| 05:01:59 | vogon_jeltz | is away: TV |
| 07:20:49 | iLLf8d | hrm |
| 07:40:36 | MrApathyCream | listserver down? |
| 07:41:20 | iLLf8d | uhm |
| 07:41:24 | iLLf8d | wasn't a second ago |
| 07:50:18 | iLLf8d | woa |
| 07:51:03 | iLLf8d | caps server has 4 guys on it 3 had 1400+ms lag times |
| 07:51:05 | iLLf8d | heh |
| 07:51:20 | iLLf8d | he's probably running his icecast stream or something |
| 07:53:15 | captain_proton | my stream shouldn't take up that much bandwidth |
| 07:57:11 | MrApathyCream | working now... hm |
| 08:32:11 | iLLf8d | heh |
| 08:32:17 | iLLf8d | now 2 have 2600ms+ lags heh |
| 08:32:18 | iLLf8d | nice |
| 08:32:25 | captain_proton | thats bizzare |
| 08:32:28 | captain_proton | are you there? |
| 08:38:34 | iLLf8d | heh |
| 08:52:19 | iLLf8d | captain_proton: I sentem over |
| 08:52:29 | iLLf8d | if you find time to run them msg me =P |
| 08:52:33 | iLLf8d | please =) |
| 08:52:45 | iLLf8d | roids has traps in them |
| 08:52:59 | iLLf8d | and the non roids is very similar to current decks |
| 08:53:06 | iLLf8d | has an escape in the center |
| 08:53:08 | iLLf8d | and |
| 08:53:30 | iLLf8d | corner (or in this case) outter ring teleporters work in a wierd fashion |
| 08:53:33 | iLLf8d | you enter one |
| 08:53:38 | iLLf8d | and back into next |
| 08:53:39 | iLLf8d | etc... |
| 08:54:02 | iLLf8d | keep entering them reguardless of how your facing |
| 08:54:13 | iLLf8d | and it'll bring you to the next quardrant |
| 08:54:26 | captain_proton | nifty |
| 08:54:28 | iLLf8d | and roids has traps too |
| 08:54:46 | iLLf8d | likes roids better but thats personal pref |
| 09:17:59 | MrApathyCream | hmmm list server is down again for me :( |
| 09:23:17 | Chestal | good morning |
| 09:25:18 | MrApathyCream | hey |
| 09:26:36 | Chestal | I am just back from supervising a written test |
| 09:26:49 | Chestal | I wonder if I should do some sports now, hmm |
| 09:27:36 | Chestal | but maybe I'd rather drive home and eat something :-) |
| 11:05:55 | vogon_jeltz | is away: sleep -- possibly |
| 18:04:35 | Chestal | hi MAC |
| 18:10:48 | MrApathyCream | hey |
| 18:16:26 | Chestal | hmm, I would really like to work with the current CVS code base (bzfs), but I guess we need to establish backwards compatibility with regard to UDP / multicastign first, or I won't be able to use this version in the real world |
| 18:16:41 | Chestal | IIRC the UDP bug was fixed in the client |
| 18:21:22 | MrApathyCream | Meaning fix the UDP mess in the server, so clients don't HAVE to be e5 clients? |
| 18:21:51 | MrApathyCream | Any idea why all these changes were made? |
| 18:23:50 | Chestal | 'all these changes'? |
| 18:24:04 | MrApathyCream | multicast removal, etc |
| 18:24:43 | Chestal | well, I guess Tim was working on the non-reconnecting network code and somehow the multicasting stuff made problems |
| 18:25:09 | Chestal | multicasting had problems for some time, I guess since UDP unicastign was introduced |
| 18:25:18 | Chestal | so he tried to disable it altogether |
| 18:25:42 | Chestal | but because of the client bug (that has been there since UDP unicastign was introduced), UDP now didn't work right |
| 18:25:56 | Chestal | so bu gin client got fixed, and there we are now |
| 18:27:25 | MrApathyCream | ah |
| 18:27:50 | Chestal | I have made some experimental server patches, but these are against 1.7e3 |
| 18:28:36 | Chestal | they're probably not for production use but still I would like to upload them, but diffs against the old version can be a bit of a hassle |
| 18:29:06 | MrApathyCream | yes. |
| 18:29:31 | vogon_jeltz | is back (gone 07:23:35) |
| 18:29:42 | MrApathyCream | I was hoping to take smaller steps from 1.e75 to 1.8, so that more progress could be seen |
| 18:29:54 | MrApathyCream | Most are in the shake up the world camp, though. |
| 18:32:41 | Chestal | I favour small steps also, i.e. we could try and replace a whole module like the network protocol |
| 18:32:49 | Chestal | or change the rendering |
| 18:33:05 | Chestal | so we coudl take the current code and try out the new modules optionally |
| 18:33:21 | MrApathyCream | YES, I'd like to decide what we want to do in the network protocol, make the changes in a pre 1.8 build and move from there |
| 18:34:06 | Chestal | it should be easily possible to take the curernt network protocol and extend it a little, e.g. for explciit UDP lag measurement etc. |
| 18:34:17 | Chestal | experimenting with mesage throttling |
| 18:34:21 | MrApathyCream | I agree. |
| 18:34:26 | Chestal | some features would only work with new clients of course |
| 18:34:53 | MrApathyCream | I don't have a problem mandating new clients. People should upgrade anyway, I still can't believe there are e1 clients out there |
| 18:34:57 | Chestal | I think Tim is the main driving force behind new network code |
| 18:35:04 | MrApathyCream | yeah |
| 18:35:19 | Chestal | I am not so sure what to do best to enhance it, though |
| 18:35:54 | Chestal | I did some experimentign with server/client side message throttling and have packet loss / out of order stats for UDP packets |
| 18:36:17 | MrApathyCream | how unreliable is udp? I'm guessing not very |
| 18:36:19 | Chestal | I plan to add explicit lag measurement and dts next |
| 18:36:26 | Chestal | haven't lost a single packet yet |
| 18:36:33 | MrApathyCream | Yeah, that's what i thought |
| 18:36:33 | Chestal | of course I am sitting very close to the server |
| 18:36:41 | Chestal | the other person using my patch is lupinator usign DSL |
| 18:36:45 | MrApathyCream | Still, i think udp is very reliable |
| 18:36:46 | Chestal | not a packet lost so far, too |
| 18:37:01 | MrApathyCream | all this talk of building reliable packet transfer over udp is silly, imho |
| 18:37:04 | Chestal | I guess the only way fo rlosign packets would be some overloaded router out there |
| 18:37:25 | Chestal | Lupinator tried and did some downloads/uploads in the background |
| 18:37:34 | Chestal | no UDP packet was dropped, only his lag got real bad |
| 18:37:45 | MrApathyCream | yeah, that's what i would expect. |
| 18:37:57 | Chestal | this only related to client->server, though, not the other way |
| 18:38:03 | MrApathyCream | how bout out of order |
| 18:38:07 | Chestal | none |
| 18:38:13 | Chestal | I hope my code is working :-) |
| 18:38:18 | MrApathyCream | heh |
| 18:38:33 | Chestal | but I am quite sure it does, I did soem random sequence number manglign in my client and the server found that |
| 18:39:02 | Chestal | hmm, maybe it would be interesting to test it with someone from the US liek you? |
| 18:39:09 | Chestal | it's only some small chnages to the client |
| 18:39:12 | MrApathyCream | fine by me |
| 18:39:26 | MrApathyCream | dcc a diff? |
| 18:39:29 | Chestal | fine, we ned to patch Player.h and Player.cxx |
| 18:39:35 | Chestal | got no diff for this |
| 18:39:41 | MrApathyCream | ok... what needs changing |
| 18:39:45 | Chestal | but it's only 5 lines or so, we can do this manually |
| 18:39:59 | Chestal | Player.h line 30 |
| 18:40:07 | Chestal | +34 -> +34+4 |
| 18:40:13 | Chestal | we ad 4 bytes to msgplayerupdate |
| 18:40:32 | Chestal | at the end of class Player, about line 178, add a member variable: |
| 18:40:35 | MrApathyCream | done |
| 18:40:36 | Chestal | mutable unsigned short seqno; |
| 18:40:58 | Chestal | then in Player.cxx find the pack() method |
| 18:41:06 | Chestal | after |
| 18:41:07 | Chestal | buf = nboPackFloat(buf, angVel); |
| 18:41:08 | Chestal | add |
| 18:41:17 | Chestal | buf = nboPackShort(buf, 0x1729); //%%% |
| 18:41:17 | Chestal | buf = nboPackShort(buf, seqno); //%%% |
| 18:41:17 | Chestal | seqno=(seqno+1)%10000; |
| 18:41:26 | Chestal | the comments are not necessary :-) |
| 18:41:48 | MrApathyCream | ok, done |
| 18:41:52 | MrApathyCream | anything else? |
| 18:41:58 | Chestal | I guess you have to do a 'make clean' in src/bzflag before rebuilding because size of class Player changed and the dependencies won't get it right |
| 18:42:01 | Chestal | that should be it |
| 18:42:12 | Chestal | I just add a 2 byte seqauence number to the packet 0-9999 |
| 18:42:29 | Chestal | and before that a magic number so I can check if you're using the extension easily |
| 18:42:30 | MrApathyCream | win32 does build deps right, just a sec |
| 18:42:37 | MrApathyCream | what server? |
| 18:42:44 | Chestal | hm, ducati 5155 |
| 18:43:14 | Chestal | I guess I should mak epublic what my experiments revealed so far somewhere |
| 18:43:36 | MrApathyCream | i'm on ducati |
| 18:43:40 | Chestal | so others may benefit frommy experiences |
| 18:43:44 | Chestal | that was fast |
| 18:55:40 | Chestal | I'll try to remmeber the 10 last msgs or so |
| 18:55:56 | Chestal | so there will be no wrongly detected lost messages |
| 18:56:47 | MrApathyCream | In real world, seems like just keep track of last PlayerUpdate seqno, and ignore older ones that come in later |
| 18:56:47 | Chestal | hm, it's not as easy as I thoguht, I will just dump for now :-) |
| 18:57:10 | Chestal | a more advanced mechanism in the client would try to compensate dts |
| 18:58:16 | Chestal | actualyl I tried this, but it didn't work very well, butit was just a very quick shot at it |
| 18:59:32 | Chestal | ducati 5157 any time you like to |
| 18:59:52 | MrApathyCream | ok, hold on |
| 19:05:42 | Chestal | Player.cxx, isDeadReckoningWrong() |
| 19:05:52 | Chestal | after |
| 19:05:53 | Chestal | if (predictedPos[2] < 0.0f) return True; |
| 19:05:53 | Chestal | add |
| 19:06:11 | Chestal | if (TimeKeeper::getTick() - inputTime < 0.04) return False; |
| 19:06:23 | Chestal | this should throttle you playerupdate rate to 25 packets/s max. |
| 19:06:28 | MrApathyCream | that it? |
| 19:06:31 | Chestal | yes |
| 19:06:44 | Chestal | let's see if the number of ooo packets reduces |
| 19:06:45 | MrApathyCream | 5157? |
| 19:06:51 | Chestal | meet you there |
| 21:46:25 | vogon_jeltz | captain_proton: you here? |
| 21:50:02 | captain_proton | yes |
| 21:50:04 | captain_proton | kinda |
| 21:57:37 | vogon_jeltz | captain_proton: how many CDs does the Red Dwarf set fit on? |
| 21:58:04 | vogon_jeltz | I'm getting some CD-Rs today, and I need to know how many I need |
| 22:03:43 | captain_proton | vogon_jeltz: 19 total |
| 22:03:48 | captain_proton | 18 if you don't want the 'extras' |
| 22:06:49 | vogon_jeltz | heh |
| 22:07:08 | vogon_jeltz | I don't care for the extras :P |
| 22:07:12 | vogon_jeltz | address? |
| 22:07:19 | captain_proton | no? you don't want the arnold rimmer tribute song? |
| 22:07:23 | vogon_jeltz | no thanks :P |
| 22:24:25 | vogon_jeltz | dude, what's your address in meatspace? |
| 22:38:56 | Ruudi | hi |
| 22:40:01 | Ruudi | is there slackware bzflag pack? |
| 22:51:07 | captain_proton | i doubt it |
| 22:51:43 | Ruudi | i get /usr/bin/ld: cannot find -lstdc++ error |
| 22:51:47 | Ruudi | in compile |
| 22:51:53 | Ruudi | last source version |
| 22:59:01 | tvon | captain_proton: Decks seems to be down....in case this isnt intentional :) |
| 23:41:25 | captain_proton | yes i know |