |
| 22:53:35 | Chestal | I am agaisnt major changes in gameplay |
| 22:53:40 | BZFlag | me too. |
| 22:53:46 | Chestal | if some are introduced, make them optional |
| 22:53:49 | roadrunner | i'd like to see it as an option |
| 22:53:49 | roadrunner | right |
| 22:53:52 | BZFlag | I would like to drop tcp entirely in 1.8 |
| 22:54:00 | roadrunner | hrm..that would suck for me |
| 22:54:07 | BZFlag | why? |
| 22:54:15 | Chestal | that reminds me, I would like an option to make the new radar changes optional, too. I think I mentioned that before |
| 22:54:16 | roadrunner | i guess i don't play outside our firewall much anyway |
| 22:54:26 | roadrunner | Chestal: that is optional |
| 22:54:37 | BZFlag | explain your fw issues. cause I am considering dropping tcp. |
| 22:54:49 | roadrunner | our firewall doesn't relay udp packets AFAIK |
| 22:54:56 | Chestal | roadrunner: nope, only the 'enhanced' part of it is. There are other changes liek coloured bullets and the new tank drawing routine |
| 22:55:02 | BZFlag | Chestal: try e5 and see if you like the radar there. |
| 22:55:13 | Chestal | BZFlag: I liek the E2 radar :-) |
| 22:55:18 | roadrunner | Chestal: ahh...oke |
| 22:55:29 | roadrunner | likes the e3 radar |
| 22:55:34 | BZFlag | roadrunner: hmmm. most any nat fw should. it needs to for things like dns etc. |
| 22:55:45 | BZFlag | roadrunner: can you "host rikers.org rikers.org" ? |
| 22:55:50 | roadrunner | BZFlag: i'm pretty sure...jas |
| 22:56:00 | BZFlag | that would be udp. |
| 22:56:08 | roadrunner | hrm...maybe it is |
| 22:56:12 | roadrunner | $ host rikers.org rikers.org |
| 22:56:12 | roadrunner | Using domain server: |
| 22:56:12 | roadrunner | Name: rikers.org |
| 22:56:12 | roadrunner | Address: 166.70.15.250 |
| 22:56:12 | roadrunner | Aliases: |
| 22:56:16 | roadrunner | but then it just stops |
| 22:56:19 | roadrunner | and sits there |
| 22:56:36 | Chestal | hmm, droppign TCP entirely? we would have to create some reliable protocol on top of UDP then? |
| 22:56:43 | roadrunner | Host not found, try again. |
| 22:56:54 | BZFlag | hmmm... that is bad. that probably means udp is dropped with no unreachable packet reply. not a Good Thing. |
| 22:57:03 | Chestal | I think the TCP connection is fine. Why drop it? |
| 22:57:33 | BZFlag | Chestal: yes. there would be an internal packet queue for the packets that require reliable transport. |
| 22:57:54 | BZFlag | then we have full control on lag issues. |
| 22:58:30 | Chestal | what I thought about once: with the current scheme UDP packets might be processed out of order |
| 22:58:30 | BZFlag | I suppose a fallback to tcp would be ok. |
| 22:58:32 | roadrunner | BZFlag: it does that with internal machines too |
| 22:58:40 | Chestal | not that big of an issue I suppose as it won't happen very often |
| 22:58:55 | BZFlag | Chestal: yes. I would add a sequence number to them and track lag with it. |
| 22:59:13 | BZFlag | roadrunner: on just "host rikers.org" ? |
| 22:59:14 | roadrunner | hrm...i guess it is still going outside the firewall to get the dns |
| 22:59:26 | roadrunner | $ host rikers.org |
| 22:59:26 | roadrunner | rikers.org has address 166.70.14.212 |
| 22:59:26 | roadrunner | rikers.org mail is handled (pri=10) by timr.dsl.xmission.com |
| 22:59:43 | BZFlag | that used udp to your internal nameserver. |
| 22:59:49 | roadrunner | ahh |
| 23:01:00 | BZFlag | ok, then a default of udp only and packet sequence numbers to handle it. and a fallback of tcp |
| 23:01:12 | BZFlag | that should work with multicast as well. |
| 23:01:50 | BZFlag | wishes he had a mbone/ipv6/6bone connected machine to develop this new stuff on. |
| 23:01:51 | Chestal | all msgs are small enough to fit into s single UDP packet I suppose? |
| 23:02:13 | BZFlag | yes, but this queue would combine packets which does not happen now. |
| 23:02:18 | Chestal | I think our university has mbone/ipv6, althoguh I am not sure |
| 23:02:29 | roadrunner | so does it seem odd to you taht 'bzfs -noudp' doesn't work? |
| 23:02:33 | BZFlag | and track the max allowed packet size for each connection. |
| 23:02:58 | BZFlag | so there would be far fewer packets than there are at present. |
| 23:03:22 | Chestal | do you think that many packets could be combined? |
| 23:03:36 | BZFlag | the server and client would throttle back multiple PlayerUpdate etc type msgs to handle bandwidth. |
| 23:04:01 | BZFlag | Chestal: some would be combined, some would be discarded. |
| 23:04:54 | BZFlag | if lag is .5 seconds, there is not much point in sending PlayerUpdates every .1 seconds. slow them down and decrease the lag. |
| 23:05:11 | BZFlag | even more so for large (say 1.5 sec) lag conenctions. |
| 23:05:33 | Chestal | but large lag does not necessarily mean that the lag is caused by congestion |
| 23:05:42 | BZFlag | I have access to some working code from SubSpace which handles hundreds of players on the same game. |
| 23:05:44 | Chestal | 500ms might be a normal round trip pign time |
| 23:06:06 | Chestal | and you cannot do anythign about the network latency |
| 23:06:20 | BZFlag | the point is the currect game does not know what the lag is well enough to handle changing what is sent. |
| 23:06:30 | BZFlag | er current |
| 23:07:13 | Chestal | I can imagine that dropping messages could be a good idea for people with low bandwidth connection (like modems) |
| 23:07:44 | Chestal | but dropping messages means less accuracy |
| 23:07:49 | BZFlag | that's what subspace does. and it has its own reliable/throttling transport on top. |
| 23:08:41 | Chestal | bzflag feels great if everyone has a sevre rroudn trip time of less than 100ms |
| 23:08:42 | BZFlag | if a 200fps player is playing a 20fps player, the 200fps player should get throttled to no more than 20 Updates/sec |
| 23:08:55 | BZFlag | that kind of thing. |
| 23:09:13 | Chestal | well, ok, but updates are only sent when needed anyway (deadreckoning stuff) |
| 23:09:35 | BZFlag | now, the 200fps player may well continue to send 200 Updates/sec, the server will know to throw away the irrelevant ones. |
| 23:09:49 | BZFlag | Chestal: yes, I'm only talking worst case here. |
| 23:09:49 | vogon_jeltz | Chestal: those checks are done every frame |
| 23:10:02 | vogon_jeltz | Chestal: and most of them require an update |
| 23:10:23 | Chestal | vogon_jeltz: I don't believe that that many MsgUpdate apckets are sent |
| 23:10:27 | BZFlag | by all means, the deadrecking code will stay there! |
| 23:10:30 | Chestal | vogon: I haven't looked that closely on that code though |
| 23:10:54 | Chestal | if you do not touc hyour mouse, there will be very few packets |
| 23:10:54 | BZFlag | recking... hehe. you knew what I meant. |
| 23:10:57 | vogon_jeltz | Chestal: they're sent every time position and/or velocity differ by more than about .1 meters or .1 m/s |
| 23:11:01 | vogon_jeltz | Chestal: IIRC |
| 23:11:15 | Chestal | vogon_jeltz: which is a godo thing, accuracy is very improtant to bz |
| 23:11:19 | vogon_jeltz | Chestal: they're also sent at least once every second |
| 23:11:37 | Chestal | vogon: if a player makes a quci evasive maneuver I want to see that |
| 23:11:45 | vogon_jeltz | BZFlag: we need to remove the no-MsgUpdatePlayer exploit in 1.8 |
| 23:12:10 | BZFlag | vogon_jeltz: you mean the msgkill? or what? |
| 23:12:13 | vogon_jeltz | Chestal: yeah, and so do I, but a lot of times, we're sending messages that aren't really needed |
| 23:12:36 | Chestal | I really wonder whether internet latency will go down in the future |
| 23:12:36 | BZFlag | particularly for a slow connect or fps player. |
| 23:12:40 | vogon_jeltz | BZFlag: by stopping the emission of MsgUpdatePlayer, one can drive around while black-capped and act as a normal player |
| 23:12:51 | Chestal | but I fear that msot is optimized at bandwidth |
| 23:13:01 | Chestal | latency doesn't seem to be a priority for most people which is a pity |
| 23:13:20 | BZFlag | vogon_jeltz: ahh. yes. the server should track player state and only allow/forward the packets that are valid. else kick the player. ;-) |
| 23:13:27 | vogon_jeltz | BZFlag: yeah... :) |
| 23:13:45 | vogon_jeltz | BZFlag: as is, I kick any players who are not responding and are either killed or kill someone else |
| 23:14:01 | Chestal | vogon: I am not sure that this is cheating in most cases |
| 23:14:06 | vogon_jeltz | Chestal: it is |
| 23:14:18 | Chestal | vogon: it seems to be a bug / network problems, althoguh I dont quite understand it |
| 23:14:18 | vogon_jeltz | Chestal: MsgKill will not be emitted by a player who's truly not responding |
| 23:14:30 | vogon_jeltz | Chestal: and neither will MsgShotUpdate |
| 23:14:35 | roadrunner | vogon_jeltz: that sounds like what i'm seeing |
| 23:14:43 | Chestal | vogon: MsgShotBegin shouldn't be, I agree |
| 23:14:45 | vogon_jeltz | roadrunner: it's either cheating or bad firewall setup |
| 23:14:53 | roadrunner | it's internal to my network |
| 23:14:57 | vogon_jeltz | heh |
| 23:15:05 | roadrunner | the only thing between clients is an HP switch |
| 23:15:12 | roadrunner | and i compiled the code |
| 23:15:14 | vogon_jeltz | *shrug* |
| 23:15:23 | roadrunner | certainly doesn't want to cheat against himself |
| 23:15:42 | Chestal | vogon: MsgKill will only kill the cheater himself :-) |
| 23:15:48 | Chestal | MsgKilled, that is |
| 23:15:50 | vogon_jeltz | Chestal: even if they're not cheating, there is still a problem with their network connectivity, and a swift boot will not worsen it |
| 23:16:03 | roadrunner | i'm also not seeing any client get kicked |
| 23:16:18 | vogon_jeltz | Chestal: yes, I know, but cheaters often don't hack the client to surpress MsgKilled |
| 23:16:18 | Chestal | vogon: I have seen this behaviour several times (just talked to BZFlag about it), and I do not understand it |
| 23:16:46 | Chestal | vogon: and at least some of these peopel werent cheating, and that's strange |
| 23:16:56 | vogon_jeltz | Chestal: so if they get hit by a bullet (randomly, as they can't be seen), they still emit MsgKilled |
| 23:17:00 | Chestal | vogon: as I can't imagine why MsgUpdate woudl be blocked, but not MsgShotBegin |
| 23:17:17 | roadrunner | the bullets are seen, the tanks aren't |
| 23:17:21 | vogon_jeltz | Chestal: precisely, so that's another issue we have to work on |
| 23:17:23 | vogon_jeltz | roadrunner: yep |
| 23:17:31 | roadrunner | hrm |
| 23:17:41 | vogon_jeltz | roadrunner: but if the invisible tank happens across someone else's bullet, they will most often emit MsgKilled |
| 23:17:53 | roadrunner | but all the tanks are invisible |
| 23:17:56 | roadrunner | hehehe |
| 23:18:07 | vogon_jeltz | roadrunner: oh, you're talking about the problems on your network |
| 23:18:16 | roadrunner | yah...sorry to confuse the issue |
| 23:18:21 | vogon_jeltz | roadrunner: np |
| 23:18:33 | roadrunner | needs to dig into the bzflag code |
| 23:18:43 | vogon_jeltz | BZFlag: I have most of the functions in BzfWindow wrapped with their SDL counterparts, btw |
| 23:18:55 | roadrunner | maybe i'll try the cvs code |
| 23:19:03 | Chestal | but I don't share mose people's optimism that soem cleverly designed protocol will help with the lag issue very much |
| 23:19:12 | vogon_jeltz | BZFlag: if I can continue my current pace, I should have an SDL port by the end of the week |
| 23:22:27 | roadrunner | compiles the CVS quick |
| 23:22:38 | vogon_jeltz | hehe |
| 23:24:59 | roadrunner | kind of a pain on bsd...but it works... |
| 23:25:00 | roadrunner | :) |
| 23:26:10 | roadrunner | odd... |
| 23:26:20 | roadrunner | says "Server is full or game is over" When i try toconnect |
| 23:26:22 | Chestal | ok, I'll go home |
| 23:26:35 | Chestal | cu |
| 23:26:38 | vogon_jeltz | bye |
| 23:26:45 | roadrunner | bye |
| 23:28:56 | roadrunner | i don't need to specify the data directory to the server too so i? |
| 23:31:26 | roadrunner | blech...oh well |
| 23:31:32 | roadrunner | guess the server is borken in cvs |
| 23:31:39 | roadrunner | c'est la vie |
| 23:31:46 | roadrunner | or something |
| 23:43:12 | BZFlag | roadrunner: cvs server requires two connects for some reason I have not tracked down. |
| 23:43:31 | BZFlag | also cvs REQUIRES the -pr option on the command line. |
| 23:43:39 | roadrunner | BZFlag: so just try again? |
| 23:44:22 | roadrunner | -pr? |
| 23:45:11 | roadrunner | oh, i see pr now |
| 23:46:37 | roadrunner | hrm...just FYI, seeign the same problem with CVS as i do with 1.7e4 |
| 23:46:43 | roadrunner | but you probably knew that would happen... |
| 23:46:50 | roadrunner | is away - heading home |