#bitfighter IRC Log

Index Search ←Prev date Next date→

IRC Log for 2011-06-28

Timestamps are in GMT/BST.

00:00:00Zoomberhmm
00:00:00Zoomberitseems
00:00:00Zoomberthe bots oh zap worked extremley well on stock maps
00:00:00Zoomberit was really the structure of the map
00:01:00Zoomberthey probably assumed if the bots worked on their maps it was fine
00:01:00Zoomberon maps with a huge barrier in the center of the map, or between flags, was harder to nagivate around
00:11:00raptorZoomber: this is a good explanation of the two main ways to do bot navigation: http://www.ai-blog.net/archives/000152.html
00:11:00raptorgotta go!
00:11:00raptor Quit (Remote host closed the connection)
00:23:00Zoomberoh that clears alot up actually
00:32:00Zoomberhey watusimoto: i just realized N doesn't create a speedzone in level editor anymore
00:32:00Zoomberwas that removed?|
00:33:00watusimotonot sure -- hover over the item on the dock and see -- shortcut will be listed in the help message
00:33:00watusimotoeditor keys will be remapped in next version
00:33:00Zoomberoh ok
00:34:00Zoomberok it isnt in instructions
00:34:00Zoomberso i guess its not in at all
00:34:00Zoomberbut maybe we should put N shortcut back in next version?
00:42:00watusimotono
00:42:00watusimotoall insertion functions will be ctrl-soemthing
00:42:00watusimotoor maybe alt-soemthing
00:42:00watusimotoinsert gofast... alt-g
00:43:00watusimotoor something
00:43:00watusimotomaybe ctrl-alt-g
01:02:00Flynnn Quit (Quit: This computer has gone to sleep)
01:04:00watusimoto Quit (Ping timeout: 240 seconds)
01:16:00Flynnn has joined
01:46:00watusimoto has joined
02:08:00sam686My last commit (mostly for windows only) was to remave inaccurate getTickCount in windows getRealMilliseconds (getting only max 64 FPS without getHighPrecisionTimerValue), so that means getTickCount updates only 64 times a second... I changed it to timeGetTime which is a lot more accurate milliseconds. Note that getHighPrecisionTimerValue in linux / mac is not high precision, it is only milliseconds precision, same as getRealMilliseconds..
02:13:00watusimotohey sam -- just wanted to know if you are still working on the split screen mode... is that something you still foresee finishing?
02:14:00sam686haven't worked on split-screen for a long time... But splitscreen might help for fullscreen users.
02:15:00watusimotoI know it's not an active project
02:15:00watusimotobut I wondered if it's an abandoned project, or just on hold
02:16:00sam686i might just hold it until after 016 is released..
02:16:00raptor has joined
02:16:00ChanServ sets mode +o raptor
02:16:00watusimotook, sounds good
02:17:00watusimotoI was lukewarm about it before, but now I'm thinking it's a good idea
02:17:00raptorSDL fullscreen means no resizing
02:17:00watusimoto?
02:17:00sam686SDL fullscreen might not be 4:3 ratio...
02:18:00sam686depending com computer monitor.
02:18:00watusimotoany fullscreen might not be that ration
02:18:00watusimotoback later
02:20:00sam686raptor, did you see my comment about my changes in todays log [2:08]?
02:20:00Flynnn Quit (Read error: Connection reset by peer)
02:23:00watusimoto Quit (Ping timeout: 250 seconds)
02:25:00raptoryes
02:25:00raptorthanks for clearing that up
02:26:00raptorit wasn't commited by sam8641, so i thought to ask
02:26:00sam686i think my hg update could have messed up my name...
02:26:00sam686software hg update..
02:26:00raptorah - you'll have to edit the INI again
02:26:00raptoror whatever the config file was..
02:30:00sam686at least at googoe code, it says Owner@sam-8759b470e6e, it used a username (which was Owner) and network name (sam-...)
02:30:00raptorinteresting
02:32:00sam686i remember an older version of hg require filling out username when commiting..
04:20:00sam686seems to be a small bug in stats: http://bitfighter.org/gamereports/graphs/ ... probably came from buggy Game ending on 20 Jun 2011 12:18:26 AM and not being able to transfer all data causing Robot.59 having -1073763280 score..
05:37:00watusimoto has joined
06:25:00raptorgood evening
06:27:00sam686hi.
06:28:00raptorwhoa what happened to players-per-game
06:28:00sam686maybe one of a stats screwed up and throws in players of what is actually a robot?
06:29:00raptorreally weird
06:29:00raptoronly watusimoto has direct access to the database and can fix it...
06:31:00watusimotowhat's the prblem?
06:31:00raptortake a look at game stats for the last 30 days
06:31:00raptorbottom graph
06:32:00sam686Game ending on 20 Jun 2011 12:18:26 AM in http://bitfighter.org/gamereports/ seems to be screwed up
06:32:00raptorsam686: did that come from your server?
06:32:00sam686server: Flynn (New level: Infantry) , is not my server...
06:32:00raptorah, ok
06:33:00raptoryeah - i suppose i should stop being lazy and look up the game on the reports page...
06:33:00watusimotook
06:34:00watusimotolook fine at a first glance
06:35:00watusimotowhat's wrong with 12:18 game?
06:35:00sam686look at Robot.59 kills deaths and suicides, it is way too high, and exactly the same..
06:35:00watusimotooh my
06:36:00watusimotorobot.59 did not do so well
06:36:00watusimotobut he killed a lot
06:36:00raptorhaha
06:36:00watusimotolooks like someting overflowe
06:36:00watusimotod
06:36:00sam68649151 kills 49151 deaths 49151 suisides? why are they the same?
06:36:00raptoryeah...
06:37:00watusimotoperhaps Flynn scored more than S16?
06:37:00watusimotoand polluted the following stats packets?
06:37:00sam686Points are in S32, kills / deaths / suicides are in U16..
06:38:00watusimotostill looks like an overflow
06:38:00sam686to me it appears that the packet got too big or got damaged..
06:38:00watusimotoyes
06:39:00watusimotois that flynn's server?
06:39:00sam686yes, look at Server anem and IP Address...
06:39:00watusimotomaybe he has a local log file that would provide more info?
06:40:00watusimotobut also notice the first record -- no name
06:40:00sam686must be related to data too big to send in one packet, and TNL RPC appears to cannot be split one data into multiple packets...
06:41:00watusimotothat may be it
07:04:00watusimotowell, I understand why the teams are broken in the editor
07:04:00watusimotono idea how to fix them
07:51:00raptormagic?
08:14:00raptor Quit (Remote host closed the connection)
08:58:00watusimoto Quit (Ping timeout: 240 seconds)
09:41:00kodax has joined
10:41:00BitfighterLogBot - Commit 22d7b8acca2e | Author: sam8641 | Log: Change TNL to support RPCGuaranteedOrderedBigData, for sending up to 256 kilobytes, for s2mSendStatistics
11:56:00sam686 Quit (Quit: .)
14:53:00watusimoto has joined
14:54:00watusimotoI tried falling asleep at the computer, but that didn't work
15:39:00raptor has joined
15:39:00ChanServ sets mode +o raptor
15:40:00raptorsleeping at the computer didn't solve the editor team problem? darn...
15:40:00watusimotono, but I've figured it out
15:40:00raptorit came to you in a dream? :]
15:40:00watusimotomake game manage the teams, rather than gameType
15:41:00watusimotothen pass pointers to the teams
15:41:00watusimotobut have the editorGame override the methods to use EditorTeams
15:41:00watusimotoand the client/serverGame use Teams
15:41:00watusimotoshould work
15:42:00raptori just found out about 'hg bisect'
15:44:00raptorfor regression testing
15:52:00watusimotobut my method does lead to some ugliness: dynamic_cast<Team *>(mGame->getTeam(i))->getScore()
15:55:00watusimotoor the slightly better ((Team *)(mGame->getTeam(i)))->getScore()
15:56:00raptoris there any other down side, other than ugliness?
15:56:00watusimotono
15:56:00watusimotoat least not yet discovered
15:57:00watusimotobut casting is often a sign of design weakness
15:57:00raptorhappy tau day!: http://tauday.com/
15:58:00raptoryou mean the object hierarchy could probably be improved?
16:32:00watusimotopossibly... though I'm finding some defects in the way I implemented it intially, it's probably right
16:32:00watusimotowe have an AbstractTeam object from which Teams and EditorTeams are descended
16:32:00watusimotobut it's kind of messy; I'm cleaning it up
16:46:00BitfighterLogBot - Commit 32251c913b98 | Author: buckyballreaction | Log: Update OpenAL-Soft git revision to support CoreAudio input - VoiceChat works on Mac now - Update project file
16:47:00raptorok phoew! that's over with...
16:47:00raptornow on to input unification...
17:08:00watusimotorats
17:08:00watusimotoI have to switch to shared_ptr to make this work
17:13:00kodax Quit (Quit: Sto andando via)
17:14:00[1]watusimoto has joined
17:17:00watusimoto Quit (Ping timeout: 244 seconds)
17:17:00[1]watusimoto is now known as watusimoto
17:18:00raptorso what is your opinion of shared_ptr now having used it for a little while?
17:19:00watusimotosafer than using raw ptrs
17:19:00watusimotomakes me a little less nervous about memory leaks
17:20:00raptordo you think the good outweighs the bad? (like code that is a little more cumbersome)
17:21:00watusimotodefinitely
17:21:00watusimotoI wouldn't be using them otherwise
18:15:00watusimotogetting sidetracked moving rendering code from gameType into UIGame
18:15:00watusimotohttps://code.google.com/p/bitfighter/issues/detail?id=4
18:15:00watusimotobah
18:32:00raptori remember that issue!
18:58:00watusimotostill plugging away on it...
18:59:00raptori don't know if I mentioned - the games main idle() loop needs some clean-up
18:59:00raptorthere seems to be two separate sleep times it uses
18:59:00raptorI am wondering if we can make it only one sleep time
18:59:00raptorwe could cut down a lot of CPU time by tuning the idle() loop appropriately
19:01:00raptori don't know how much knowledge you have of the idle() functions
19:03:00watusimoto Quit (Ping timeout: 260 seconds)
19:13:00koda has joined
20:06:00watusimoto has joined
20:38:00sam686 has joined
20:38:00ChanServ sets mode +v sam686
23:06:00raptor Quit (Remote host closed the connection)
23:29:00koda Quit (Quit: k thx bai)
23:51:00Flynnn has joined

Index Search ←Prev date Next date→

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