| 01:45:15 | Widget | hihihi, captain_proton. |
| 01:46:34 | captain_proton | hiya |
| 03:49:24 | Widget | hihihi, jpablo. |
| 03:49:30 | jpablo | what's up ? |
| 05:14:43 | MrApathyCream | wow look at all the new checkins! |
| 05:15:01 | vogon_jeltz | all none of them! |
| 05:23:45 | MrApathyCream | hmmm, shocker! |
| 05:24:08 | MrApathyCream | BZFlag going the way of the 8-track? |
| 05:24:17 | vogon_jeltz | hah. |
| 05:24:24 | MrApathyCream | Whoops, perhaps vogon is unfamiliar with 8-track |
| 05:24:35 | vogon_jeltz | what are you talking about? I have all the newest hits on 8-track! |
| 05:24:52 | MrApathyCream | A true renaissance man |
| 05:25:02 | vogon_jeltz | that Devo rules! I think they're the wave of the future! |
| 05:27:52 | MrApathyCream | whip-it good |
| 05:28:40 | vogon_jeltz | cranks up the Safety Dance |
| 05:28:48 | vogon_jeltz | different group, but it still rocks! |
| 05:32:35 | MrApathyCream | you can dance if you wanna |
| 05:32:37 | MrApathyCream | ah, ok |
| 05:33:10 | MrApathyCream | grampa is giggy with it |
| 05:33:36 | vogon_jeltz | o/~ You can dance if you want to / You can leave your friends behind / 'Cause your friends don't dance, and if they don't dance / then they ain't no friends of mine o/~ |
| 05:40:55 | Widget | Sigh. |
| 07:26:21 | MrApathyCream | Hmmm, How to? |
| 07:26:53 | MrApathyCream | Convert BZW object definition to general matrix for 1x1 object at -1,0,-1 to 1,0,1 |
| 07:26:55 | MrApathyCream | ? |
| 07:27:47 | MrApathyCream | Hmmm that 2x2... |
| 07:27:54 | MrApathyCream | maybe that's the problem... |
| 13:36:13 | Chestal | hmm |
| 14:05:44 | vogon_jeltz | is away: trying to sleep. |
| 18:59:56 | | niihau, Multitronic |
| 18:59:56 | Multitronic | Hi |
| 19:04:40 | Chestal | hi MUlti |
| 19:04:47 | Chestal | fiddles with e5 |
| 21:16:05 | Chestal | heh captain |
| 21:31:13 | captain_proton | heya tim |
| 21:31:31 | Chestal | hi Tim. are you alive? :-) |
| 21:31:58 | Chestal | I've been looking at e5/e4 UDP problem and ID patching bugs today |
| 21:32:16 | BZFlag | barely. ;-) had surgery on my wrist so I've been taking it easy. |
| 21:32:28 | BZFlag | cool. progress? |
| 21:32:31 | Chestal | yes |
| 21:32:42 | Chestal | I fixed one severe id bug |
| 21:32:57 | Chestal | and I think I have a workaround that allows e4 and e5 clients to work with e5 server |
| 21:33:03 | BZFlag | I planned to build an e2 that printed the id on UNKNOWN msgs. |
| 21:33:18 | | bonjour, [TeA|TrEE] |
| 21:33:18 | [TeA|TrEE] | hello |
| 21:33:24 | Chestal | there's only one problem left which I don't know how to solve best right now |
| 21:33:38 | Chestal | and then it needs some testing |
| 21:33:53 | Chestal | BZFlag: care to hear the details? |
| 21:34:12 | BZFlag | yes, please |
| 21:34:35 | Chestal | ok, the id patching bug: pwrite calls patchmessage if the old client is the target of the message |
| 21:35:01 | Chestal | problem with this: the message buffer might be used for another pwrite, e.g. when broadcastmessage is used |
| 21:35:17 | Chestal | so if the first target is an old player, id gets patched and further clients get patched id, this is bad |
| 21:35:28 | Chestal | I fixed this by patching, then sending, then reverting the patch |
| 21:36:07 | Chestal | pwrite does patch(new,old); pwrite2(...); patch(old,new); now |
| 21:36:21 | BZFlag | ahh! good point. |
| 21:36:41 | Chestal | the problem might have been introduced alogn with the shared message buffer optimization, I don't know |
| 21:37:10 | Chestal | however, this works ok now, I tested with one e4 and e5 client |
| 21:37:19 | BZFlag | well, it did not show up till we had new clients that knew thier ids. |
| 21:37:34 | Chestal | there's one problem with the id still: |
| 21:38:02 | Chestal | the server marks the client with knowid=true as soon as it recieved msgidack |
| 21:38:22 | BZFlag | and? |
| 21:38:31 | Chestal | but there are some mesasges sent from server->client after client uses new id and server receives msgidack |
| 21:39:03 | Chestal | one of the messages which causes problems are the informational messages, bzfs sends at the end of addplayer |
| 21:39:20 | Chestal | the ids will be patched there and the client thus gets the wrong id |
| 21:40:05 | BZFlag | that should just cause unknown instead of direct though, right? |
| 21:40:10 | Chestal | yes |
| 21:40:18 | BZFlag | no biggie then. |
| 21:40:35 | Chestal | I wonder if it's possible, that other messages could be sent from bzfs to the newly joined client, though |
| 21:40:53 | Chestal | for example if two players join at nearly the same time, it might happen? I don't know |
| 21:41:12 | | Howdy Bub |
| 21:41:12 | [TeA|TrEE] | hello ? |
| 21:41:13 | BZFlag | from other players, like a direct msg from another player, |
| 21:41:27 | [TeA|TrEE] | which othe url das the league have ? |
| 21:41:39 | Chestal | I guess the probability is very low |
| 21:41:42 | [TeA|TrEE] | http://www.bzflagleague.com/.................. |
| 21:41:52 | [TeA|TrEE] | what was it |
| 21:42:03 | BZFlag | http://bzflag.org/league/ |
| 21:42:06 | Chestal | because we'd only get a problem, if the newly joined player's id is part of a message |
| 21:42:18 | [TeA|TrEE] | i know that wone |
| 21:42:22 | [TeA|TrEE] | teh other wone |
| 21:42:39 | Chestal | tea: which other one? |
| 21:44:28 | Chestal | BZFlag: maybe we can live with it, testing will show. It's a bit udly to get the unknown-message |
| 21:45:09 | BZFlag | I'm not worried about it for now. so long as clients don't crash. |
| 21:45:25 | Chestal | BZFlag: I had the idea of dropping msgidack and using a modified MsgEnter which new clients would send. In that way, the server would know from the start that it is a new client. But this needs another ugly hack later and it would break compatability to current E% |
| 21:46:29 | Chestal | I'll run this e5 server on free-for-all ducati for a few days and see how it works |
| 21:47:37 | Chestal | the UDP problem: e4 clients cannot handle msgnetworkrelay correctly when sent during addplayer handling |
| 21:47:40 | BZFlag | in cvs now? |
| 21:47:46 | Chestal | not yet |
| 21:48:14 | Chestal | I'm reworkign the code and removing all those 50 ugly printfs :-) |
| 21:48:16 | BZFlag | udp, yeah. got a thought to work around that one? |
| 21:48:50 | Chestal | yes, I just moved the sending of MsgNetWorkRelay in server's addplayer() fter the last AddPlayer message |
| 21:49:13 | BZFlag | sounds good to me. |
| 21:49:56 | Chestal | it seems to work |
| 21:50:08 | Chestal | I had a hard tiem figuring this out, it's quite a mess, luckily it's all gone with 1.8 |
| 21:50:21 | BZFlag | If these are working ok, then it would be nice to get e6 out with these fixes. |
| 21:50:32 | BZFlag | I'd like to see the new HUD in the release as well. |
| 21:50:45 | BZFlag | the patch is looking better all the time. ;-) |
| 21:50:47 | Chestal | I'm not sure about the new HUD; haven't tried it yet |
| 21:50:58 | Chestal | it's in MrApathyCreams's f5 release righ tnow |
| 21:51:06 | Chestal | some people seem to liek it, some don't |
| 21:51:16 | Chestal | but there is a bunch of other patches whch should go in |
| 21:51:17 | BZFlag | check it out and tell me what you think. |
| 21:51:45 | BZFlag | agreed, send a summary of this and a list of patches you would like to see in e6 to the dev list? |
| 21:51:47 | Chestal | in fact all of my oatches thar are waiting on sf :-) |
| 21:52:48 | Chestal | some of these (like the observer patch) are a bit ugly, but IMHO that's not that big a problem, because 1.7 should be a dead-end anyway |
| 21:53:15 | Chestal | they work fine, just conceptually they would need bigger changes to client+server, but that's not possible withotu breaking compat. |
| 21:53:57 | Chestal | I'll put all of them into my e5 server here, so I can test them |
| 21:54:10 | Chestal | and with several peope using e5 or f5 clients right now, most of the code should be well-tested, too |
| 21:57:35 | captain_proton | perhaps enhanced hud could be an option? (yes, i know, eew) |
| 21:57:38 | Chestal | ok, I will checkin the fixes and then write to the list |
| 21:57:47 | BZFlag | sweet. |
| 21:57:54 | Chestal | captain: if it's possible to do that in a clean way, I would liek it |
| 21:58:16 | captain_proton | the best way i can think of is to just take all the stuff from the patch and conditionalize it |
| 21:58:18 | Chestal | I will give it a chance first, though |
| 21:58:43 | Chestal | ah, btw. what's your opinion on the new kill messages? |
| 21:58:53 | captain_proton | most of them are ok |
| 21:58:54 | Chestal | I wrote somethign about it to the bug that is now clsed |
| 21:59:02 | captain_proton | 'skewered by a super bullet' just doesn't work for me |
| 21:59:03 | Chestal | the problem is, that they cannot work ok in all cases |
| 21:59:22 | Chestal | because they rely on the shot object to be alive still |
| 21:59:31 | Chestal | the the shot is expired already, when MsgKill is received |
| 21:59:32 | captain_proton | yeah i noticed |
| 21:59:37 | Chestal | so the shot might be there or already gone |
| 21:59:42 | BZFlag | they revert if the shot has died. |
| 21:59:52 | Chestal | I'm not 100% sure if the shot might even be replaced by a new one |
| 22:00:13 | Chestal | with the curent code, I get by unknown about every 5 minutes |
| 22:00:44 | captain_proton | yep |
| 22:02:33 | BZFlag | I'm not attached to them, but they are kinda funny. |
| 22:02:51 | Chestal | it could be done better in 1.8 |
| 22:02:56 | BZFlag | agreed, |
| 22:02:59 | Chestal | I guess we might have a kill reason there |
| 22:03:27 | Chestal | it would be anice to have a different message for steamroller, but that's not possible this way |
| 22:03:34 | BZFlag | yeah, there will be a kill reason. |
| 22:05:10 | Chestal | I guess the unknowns could be fixed by not reverting to 'destroyed by unknown' if the shot is gone, but by using the old method instead |
| 22:05:45 | T-34 | Hello everybody ! |
| 22:05:51 | MrApathyCream | heya |
| 22:07:15 | Chestal | hi Cream |
| 22:07:51 | Chestal | someoen with e4 client care to join ducati 59000 for a moment? |
| 22:08:27 | MrApathyCream | hey |
| 22:08:35 | MrApathyCream | ok |
| 22:08:41 | Chestal | fne |
| 22:08:42 | Chestal | fine |
| 22:11:42 | MrApathyCream | one sec |
| 22:12:00 | Chestal | ok |
| 22:14:26 | BZFlag | Chestal: I hacked a simple patch workaround per your description to test pwrite |
| 22:14:28 | MrApathyCream | alrite i'm in |
| 22:14:32 | BZFlag | it's in cvs now. |
| 22:17:28 | Chestal | ah, fine. Looks like my first version, I used goto too |
| 22:17:43 | Chestal | then I thought peopel wouldn't like me to check in somethign using gotos and I wrote a wrapper function :-) |
| 22:17:52 | captain_proton | ahh |
| 22:17:58 | captain_proton | goto is evil |
| 22:19:53 | MrApathyCream | sorry i'm late, this fixes the e4 client domed problem? |
| 22:21:20 | Chestal | nope, but I'm commiting a fix for that right now |
| 22:21:35 | MrApathyCream | cool |
| 22:21:40 | Chestal | it fixes id patching bugs in mixed e4/e5 setups |
| 22:21:59 | MrApathyCream | nifty |
| 22:22:11 | MrApathyCream | are these the last e5/6 issues? |
| 22:23:20 | Chestal | I hope so, except for some patches we'd liek to integrate |
| 22:23:29 | Chestal | but e5 server needs some testing |
| 22:23:32 | | RCAM was last seen on #tuxboxproject 3 days, 18 hours, 49 minutes and 39 seconds ago, saying: i am off to an old far away land called bed, i will be in a state known as sleep. Call me and wake me up at 407-448-2555 and you shall face your destiny. Have a nice day. [Tue May 14 04:33:53 2002] |
| 22:23:32 | captain_proton | ibot, seen RCAM? |
| 22:24:00 | MrApathyCream | which patches? |
| 22:24:57 | Chestal | all of mine :-) and possible some others |
| 22:25:02 | captain_proton | haha |
| 22:25:06 | MrApathyCream | heh |
| 22:25:22 | Chestal | I'll check th epatch list and suggest something on the dev ml |
| 22:25:34 | MrApathyCream | captain_proton: I'm guessing Chestal won't like your hud patch! |
| 22:25:46 | captain_proton | prolly not |
| 22:25:50 | Chestal | Cream: some people usign the latest f5 have been complaining about it |
| 22:25:55 | captain_proton | btw new radar r0x0rz |
| 22:26:12 | Chestal | mostly about the fonts I think |
| 22:26:20 | Chestal | I haven't tried it yet |
| 22:26:46 | Chestal | and there's a bug with team mesages in f5, don't knwo if this is realetd to the HUD, it doesn't say [Team] anymore |
| 22:27:03 | captain_proton | my code doesn't change that |
| 22:27:16 | MrApathyCream | Yeah, also in f5 some messages are misplaced |
| 22:27:27 | MrApathyCream | like fps, flag help |
| 22:27:43 | MrApathyCream | captain_proton: are you sure? |
| 22:27:51 | captain_proton | yes |
| 22:27:59 | captain_proton | i don't change what gets written to the control panel |
| 22:28:15 | MrApathyCream | hmm, maybe a merge problem, cause it showed up after your patch... |
| 22:28:25 | Chestal | it works fine with e5 |
| 22:28:32 | Chestal | I did some changes to the mesage code recently |
| 22:28:59 | MrApathyCream | I applied some of your patches to chestal in f5.. so maybe... |
| 22:29:05 | Chestal | ok, ducati 5155 is now e5 server |
| 22:29:40 | Chestal | Cream: my other patches shouldn't affect this, my changes are in CVS, but it's a minor issue anyway and it's probably a merge problem |
| 22:30:08 | Chestal | hmm, double fixing in CVS |
| 22:31:39 | BZFlag | heh. I reverted. |
| 22:32:51 | Chestal | ok |
| 22:33:23 | Chestal | Cream: I never tried the message verbosity patch, is it good? |
| 22:33:36 | MrApathyCream | The idea is good, the implementation isn't so hot |
| 22:33:43 | MrApathyCream | I'd skip it for now |
| 22:35:05 | MrApathyCream | message verb, should be implemented as a bitset, where you (as a user) can pick any combination of message types |
| 22:35:24 | Chestal | I guess the deadreckoning patch is a good thing |
| 22:35:27 | MrApathyCream | as it is, there's a predefinied importance of each type of messages that you must live with |
| 22:38:37 | MrApathyCream | I'll try repatching to f5 and see if those 'anomolies' go away, later tonite |
| 22:39:07 | Chestal | ok |
| 22:40:16 | MrApathyCream | anyone interested in world caching? for e5 |
| 22:42:04 | MrApathyCream | guess not |
| 22:42:18 | Chestal | hmm, is there good code already? |
| 22:42:37 | Chestal | it's not a very high priority for me, but if it's workign, why not? |
| 22:42:42 | MrApathyCream | everythings there except unix file paths cache files |
| 22:42:53 | Chestal | and it's compatible with oler clients, too? |
| 22:43:00 | MrApathyCream | yes |
| 22:43:07 | Chestal | is there any change in the server? |
| 22:43:15 | Chestal | I guess there has to be? |
| 22:43:30 | MrApathyCream | a little |
| 22:44:47 | MrApathyCream | client asks for a md5 value, server responds, if match client doesn't ask for world |
| 22:45:04 | MrApathyCream | if server doesn't respond in some time, client just asks for world |
| 22:45:17 | MrApathyCream | old clients never ask, and never know better |
| 22:45:48 | MrApathyCream | I think it worth it, as long as someone is willing to fix the file code for linux side |
| 22:46:03 | captain_proton | i can take a look at it |
| 22:46:34 | captain_proton | what did we decide on for the cache directory in linux? |
| 22:46:38 | Chestal | where do you store it on windows? |
| 22:46:41 | captain_proton | $HOME/.bzcache ? |
| 22:47:02 | MrApathyCream | iirc name the directory, what the pref file is, and put the pref file in that directory... but i don't care |
| 22:47:12 | Chestal | hmm, I think we should really increase the DNS timeout from 3s |
| 22:47:53 | Chestal | and md5 encoded in filename? |
| 22:47:58 | MrApathyCream | right now it's in TEMP, but be glad to change it |
| 22:48:04 | MrApathyCream | md5 is the name of the file |
| 22:48:14 | Chestal | no automatic expiring yet? |
| 22:48:20 | MrApathyCream | no |
| 22:48:50 | Chestal | ah, I nearly forgot one importnat hing about e6: the documentation needs to be updated |
| 22:49:01 | MrApathyCream | heh |
| 22:49:07 | Chestal | the manpages are not uptodate, fo rexample the quickmessages are documented nowhere |
| 22:49:27 | Chestal | and there probably is some stuff which does not apply anymore |
| 22:49:36 | Chestal | for example, multicastign is disabled AFAIK |
| 22:49:43 | MrApathyCream | server commmands probably not updated too |
| 22:49:49 | Chestal | some of it is |
| 22:49:52 | Chestal | but not all |
| 22:50:04 | Chestal | someoen should read it all and check :-) |
| 22:51:46 | Chestal | ok, changign DNS timeout righ tnow, what is a good value? 5s? 10s? |
| 22:52:00 | MrApathyCream | world is 5s |
| 22:52:20 | Chestal | rmarco's patch (which isn't there :-) uses 10s |
| 22:52:31 | MrApathyCream | fine. |
| 22:53:02 | MrApathyCream | thanks chestal-> for the fixes, bbl.... |
| 22:53:04 | Chestal | hmm, I'll use 8s |
| 22:56:16 | Chestal | hmm, something is wrong |
| 22:56:57 | BZFlag | with no udp I still get seg faults. don't know why yet. |
| 22:57:10 | Chestal | uh? that#s strange |
| 22:57:40 | Chestal | my client just segfaulted on ducati ctf, I played with it a logj time this afternoon with no problems in my office, however, just checkign if my latest changes broke it |
| 22:59:11 | vogon_jeltz | is back (gone 08:53:24) |
| 23:03:08 | Chestal | looks like my debug change caused it, strange |
| 23:03:33 | Chestal | ah, I think I see, oh well |
| 23:21:43 | BZFlag | stomped on your msg patch. I did just about the same thing here. |
| 23:25:11 | Chestal | hmm :-) |
| 23:25:58 | Chestal | posting to the dev-ml right now |
| 23:51:55 | vogon_jeltz | hmm. |
| 23:52:19 | vogon_jeltz | I wonder if I should try my hand at implementing Chestal's "server locking" patch. |
| 23:53:50 | vogon_jeltz | Chestal: I suppose it could be done with passwording, but I'm not sure exactly what you have in mind. |
| 23:54:11 | Chestal | I'm not sure, either |
| 23:54:24 | vogon_jeltz | heh. |
| 23:54:25 | Chestal | my poin tis: I want to have this server available and running at all times |
| 23:54:31 | vogon_jeltz | that's always a good sign. |
| 23:54:43 | Chestal | so peopel can do league matches there without requriing me to start the server manually |
| 23:54:45 | vogon_jeltz | Chestal: but only allow people to connect to it during league matches? |
| 23:54:50 | Chestal | yes |
| 23:54:53 | vogon_jeltz | ahh. |
| 23:55:02 | vogon_jeltz | but you don't want general riff-raff getting in. |
| 23:55:06 | Chestal | peopel will need an admin/referee for league matches anyway, i.e. someone who knows the server password |
| 23:55:07 | vogon_jeltz | hmm. |
| 23:55:14 | vogon_jeltz | yeh. |
| 23:55:18 | Chestal | but I don't awnt the server to be used for normal playing all day |
| 23:55:29 | Chestal | it will use timed game anyway |
| 23:55:32 | vogon_jeltz | ponders a bit. |
| 23:55:39 | vogon_jeltz | I think I may have it. |
| 23:55:43 | Chestal | so after a game it could automatically fall in a game-over state |
| 23:55:54 | vogon_jeltz | the admin can unlock the server with the command /unlock. |
| 23:56:05 | Chestal | but the admin needs to be able to connect to it anyway, maybe only allow observers or something before unlocking? |
| 23:56:06 | vogon_jeltz | people joining as observers will be allowed in unconditionally. |
| 23:56:14 | Chestal | ah :-) |
| 23:56:15 | vogon_jeltz | precisely. ;) |
| 23:56:24 | vogon_jeltz | get out of my mind. :) |
| 23:56:35 | Chestal | there already is a game over state |
| 23:56:36 | vogon_jeltz | I'll probably get to it after I finish downloading OpenOffice. |
| 23:56:45 | Chestal | after a timed game, noone can join the server anymore |
| 23:56:56 | Chestal | until everyone has left, then the game is resetted |
| 23:57:16 | vogon_jeltz | yeah. |
| 23:57:29 | Chestal | it should probably autolock then |
| 23:57:36 | vogon_jeltz | personally thinks we should change that behavior, but that's a fix for 1.8. |
| 23:59:27 | BZFlag | agrees |