#bitfighter IRC Log

Index Search ←Prev date Next date→

IRC Log for 2012-04-18

Timestamps are in GMT/BST.

01:17:16mollie has joined
01:17:21molliehi
01:17:37molliehello
01:39:36mollie Quit (Quit: Page closed)
02:19:57mollie has joined
02:20:04molliehi
02:20:45kodaws Quit (Ping timeout: 265 seconds)
02:23:52mollie Quit (Client Quit)
03:55:48zoomber_mbp has joined
04:01:30zoomber_mbphi
04:01:45zoomber_mbphi raptor
04:01:52zoomber_mbpoh, i just said hi
04:01:53zoomber_mbpooohkay
04:02:16zoomber_mbphey raptor, did Koda ever get involved in the bit fighter code? curious
05:02:56sam686hi
05:15:08zoomber_mbp Quit (Quit: zoomber_mbp)
05:17:44zoomber_mbp has joined
05:42:11zoomber_mbpi guess sam is tired of hellos
05:42:12zoomber_mbphi
05:42:12sam686hello (this is an automatic message)
05:43:49zoomber_mbp Quit (Quit: zoomber_mbp)
05:55:04raptor Quit (Remote host closed the connection)
05:55:21raptor has joined
05:55:22ChanServ sets mode +o raptor
05:55:42sam686hi
05:55:49raptorhi
05:55:50sam686hello (this is an automatic message)
05:55:57sam686and bye (its like nidnight already)
05:55:59raptorhi automatic message
05:56:03raptorgood night!
05:56:07sam686and i put an auto message for my client
05:57:59raptorhi Zoomber: koda never did get directly involved other than answering some Mac questions I had (he is the Mac guy for his game, hedgewars)
06:01:30sam686 has left
06:28:25raptor Quit ()
07:00:49Watusimoto has joined
07:11:20Watusimoto Quit (Ping timeout: 252 seconds)
07:53:10watusimoto has joined
07:53:10ChanServ sets mode +o watusimoto
10:20:31Watusimoto_ has joined
11:47:11Watusimoto_ Quit (Ping timeout: 276 seconds)
15:05:44watusimoto Quit (Read error: Operation timed out)
15:43:39raptor has joined
15:43:39ChanServ sets mode +o raptor
15:52:06kodaws has joined
15:53:33Watusimoto has joined
15:54:02kodabbws has joined
15:57:48kodaws Quit (Ping timeout: 265 seconds)
16:17:16watusimoto1 has joined
16:20:15raptorsorry i ran off yesterday... I was the unwilling recipient of a migraine (a real one)
16:21:59raptorAnd I am looking for ways to alleviate them in the future, but no one seems to have any good ideas...
16:46:51watusimoto1vodaka or bullets. luckily I've never had one; hope you are feeling better.
16:47:02watusimoto1I'm heading home, so later!
16:51:56watusimoto1 Quit (Ping timeout: 265 seconds)
17:10:05raptorbye
18:02:16kodabbws Quit (Ping timeout: 246 seconds)
18:19:00LordDVG has joined
18:56:24kodaws has joined
19:01:05kodaws Quit (Ping timeout: 260 seconds)
19:06:57LordDVG Quit (Remote host closed the connection)
19:29:02kodaws has joined
20:22:27Watusimotohi
20:22:31raptorhi
20:27:34raptori need to try your latest changes..
20:29:02raptorFatal error running Lua code: . Possibly out of memory? Shutting down Bitfighter.
20:29:19raptorthat was with pressing Ctrl + ;
20:29:30raptorin the editor to bring up the arc plugin
20:29:50raptoroh oops, forgot to copy over new scripts..
20:30:30raptorah, that fixed it..
20:30:42sam686 has joined
20:30:42ChanServ sets mode +v sam686
20:32:35raptormazegen works nicely now
20:35:26sam686when i tried the latest changes last night, my level gen looks broken (especially for non-bitmatch) as if it is unable to read the arguments the level have
20:35:45raptoryes, i think Watusimoto made main() mandatory
20:37:38Watusimotono, main() is not mandatory, I don't think
20:37:42Watusimotothough
20:37:45Watusimotoand this is a big one
20:38:19raptori think it's mandatory now because of a 'return true' that was changed to 'false' somewhere...
20:38:30Watusimotoall code outside of main() is executed and cached and shared amongst all bots
20:38:37Watusimotoof that same script
20:38:56Watusimotoso if you do some bot-specific initializations outside of main, you are probably asking for trouble
20:39:17raptorbubble bubble toil and trouble
20:39:29Watusimotojust put everything in main
20:39:41Watusimotoexcept possibly general variable delcarations
20:39:49Watusimotobut
20:39:59Watusimotowhich levelgen is it, I can make sure it;s working
20:40:41sam686http://sam686.maxhushahn.com/bitfighter/levels/ the 6357.levelgen, and one of 6357 .level
20:40:54sam686mostly, the non-bitmatch one
20:41:39raptori use 6357core
20:41:50raptorit draws nice random walls now...
20:41:54raptorbut nothing else
20:54:18Watusimotodoes it even work the first time?
20:54:21Watusimotoprobably not...
20:55:12Watusimotoboy, that levelgen has a lot of "unwrapped" code!
20:55:29Watusimotowhich is almost certainly the problem
20:56:19raptorit does not the first time - i believe the problem is that main() is mandatory
20:57:13Watusimotono, it isn't
20:57:17Watusimotoat least not always
20:57:34Watusimotobut... why should I argue with you? always use main()!
20:57:46raptori agree
20:57:55Watusimotoit makes the code much easier to follow
20:58:13sam686just to let you know, i coded that back on, in bitfighter 015
20:58:17Watusimotoanything that can be compiled and cached, like a data structure, can be done outside main
20:58:44Watusimotosam686: not picking on you -- my own stuff (little of it there is) is probably even more broken
20:58:50Watusimotofor the same reasons
21:01:23Watusimotoactually, I'm not sure if what I said about running stuff outside main is actually correct
21:01:35Watusimotook, main may be mandatory
21:02:06raptori think you made a minor change to the main() check function that made it mandatory a week or two ago... still trying to find it
21:02:14raptorbut i think we should always use main() anyways
21:02:25Watusimotook. well we can all assume it is and everything will be fine
21:02:45raptorthat was the plan anyways, right? 017 main() not mandatory, 018 mandatory?
21:03:29Watusimotowas it? well, it's the plan now!
21:03:39raptorhaha
21:10:34sam686there is another problem with catched script: making changes to the script and it errors, fix that error in script, and the catched script won't reload the new, modified script, having to restart bitfighter to get it to see the new script
21:10:58raptoryes - Watusimoto suggested maybe disabling caching for editor testing...
21:12:17raptoror maybe we could be smart? like calculate a cksum for the file real quick and comparing it with one we have cached: if different, reload
21:12:44sam686maybe it can check the modified date and compare (maybe don't check if modified if there is already one or more robots)
21:13:03raptori guess that'd be even faster...
21:16:15raptori wonder, do we have modification checking capability already written somewhere?
21:18:57sam686i don't think there was one on bitfighter for file checking modified file (date, file size, and other checks to see if modified)...
21:20:58raptori wonder if dirent.h has something...
21:22:46sam686http://sam686.maxhushahn.com/upload/6357.levelgen a quick modify to put everything in main, I wonder how it works? Somehow it creates global variables inside a function main(), that can be used anywhere, like "sizex", something that can't be done on c++
21:23:23raptori'll be back in a bit...
21:23:26raptor Quit ()
21:39:24Watusimotomy idea was no caching in the editor, and a /disablecache option in the game for dev purposes
21:40:01Watusimotoby caching you avoid the cost of loading the file
21:41:00Watusimotoif the vars are in main, when the script executes, the "globals" are written to the function's environment table
21:41:17Watusimotoeach instance of the script has a private environment table that acts like a global environment
21:41:31Watusimotoit's all a bit confusing, which is why it took years to get it figured out
21:42:33raptor has joined
21:42:34ChanServ sets mode +o raptor
21:43:13Watusimotoso sam686 -- did you mods fix the script?
21:43:41sam686i jsut changed my 6357.levelgen script to fix
21:43:50Watusimotodid it work?
21:43:52sam686but other levelgens might be broken?
21:43:57Watusimotoprobably
21:43:58sam686yes, it works for my levelgen
21:44:13Watusimotomazegen was fixed a few releases back
21:44:52WatusimotoI don't know about any others
21:45:00raptors_bot works
21:45:11raptorarc script works
21:45:20raptorso no one created editor plugins yet...
21:45:35raptordid we not advertise that?
21:46:02Watusimotowho's going to create plugins?
21:46:16Watusimotothe whole plugin dev community is in this chatroom
21:46:17raptorthose myster lua programming bitfighter fiends we talked about...
21:46:26Watusimotooh right
21:46:58Watusimotothe next step in plugins is let them have access to selected items
21:47:09Watusimotoso you can write a script that manipulates items
21:47:19raptorso item IDs
21:47:23Watusimotooh yes
21:48:01Watusimotowhat about them?
21:48:13sam686other scripts i have works without main, only because it does not depend on having to have arguments, but in editor, missing walls when a level generates walls without main()
21:48:40raptorthat's how you get them to manipulate items? by ID, right?
21:48:55raptoror i may be misunderstanding you...
21:50:04sam686well... the walls in editor that came from levelgen without main() does appear, only after pressing ctrl + i
21:50:21sam686ctrl + i is slow for levelgen that creates lots of walls
21:53:18Watusimotoright now, ID serves no purpose whatsoever
21:57:09Watusimotosam686: do any of your levelgens depend on gridsize at all?
21:57:38raptorthinking of getting rid of it?
21:57:48WatusimotoI'd like to
21:57:53sam686mostly no, the level itself does the grid size, and the levelgen which adds barriers and such will automatically scale to grid size
21:57:56WatusimotoI think its an unnecessary complication
21:58:09raptori know _k loved it because it allowed dimensional scaling with scaling the items...
21:58:14raptor*without
21:58:20Watusimotowe can still have the parameter in the levelfile
21:58:30Watusimotobut levelgens shouldn't need to worry about it
21:58:54Watusimotoit's really more of a kind of "spread factor"
22:01:06sam686i can probably do levelgen:addLevelLine("GridSize 250") to change grid size from level editor (but, it only changes for new items that get added, not existing objects)
22:01:13Watusimotook, trying to push a shared_ptr through Lunar
22:01:18sam686and, i can make use of passing arguments from level to levelgen
22:02:15Watusimotoof course, you can just multiply levelgen coords by whatever factor to mimic the effect directly
22:02:16raptorgood luck!
22:02:38Watusimotolots of errors
22:03:17Watusimotobecause I used shared_pointer instead of shared_ptr
22:03:23raptorheh
22:03:39raptori was working on something recently... ijust have to remember
22:03:47Watusimotoif this works, it will save a lot of work and simplify our design
22:05:29raptoroh yeah - bot body shapes... did we really want those?
22:06:08WatusimotoI think we do
22:06:30WatusimotoI my kids were getting all amped up over this stupid driving game today because they found the custom bodies
22:07:03Watusimotoit's a little different, in that registered (i.e. paying) players could design their own, which is not what we're proposing, but nonetheless...
22:07:25Watusimotoyou already did one, right?
22:07:56raptoryeah, a klingon bird of prey
22:08:11Watusimotoso let's try that and see how it works
22:08:36Watusimototry making all bots use it (for the moment) and see if you like it
22:10:14raptori remember the problems i was having: some ship rendering is done in Ship, other in gameObjectRender
22:10:28raptorand i was having quite the time trying to pull it all together...
22:10:32Watusimotowhat a mess
22:10:45Watusimotoyou could start by consolidating all the code
22:10:52WatusimotoI've been doing it bit-by-bit
22:11:00Watusimotothere is no reason for it to be spread out
22:11:11raptorexactly
22:11:12Watusimotocould all be in gameObjRenderer
22:11:17raptornow to find my diff...
22:11:23Watusimotoso that would be good regardless of what bodies we use
22:18:31raptoroh no
22:18:33raptorlayerIndex
22:18:36raptorok
22:18:43raptortell me what layerIndex is really for?
22:18:58raptorrendering in a specific order in a for loop?
22:20:15Watusimotosome items get rendered above or below others
22:20:37WatusimotolayerIndex controls what gets rendered when
22:20:44raptorin Ship::render() it only allows -1 and 1
22:20:48raptorfor layerIndex
22:20:56raptorare there three layers then?
22:21:04Watusimotoprobably
22:24:12raptordoesn't opengl natively support layers?
22:24:23raptorbeing 3d and all...
22:25:41Watusimotoyes; sam uses them in a clever way
22:26:11Watusimotodon't know if we can use them instaed of multiple passes and levelindex, but it owul dbe cool if it worked
22:26:16raptorah ha! found the layers
22:26:31raptorClientGame.cpp:1846
22:27:15sam686 Quit (Ping timeout: 245 seconds)
22:28:07sam686 has joined
22:28:07ChanServ sets mode +v sam686
22:34:47raptorI think Ship has got to be one of the most complicated classes...
22:37:57raptoroh as a side note, how is SDL2 working for you guys on windows?
22:38:14raptorany other quirks yet? (except for my blindly coded mouse wheel direction)
22:39:24sam686cracatoa seems to complain about the fullscreen displays dock in mac, and clicking dock minimizes bitfighter (i guess bitfighter 017a is not SDL 2.0)
22:39:58raptoryeah, it's a temporary solution on mac
22:40:04raptorbut at least it doesn't crash anymore...
22:40:19sam686sdl 2.0 does appear to work just as good (probably better with all that stupid hide mosue pointer bugs) sdl 1.something
22:40:31raptorhaha
22:40:33raptorok good
22:46:10BFLogBot - Commit 76edac7bfe11 | Author: buckyballreaction | Log: Remove some debugging
22:53:57raptori cannot seem to find where the ship contrails are being rendered...
22:54:13sam686ship.cpp ship::render ?
22:54:25raptorthat's where i'm looking...
22:58:03raptorah, mTrail
22:58:09raptorinteresting...
22:59:12raptorha!: FXTrail::render()
22:59:21raptorhard-coded to the ship's dimensions...
23:01:02raptorwait, maybe not... hmm..
23:52:14sam686 Quit (Ping timeout: 245 seconds)
23:53:29sam686 has joined
23:53:30ChanServ sets mode +v sam686

Index Search ←Prev date Next date→

These logs were automatically created by BFLogBot on irc.freenode.net.