#bitfighter IRC Log

Index Search ←Prev date Next date→

IRC Log for 2012-03-02

Timestamps are in GMT/BST.

00:00:16Aggropanda Quit (Quit: Page closed)
01:19:11BFLogBot - Commit 8f803d4a7f91 | Author: watusim...@bitfighter.org | Log: Changed signatures
01:19:12BFLogBot - Commit 22956ae63043 | Author: watusim...@bitfighter.org | Log: Added color
01:19:14BFLogBot - Commit 3ecf80a8c249 | Author: watusim...@bitfighter.org | Log: Move spark emissions to CoreItem, other consolidation... somehow, this makes "leak sparks" worse
01:24:10Watusimoto Quit (Ping timeout: 272 seconds)
02:19:20BFLogBot - Commit 42ced7fe1f2b | Author: sam8641 | Log: Fix a possible crash when getClientInfo returns NULL
02:27:39sam686 Quit (Ping timeout: 245 seconds)
02:27:44sam686 has joined
02:27:44ChanServ sets mode +v sam686
02:31:49Little_Apple has joined
02:31:55Little_Appleaww
02:31:57Little_Applehellooo
02:40:26Little_Apple Quit (Ping timeout: 245 seconds)
03:18:49raptor has joined
03:18:49ChanServ sets mode +o raptor
03:50:40raptorgood evening
03:56:50raptorsomehow Core is rendered in reverse...
03:57:57sam686or is it that core is moving in reverse?
03:58:05raptoryes, that
03:58:37raptorand your center kaleidoscope graphic is in reverse, too
03:59:26sam686it was going the right way on 016
03:59:31raptoryep
04:00:00sam686also, a problem with /settime 0 (unlimited / no time limit) makes core not move.
04:00:10raptoryeah, that a larger problem
04:00:24raptorthat will require synchronizing a different timer between server and client
04:03:04raptorthat is so weird...
04:03:09raptorwhat is making it go backwards?
04:03:51sam686taking from time remaining which count down (where in 016, a simple milliseconds is used which count up)
04:04:12raptorahh
04:07:05raptormaybe I can change it to getTimeElapsed
04:13:38raptornope, guess not...
04:13:59raptoroh wait, yes i can
04:16:35raptorha! i had no idea, that to render the Bitfighter logo, you just have to add a TextItem called 'Bitfighter'
04:19:27Little_Apple has joined
04:19:44Little_AppleHELLOOOO
04:19:48sam686hi
04:19:52raptorhola
04:20:06Little_AppleRaptor
04:20:27Little_AppleDo you ever give your kids piggy back rides?
04:20:38Little_Apple:3
04:20:39raptormy oldest is 2
04:20:43raptorso not yet
04:20:56Little_Applehttp://www.geekalerts.com/daddle-the-dad-saddle/
04:20:59Little_Apple:D
04:21:43Little_Apple Isn't it genius?
04:22:09Little_AppleErgo
04:22:11Little_AppleLaters
04:22:18Little_Apple Quit (Client Quit)
04:27:06raptoractually it may just be easier to reverse the kaleidoscope..
04:27:19raptorsorry, spirograph thingy
04:34:48sam686maybe, as easy as using -time_remaining to reverse the effects.. but that won't solve the unlimited time / no time limit..
04:35:13raptoryeah, that's a different problem we may want to tackle for after 017
04:40:12sam686i don't see anything in TNL bitfighter code that sends server current getMilliseconds to client
04:40:23sam686maybe i could add that..
04:40:32raptorthere isn't
04:40:47raptorbut there needs to be - there may be future things we want to sync
04:42:23sam686another problem to solve is a problem with how sending a clientGame's time remaining can be inaccurate, i frequently see 9:32 jump to 9:30 skipping 9:31..
04:43:05raptoryes, maybe people have their cpu running faster, slower, and the time sync from server is slow sometimes
04:43:35sam686it is mostly something to do with, internet lag delaying receiving..
04:44:34raptoryou think so?
04:46:37sam686the tnl ping system only happens about once every second, that makes getting ping number somewhat inaccurate
04:46:49raptorah
05:04:35BFLogBot - Commit c02e8a5bbc14 | Author: buckyballreaction | Log: Make Core center atomic-like graphic render correctly (not reversed)
05:04:36BFLogBot - Commit 29353375e747 | Author: buckyballreaction | Log: Fix compiler warning
05:29:40BFLogBot - Commit 7ce889d351a8 | Author: buckyballreaction | Log: Consolidate panel damaged logic
06:02:41sam686good night, i think we will leave with BFLogBot being lonely...
06:02:47raptoryes
06:02:49raptorgood night
06:02:54raptor has left
06:19:58sam686 Quit ()
08:41:42watusimoto has joined
08:41:42ChanServ sets mode +o watusimoto
14:26:54raptor has joined
14:26:54ChanServ sets mode +o raptor
14:58:31LordDVG has joined
15:03:51LordDVG Quit (Remote host closed the connection)
15:06:32LordDVG has joined
15:16:43LordDVG Quit (Ping timeout: 245 seconds)
15:19:15LordDVG has joined
15:42:40LordDVG Quit (Read error: Connection reset by peer)
15:43:11LordDVG has joined
16:21:58LordDVG Quit (Remote host closed the connection)
17:01:10Watusimoto_ has joined
17:28:16watusimoto Quit (Read error: Operation timed out)
17:31:37raptor Quit ()
18:10:00LordDVG has joined
18:26:31Watusimoto_ Quit (Ping timeout: 260 seconds)
18:30:07raptor has joined
18:30:07ChanServ sets mode +o raptor
19:32:15Watusimoto has joined
19:35:45LordDVG Quit (Read error: Connection reset by peer)
19:39:52LordDVG has joined
19:49:43raptorhi Watusimoto
19:50:08raptorso I can get the ship location in CoreItem::damageObject() to properly heal only those panels that are nearest
19:50:22raptorbut i cannot get the ship location in getRepairTargets()
19:50:33raptorso the repair rays are always going to all damaged panels
19:57:14Watusimotohi
19:57:29Watusimotook
19:57:36Watusimotois your code checked in?
19:58:04raptori did not check in that enhancement because of those problems
19:58:13raptori did save it as a patch though
19:59:00Watusimotoso if I work on the getRepairTargets, will that overly conflict with what you did?
19:59:05raptornah
19:59:14raptori'm good with managing patchsets
19:59:15Watusimotook, let me take a look
20:02:07raptoralso, if we get this to work - do you want the health divided amongst panels, still? I'm figuring no, since we'll be making panels have the same behavior as other objects
20:03:01Watusimotocorrect -- no
20:03:09raptorok, that's what i coded
20:03:17Watusimotocores are so easy to kill, and hard to repair...
20:03:36raptora little harder to kill now with targets..
20:07:39raptorwe're definitely closer to your star castle thingy
20:08:02Watusimotoa little
20:08:28WatusimotoI have more (feasible) ways to get closer
20:08:59raptori would like my ship to be able to leave the screen on the left and appear on the right when attacking Cores
20:09:09raptor:)
20:09:31Watusimotogreat idea
20:09:52Watusimotook, I think the solution is simply to pass the ship's pos into getReapirLocations()
20:10:43Watusimotothen compare pos.distanceSquared(mid) to repair radius
20:11:04Watusimotoin CoreItem::getRepairLocations()
20:11:15raptoryes, that is what i do in damageObject
20:11:18Watusimotoor, rather, CoreItem::getRepairLocations(const Point &repairOrigin)
20:11:21Watusimotoor somesuch
20:11:38Watusimotowell, that seems an easy and benign solution
20:11:41raptorare you adding this in?
20:11:52raptorbecause i can, if your not there
20:12:00Watusimotogo for it
20:12:09Watusimotoyou have all the other related stuff
20:12:21raptorok
20:12:49Watusimotoof course, you will need to change all the other places where we define gRL() to have the same sig, but you certainly know this
20:18:47raptorthis might actually work...
20:19:28raptorexcept, i'm checking to see if each panel midpoint is within RepairRadius
20:20:44Watusimotodo you think we should check at the repair pioint? We could create a function to caluclate it
20:21:00Watusimotowe use it elsewhere; for rendering, for example
20:21:02raptorthe ship may be too far away
20:21:05Watusimotoand for sparks
20:21:14raptorrepair point is the health bar stake?
20:21:19Watusimotoyes
20:21:19raptor(or it should be?)
20:21:36Watusimotowe currently use the center of the core for repair, do we not?
20:21:42raptoryeah, RepairRadius is too small...
20:21:57raptorwell... object detection is at RepairRadius
20:22:01raptorso any part of the object
20:22:14raptorwhich is why you get a longer beam with Core
20:23:01raptorhey it works...
20:23:03raptorcool
20:23:12raptorbut you can only heal at most 2 panels
20:23:31raptorbecause repair radius is only comparing the midpoint instead of the whole line
20:23:45raptormaybe i should compare endpoints
20:25:07raptoris there an easy way to get those stake points?
20:26:13raptorahhh... much better repair behavior...
20:26:26raptori'm going to chekc in what i've got
20:29:56Watusimotook
20:30:01Watusimotono way to get the stakepoints
20:30:08Watusimotobut we could create one easily
20:30:24Watusimotoin fact, I'll do that now
20:31:13raptorok, i just pushed
20:31:19raptorif you want to work off of that
20:32:25BFLogBot - Commit 9d82b6bf3151 | Author: buckyballreaction | Log: Only repair the panels closest to the ships. This brings panel repair more in line with standard gameObject behavior
20:45:28Watusimotook
20:45:32WatusimotoI think it's done
20:45:41Watusimotocompiling now
20:59:09Watusimotochecked in
20:59:18raptoroh boy
20:59:19Watusimotoit compiles, but I haven't actually tested it yet
20:59:35Watusimotowe need to use those same calculations in the renderer
20:59:42Watusimotoand I'm a little torn
21:00:08WatusimotoI don't think the renderer should be calling methods on the core -- generally, we pass everything the renderer needs and it just draws
21:00:12WatusimotoI think we shoul dkeep this
21:00:14Watusimotobut
21:00:28WatusimotoI don;t like duplicating the calcuations in the renderer metohd
21:00:39Watusimotoand passing 40 points to the renderer seems excessive
21:00:50raptoryes, this was my dilemma, too
21:00:59Watusimotoand I don't like the fact that we do these relatively expensive calcs over and over
21:01:13raptoryep
21:01:19Watusimotoso I was thinking about doing them once (in idle) and caching them in a vector
21:01:28Watusimotowe could pass that vector to the renderer
21:01:37raptorI thought about creating a CoreItem::PanilInfo struct
21:01:48Watusimotothat would probably be better
21:02:21Watusimotobut we only really want to calc them when we need them -- most of the time during a game they are off screen and out of sight
21:02:27Watusimotoand who cares about the angle of a panel?
21:02:29BFLogBot - Commit 0de6dd4c3701 | Author: watusim...@bitfighter.org | Log: Centralize the calculation of the "stake" point, called repair point here.
21:02:32Watusimotoso
21:03:31Watusimotomy next idea is to create a getPanelStruct() function that lazily inits itself; if used, it caches the result. if unused, the points never get calced
21:03:37raptoryou broke repair...
21:03:40Watusimotoseems excessive
21:03:51Watusimotobut probably not too hard
21:03:54Watusimotowhat broke
21:03:55Watusimoto?
21:04:08raptori fix
21:04:17Watusimotothis was such a straightforward fix!
21:04:31raptoryou changed my distance test to: repair.distSquared(repairOrigin)
21:04:43raptorwhich is *always* outside RepairRadius
21:05:12raptorhave to test against the start and end panel points for determining if the panel is in range
21:05:47Watusimotowait.. what?
21:06:12Watusimotobefore this latest round of mods, we checked if the center of the core was in repair radius, no?
21:06:16raptori fix
21:06:23raptorno
21:06:27Watusimotono?
21:06:28raptorrepair radius is only 65
21:06:32Watusimotooh
21:06:35Watusimotocore rad is 100
21:06:39Watusimotook
21:06:41raptorthe test is against the the extents of the gameobject
21:06:52raptori fix
21:06:53Watusimotook, so I "fixed" something that wasn
21:06:56Watusimotot broke
21:07:00raptorhehe, yes
21:07:14Watusimotohrm
21:07:16Watusimotosorry
21:07:25Watusimotook, why is mCurrentExplosionNumber static?
21:07:32Watusimotowhat happens if two cores explode at once?
21:07:47raptorahh, much better
21:08:17WatusimotoI think it should not be static
21:08:27raptoragreed
21:08:31raptormy fault
21:08:51raptorand that explains some oddities i've noticed...
21:08:57Watusimotook, I'll remove
21:09:14raptorjust pushed my repair fix...
21:09:33Watusimotorats -- you beat me
21:10:18raptori win!
21:10:22Watusimotowas using svn today... it feels so clunky
21:10:33raptoroh my goodness yes
21:10:46raptorall of my work code is in SVN, and i feel restricted with it
21:10:58raptori yearn to use hg or git
21:11:01Watusimotowe use git hub, except for documents
21:11:18raptorcan you use github for work?
21:12:21Watusimotosure
21:12:32BFLogBot - Commit a1224f036887 | Author: buckyballreaction | Log: Fix repair distance test
21:12:33BFLogBot - Commit 5260aa5e7f17 | Author: watusim...@bitfighter.org | Log: whitespace
21:12:35BFLogBot - Commit 774631e14952 | Author: watusim...@bitfighter.org | Log: should not be static
21:12:36BFLogBot - Commit 8a7fbaea7cb1 | Author: watusim...@bitfighter.org | Log: Merge
21:12:36Watusimotoeverything I do is opensource
21:12:45raptorinteresting... ok
21:12:55raptormakes sense then
21:13:20WatusimotoI
21:13:29Watusimoto'm working on 2 projects in github
21:13:42Watusimotowell, 1 isn't there yet, but will be soon
21:18:45raptorCore sparkles have been increased again?
21:24:36jan______ has joined
21:24:57jan______ Quit (Client Quit)
21:31:08LordDVG Quit (Ping timeout: 240 seconds)
21:31:54Watusimotono, I think I just screwed it up last night
21:32:11WatusimotoI made an inconsequtial change, and it all just went to hell
21:32:24Watusimotoso yes, actually, but not intentionally
21:32:30Watusimotoor rather yes maybe
21:32:43Watusimotomaybe the same number, just the linger longer
21:32:46Watusimotoor something
22:07:17raptorvnc + stuck modifier are driving me mad!
22:42:45BFLogBot - Commit 48ce182c6faf | Author: watusim...@bitfighter.org | Log: Comment
22:42:46BFLogBot - Commit 185692a3ebe2 | Author: watusim...@bitfighter.org | Log: Reduce redundant calculations boy was this a pain!
22:43:10raptoryay
22:53:36Watusimotothis is a nightmare
22:53:45raptornightmare
22:56:58raptorhey explosions work better now
22:59:02sam686 has joined
22:59:02ChanServ sets mode +v sam686

Index Search ←Prev date Next date→

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