#bitfighter IRC Log

Index Search ←Prev date Next date→

IRC Log for 2012-11-11

Timestamps are in GMT/BST.

00:04:02raptor has joined
00:04:02ChanServ sets mode +o raptor
00:04:27raptorok i awoke
00:04:31raptor reads logs
00:06:03raptorWatusimoto: render position is needed client side, it is almost always used to place an object first, then correct once an update is received from server
00:07:52Watusimotoyes
00:08:07Watusimotoactually, I was able to remove the method where I was fretting about this
00:09:48raptorok
00:13:21Watusimotothis little minor glitch has turned into so much more
00:13:31raptorthe soccerball?
00:13:48Watusimotoyes -- it illustrated an unseen problem with the way we manage extents and bounds and such
00:13:58raptorit wasn't in 017b
00:14:03Watusimotothe bouds were being set twice, differently, hence the flickering
00:14:12Watusimotono, probably not
00:14:31Watusimotobut I refactored a lot of this and may have introduced the problem a while back
00:14:49Watusimotobut... maybe... I've got it fixed
00:15:05Watusimotocompiling now to test
00:15:36Watusimotoif so, it will get rid of a couple of todos and other pesky comments where I sensed things were a bit wrong
00:16:34raptorok good
00:16:55Watusimotoyes, probably good
00:17:36Watusimotoi'm falling asleep again...
00:17:58Watusimotoso I'll preemptively say goodnight
00:25:30raptornight
00:38:49koda Quit (Quit: koda)
00:48:38sam686|-| i
01:02:35Little_Apple Quit (Quit: Page closed)
01:14:39sam686 Quit (Quit: :))
01:23:17amgine1234567890did oyu read my reprot on the bots
01:23:34amgine1234567890osmtimes they seem to get confused and not know where to go=p
01:23:43amgine1234567890thats using the version same hgave me last
01:25:21BFLogBot Commit: 959c307f26a8 | Author: watusimoto | Message: Comment
01:25:22BFLogBot Commit: 2406f18e553c | Author: watusimoto | Message: Whitespace
01:25:24BFLogBot Commit: 772cbe84e36e | Author: watusimoto | Message: Get extent for objective arrow rather than bounds. Makes no practical difference, as far as I can see, and lets us get rid of the getBounds method that is otherwise unused.
01:25:25BFLogBot Commit: b97849250104 | Author: watusimoto | Message: Fix soccerball issue; mines currently have wrong scope, but I think this is easily fixed tomorrow. Want to get this in so if there are larger issues we can spot them sooner. This should be slightly more efficient, and makes the logcial move of storing an object's radius on the PointObject... after all, every point item should have a radius, no?
01:25:59raptorflags don't have a radius
01:26:10raptorwell, technically one of zero
01:26:12Watusimotowell, they have a small one
01:26:20Watusimotothey have one
01:26:22raptorreally?
01:26:24raptorhmmm
01:26:26Watusimotothey are moveItems, no?
01:26:30Watusimotoall moveItems have a radius
01:26:35raptoryes
01:26:43Watusimotoso this changes nothing for flags
01:26:46raptorok
01:26:51Watusimotoit makes a change for flagspawns
01:26:55Watusimotobut it doesn't matter
01:27:02raptorok
01:27:15Watusimotobecause they only interact spatially with lua scripts doing a select of some area
01:27:18raptordinner! i'll check the commits a bit later..
01:27:21Watusimotook
01:32:07Watusimoto Quit (Ping timeout: 246 seconds)
02:04:53raptoramgine1234567890: yes, i saw your report, thanks!
02:05:01raptorthat problem with bots has been around for a while
02:05:14raptorbut you say it didn't crash in any of the gametypes, right?
02:14:03amgine1234567890correct
02:21:08amgine1234567890 Quit (Quit: Page closed)
03:47:33sam686 has joined
03:47:36ChanServ sets mode +v sam686
04:13:28Wuzzy2 has joined
04:16:07Wuzzy Quit (Ping timeout: 245 seconds)
04:58:59Wuzzy2 Quit (Quit: Wuzzy2)
05:38:52raptor Quit (Ping timeout: 246 seconds)
05:39:20raptor has joined
05:39:20ChanServ sets mode +o raptor
06:09:30raptorhi sam686
06:10:36sam686hi
06:10:52raptordid your 017b server run on a VPS?
06:11:38sam686oh it was on one of my computers..
06:14:16sam686also, it took 2 hours to copied the windows partition to a 2 TB hard drive, rather then keeping that on old (was on broken laptop) 100 GB having about 200 bad sectors that was put into the desktop tower computer because of ran out of hard disks..
06:15:40raptordo you have almost everything set up on your new computer?
06:17:50sam686on my old pentium D computer, it came with a now-dead hard disk, so I put a small 2.5 inch hard disk with already some re-allocated sectors cause that all I had back then a year ago..
06:18:06sam686now my pentium D has 2 TB disk
06:18:20raptorah
06:19:31sam686as for my even older pentium 4 with 2 years old PATA disk, not sure if I need that anymore.. so I put that PATA 80 GB to my pentium D also...
06:19:52raptorwhat about your i5?
06:20:04sam686my intel i5 computer came with 1 TB disk already in it as windows 7, still works much faster..
06:28:44raptorthere were some people asking where your 017b server was the other day
06:36:04sam686 has left
06:36:15sam686 has joined
06:36:15ChanServ sets mode +v sam686
06:38:37raptorok, time for bed
06:38:39raptornight
06:40:15raptor Quit ()
09:10:09Watusimoto has joined
09:45:03BFLogBot Commit: 4b9502dcd6d7 | Author: watusimoto | Message: Rename BurstProjectile to Burst because it isn't a projectile!
09:45:05BFLogBot Commit: 73b4357d3f3f | Author: watusimoto | Message: Fixes mine scope issue, seems teleporter has similar issue
09:45:06BFLogBot Commit: 754967dc906c | Author: watusimoto | Message: Comment
09:52:42Watusimoto Quit (Ping timeout: 265 seconds)
12:04:12Watusimoto has joined
13:22:44BFLogBot Commit: 11d826463b9d | Author: watusimoto | Message: Cleanup and quizzical comment
13:22:45BFLogBot Commit: 560b8e3eb3ae | Author: watusimoto | Message: Add some low utility memory corruption protection
13:22:47BFLogBot Commit: cd5ebf33fbfd | Author: watusimoto | Message: Comments and var names -- why should squares be so mysterious?
13:22:48BFLogBot Commit: 6fbcd14bbe33 | Author: watusimoto | Message: Clarifying comment
13:22:50BFLogBot Commit: ea6aab87e271 | Author: watusimoto | Message: Whitespace
13:22:51BFLogBot Commit: a0b6c00abaa0 | Author: watusimoto | Message: This seems to have no effect whatsoever...
13:22:53BFLogBot Commit: 7c7a2d64cb99 | Author: watusimoto | Message: Correct these misleading and incorrect comments
13:37:44BFLogBot Commit: 9ef19d6b67bb | Author: watusimoto | Message: More clarification of comments and var names
13:37:46BFLogBot Commit: 363902b641f7 | Author: watusimoto | Message: Now that we've clarified that TELEPORTER_RADIUS is indeed a radius, and that the Rect(point, size) constructor takes a diameter and not a radius, it becomes plain to see why teleporters disappear before they are completely off the screen. And once we understand the cause, the fix is trivial, and is the main focus of this checkin, along with finalizing the removal of yet another relic of this confusion. All of these problems could have been avoided with better variable names and clearer comment
13:39:36BFLogBot Commit: be92cd596f19 | Author: watusimoto | Message: More commentary
13:48:34koda has joined
13:52:39BFLogBot Commit: 36ab106ea94a | Author: watusimoto | Message: Another correction in a similar vein to the others I've just done... this also seems to have no effect.
13:52:40BFLogBot Commit: 908b5c83bb92 | Author: watusimoto | Message: More efficient this way
13:52:42BFLogBot Commit: 9e59ee332687 | Author: watusimoto | Message: Fix internal inconsistency in how we determine if a ship hit a teleporter. This makes teleporters a little more sensitive, and is a change of behavior -- it is no longer possible to fly almost entirely into a teleporter without teleporting. If you disagree with this change, we can revert it, as long as we keep things consistent.
14:00:24raptor has joined
14:00:24ChanServ sets mode +o raptor
14:01:37Watusimotohello everyone
14:01:43raptorhi
14:01:45raptorwow
14:01:49raptoryou've been busy
14:01:56WatusimotoI just found a very rare and not too serious bug, that will be hard to fix, and I'm trying to decide what to do
14:02:16raptorok
14:02:43Watusimotoif two ships hit a teleporter at the same time, they both get teleported, each their own desitnation. But the teleporter effect only gets played at one of those destinations.
14:02:49Watusimotowe could fix in 3 ways
14:03:03raptoreffect?
14:03:17Watusimoto1) only teleport one ship, regardless (an 99.999% of the time this would be exactly what happens in practice)
14:03:31Watusimotoeffect = swirly effect by teleported ship, plus sfx
14:03:45Watusimoto2) teleport all ships to same destination
14:03:55Watusimoto3) ignore and not worry about it
14:04:18raptoryou mean the same teleporter at the same time?
14:04:29Watusimoto4) create some complex contrivance to send multiple destinations through the server-> client mechanism
14:04:31Watusimotoyes
14:04:34Watusimotoat the same frame
14:05:11Watusimotothis is very rare
14:05:45Watusimotocurrently, we do #3
14:05:45WatusimotoI am voting against #4
14:05:53raptorsame place.. as in same destination, or same point?
14:06:22raptorbecause offsets from center are kept if two ships enter the sam telepoter
14:06:33Watusimotommm... would probably same dest, as ship transmits its own coords separately, I think
14:06:35raptorused in dungeons a lot
14:07:00raptorsame dest makes sense to do
14:07:10Watusimotoso I think if two ships hit the same teleporter at the same time, and went to the same dest, they'd have the same relative coords when they arrived
14:07:24raptorsounds good to me
14:07:26WatusimotoI'm leaning that way too
14:07:32Watusimotook, I'll fix it that way
14:08:09Watusimotobut you and sam686 need to pay attention to my most recent checkin... read the comments and let me know if you have a problem
14:08:26Watusimotoas it might break dungeons
14:08:32raptorreviewing them now
14:08:45Watusimotook, I'll be back in a while... going to play games with my kids
14:08:56Watusimotoyou should have enough to chew on in the meanwhile :-)
14:09:48raptorbut Bursts *are* projectiles
14:09:59raptorthey fly through the 'air'...
14:10:19raptorbut not in the object hierarchy...
14:15:08raptorheh
14:15:21raptori like the long comment about teleporter radius / rect diameter..
14:17:28raptorhmmm.. i think you've just broken any dugneon ever made...
14:35:02raptori think people are going to scream with that last change...
14:35:17raptorit's a very big change
14:51:53raptorfrom the user's perspective..
15:15:00Wuzzy has joined
15:36:57BFLogBot Commit: d57b2c46eac2 | Author: buckyballreaction | Message: Add ability to store and display configuration errors on start-up. Add possible joystick preset error to it
15:41:07raptorWatusimoto: i created a member on GameSettings: mConfigurationErrors
15:41:27raptorit can be used to load up any configuration errors for the error message on game start-up
16:21:15Watusimotohi
16:21:20raptorhi
16:21:46Watusimotowe can easily revert the calculation of where you hit the teleporter
16:22:13WatusimotoI've done a brief refactor that makes it even easier to revert (inlcuding a comment as to where and how)
16:22:35WatusimotomConfigErrors sounds like a great idea
16:23:15raptorgood
16:23:22raptorone more layer of complexity!
16:23:54Watusimotowell, I've reduced complexity in teleporter
16:23:58Watusimotoso we're even
16:24:03raptor:)
16:24:34raptori think that we should revert to old teleporter behavior... i do favor having the entire ship in the teleporter before teleport
16:30:13Watusimotook, reverted
16:30:14raptorwas there a specific reason why you changed it? i think sam686 and I tinkered with that before some time, too...
16:30:48WatusimotoI changed it because we had two checks for intersection (now reduced to one), and one was one way and the other was the other... I picked one semi-randomly
16:30:57raptorok
16:31:09raptorso i don't feel bad for requesting old behavior
16:31:14Watusimotono
16:32:11WatusimotoI independently concluded I like the old behavior better as well
16:32:23Watusimotomore like you fly in, rather than come into contact with
16:32:38Watusimotothe new way seems a bit twitchy
16:33:10raptoryes
16:43:45BFLogBot Commit: 0bc4cd183883 | Author: watusimoto | Message: Simplify teleporter logic; fix bug with multiple ships hitting teleporter at one time (very rare, never actually observed); now if multiple ships hit at once, they all go to same dest; more comments; reverted teleport radius to 017b behavior
16:43:47BFLogBot Commit: b86cc0c8ae30 | Author: watusimoto | Message: Merge
16:48:11Watusimoto Quit (Ping timeout: 255 seconds)
17:13:28LordDVG has joined
17:16:51Watusimoto has joined
17:38:09BFLogBot Commit: 7bb77a3f67ac | Author: watusimoto | Message: Formatting
17:38:10BFLogBot Commit: b4c6edfa031b | Author: watusimoto | Message: Remove dead code relating to soccer ball carrying
17:38:12BFLogBot Commit: d86d330ea7c5 | Author: watusimoto | Message: Whitespace
17:38:13BFLogBot Commit: 3c4977da49f5 | Author: watusimoto | Message: Stop sending postion updates if soccer ball isn't actually moving. This was the other underlying cause of the flickery soccer ball.
17:38:15BFLogBot Commit: d78df3afb577 | Author: watusimoto | Message: Formatting
17:45:27BFLogBot Commit: 50d4db427aa4 | Author: watusimoto | Message: Add debugging section to lua docs
18:14:05koda Quit (Quit: koda)
18:14:42raptorhi again
18:30:58LordDVG Quit (Remote host closed the connection)
18:31:16LordDVG has joined
19:41:58Watusimotohi
19:42:12raptorhello
19:42:15WatusimotoI'm running out of stuff I can work on
19:42:26raptorhttp://bitfighter.org/wiki/index.php/Running_Bug_List
19:42:31Watusimotoyes
19:42:36Watusimototwoplayer.bat
19:42:39Watusimotoand sound stuff
19:42:46Watusimotobut the other things will require some explanation
19:43:02WatusimotoI guess #4 I should remmeber, as I think I reported it
19:43:32raptorit was some inconsistencies with starting new levels and random people would show with the '+', or not
19:43:35Watusimotomaybe I'll spend some time looking for sound
19:44:00raptorwe need to do client/server testing
19:44:03Watusimotoit's unpleasant, but someone needs to do it
19:44:11Watusimotowhat do you mean?
19:44:37raptorwe start a server and play through various levels
19:44:59WatusimotoI think big problems will be quickly evident
19:45:18raptoryes
19:45:29WatusimotoI guess what I shoudl have said was that I was running out of stuff I want to work on :-)
19:46:07Watusimotothough the list is tantalizingly short
19:46:40raptorhaha
19:46:42raptortoo true
19:46:59raptorsam686 and I usually go through some crazy levels to test things - want to do that with me?
19:47:21Watusimotosure, if lag isn;t so great as to make it a useless excersize
19:47:55raptoroh yeah - i keep forgetting you're in LU
19:48:13raptori can start a server on the master server instead - should reduce lag some
19:48:27Watusimotohow so?
19:48:34Watusimotois the master closer to europe than you are?
19:48:45raptorbecause then it isn't relaying packets through the master for punching through the firewall..
19:49:49Watusimotooh, I see... makes connection easier
19:50:03Watusimotothoguh it did work last time we tried on your machine, thoguh I thought it wouldn't
19:50:18raptorwant to just try my machien first?
19:51:03Watusimotosure
19:51:25Watusimotolisten to swosh-22 here
19:51:26Watusimotohttp://www.freesound.org/browse/
19:51:26raptorserver is up - unless you have something to checkin and i should recompile..
19:51:29Watusimotook
19:51:39Watusimotoeverything is in
19:51:54Watusimotosomething like that for firing a seeker?
19:52:45raptori can't seem to get at it..
19:52:56raptornew browser..
19:53:03Watusimotosounds like someone whipping a stick through the air
19:53:09raptorok
19:53:11raptornot too bad
19:53:22Watusimotosamtest2
19:53:28raptorno
19:53:30Watusimotoor ping timed out?
19:53:31Watusimoto:-)
19:53:51raptorthe ping timed out
19:54:22Watusimotoor 2nd or 3rd of these
19:54:23Watusimotohttp://www.freesound.org/people/man/sounds/14609/
19:54:54Watusimotocan't join the server
19:55:11raptor2nd isn't so bad
19:55:20raptorok, setting up on master...
19:56:24raptorwill take a few min. to compile
19:56:31Watusimotook
19:56:53raptoryay for Linux and non-backwards ABI compatibility!
19:57:12Watusimotoor this for firing seeker
19:57:13Watusimotohttp://www.freesound.org/people/Walter_Odington/sounds/25649/
19:57:37raptorthat's not bad
20:01:58raptorrunning into OOM problems compiling...
20:02:28raptormaybe i should compile with only one thread
20:07:08Watusimotostill compiling?
20:07:36raptorugh - i compiled a release build by accident - about 1/2 way done with debug buld now
20:07:44raptorerr 90%
20:07:49raptordone!
20:07:52raptordebug is MUCH faster
20:07:58raptor(compiling)
20:08:41raptorOK
20:08:43raptorit's up!
20:25:14raptorsam686: you're welcome to join us
20:25:18raptortesting on a server
20:37:25Watusimotolost connection to server
20:37:39raptoryeah it died... i wonder if because i was in an ssh session
20:38:09raptorok, testing good for now?
20:38:15raptoror want to test more?
20:39:02Watusimotook for now
20:39:46raptororbitbot borken
20:39:47raptori fix
20:40:11LordDVG Quit (Remote host closed the connection)
20:40:58Watusimotois et a player or an artifact?
20:41:06raptorplayer
20:41:13raptori've seen him/her before in-game
20:41:29Watusimotowell, they were lurking in the 018 lobby
20:41:38Watusimotounless all lobbies aer shared?
20:41:45raptorlobby protocols go waaaay back
20:41:49raptorshared
20:42:09raptormaster is fully backward compatible
20:43:12BFLogBot Commit: 28650adf07aa | Author: watusimoto | Message: Better disconnect message
20:50:11koda has joined
20:56:46LoneWolfy has joined
20:56:53LoneWolfyhi
20:57:03raptorargh too much debug output!
20:57:05raptorhi
20:57:12LoneWolfyIt says your onine raptor, i dont see you
20:57:21raptor018 testing... :)
20:57:23LoneWolfy018 testing?
20:57:25LoneWolfy:o
20:57:35LoneWolfyCan i get 018 beta?
20:57:37raptoryes
20:57:39raptoruhh
20:57:42raptorwell
20:58:19raptorwe're pretty close to release...
20:58:22LoneWolfycool
20:58:26raptor(we always say that...)
20:58:36LoneWolfy(Oh... Ok then.)
21:03:07LoneWolfybrb
21:04:50LoneWolfy_ has joined
21:04:55LoneWolfy_ok im back
21:04:59LoneWolfy_:O
21:05:07raptorhi
21:05:10LoneWolfy_old me will auto logout right?
21:05:23raptornot unless you closed the program
21:05:26LoneWolfy_i closed that window
21:05:38raptorWatusimoto: where is the output for the moving testitems... it's so noisy
21:05:48raptorthere is console output drowning out my bot tests
21:06:07Watusimotodid I not remove the debugging juink in the console?
21:06:28Watusimotolet me do that now
21:07:08raptorFYI for anyone interested - don't rub your eye when your finger has peppermint oil on it
21:07:23LoneWolfy Quit (Ping timeout: 245 seconds)
21:08:03Watusimotocrashed audacity twice creating this stupid sfx
21:08:18Watusimotogood advice with the oil. avoid vinegar and tobasco too
21:10:44Watusimotooops
21:10:50Watusimotothat code wasn't supposed to be checked in
21:10:55raptori still can't find that print...
21:11:25Watusimotoit's in griddb
21:11:29Watusimotoand I removed it
21:11:31BFLogBot Commit: 60ce24924ccd | Author: watusimoto | Message: Remove debugging code
21:11:42Watusimotoare you trying to fix orbitbot?
21:11:48Watusimotobecause I'll bet I know the problem
21:12:12raptoryeah - but i'm using it as an exercise in the new Lua objtypes
21:12:19Watusimotosimple edit
21:12:38Watusimotochange someting like shiptype to ObjType.ship
21:12:43raptorok
21:13:11Watusimotoyou could run the document generator and see the enum docs
21:13:15Watusimotothey're almost readable
21:13:18Watusimotobut not quite
21:13:27raptorheh
21:13:43Watusimotoin the meantime, see BfObject.h line 59
21:13:54Watusimotothird column
21:13:58WatusimotoObjType.xxxx
21:14:05raptorok
21:14:07raptorthanks
21:14:20Watusimotothe code is much more readable, I thin
21:14:21Watusimotok
21:15:55BFLogBot Commit: e75b8e60f9c9 | Author: buckyballreaction | Message: Fix orbitbot
21:16:06raptorcan we still do subscribe(MsgReceivedEvent) ?
21:16:43raptorwait
21:16:45raptorsame problem
21:16:54WatusimotoI think so
21:17:00Watusimotowhich problem?
21:17:08WatusimotoI didn;t change any events
21:18:52BFLogBot Commit: 1ee8eebb622a | Author: buckyballreaction | Message: Fix our friend Eliza
21:21:29raptorsorry sam686, i pressed ctrl + c in the wrong window..
21:22:11raptorWatusimoto: i got this warning: Finding objects will be far more efficient if your script provides a table -- see scripting docs for details!
21:22:28Watusimotoah
21:22:33Watusimotoyes... it will
21:22:34raptorwhat was the fix for that?
21:22:42Watusimotowhat's your lua statement?
21:22:55Watusimotoor does it not tell you?
21:22:56raptoroh yeah
21:22:59raptoritems = { }
21:23:03raptorcreate a global table
21:23:04raptorok
21:23:08raptor(found it in s_bot
21:23:11WatusimotofindObjects(items, blah)
21:23:13Watusimotoyes
21:23:26Watusimotomaybe not useful?
21:23:40Watusimotoit's only more efficient if you reuse the table
21:23:52Watusimotoif you use it once, might as well let c++ create it for you
21:26:25raptororbitbot uses it once; retrieve bot, over and over
21:27:21BFLogBot Commit: 326c990708c0 | Author: buckyballreaction | Message: Fix retrieve bot
21:27:29LoneWolfy_i'm playing in a game on Steam
21:27:46LoneWolfy_im on the built in browser :P
21:30:29BFLogBot Commit: 883f35629b0d | Author: buckyballreaction | Message: Fix AegisBot
21:32:42raptorhaha
21:32:53raptor3 aegisbots vs 5 seeker turrets
21:33:01raptoraegisbots win hands down
21:34:06raptor1 against 2 close range: http://sam6.25u.com/upload/5screenshot_0.png
21:36:02raptor1 against 5: http://sam6.25u.com/upload/5screenshot_1.png
21:38:14Watusimotothe seekers tend to come straight in, so if you can line it up, you'll hit every time
21:38:35Watusimotook, got new sfx... need to see if they are good
21:38:48raptorbetter shot: http://sam6.25u.com/upload/8screenshot_2.png
21:38:51raptorok
21:40:34Watusimotothat's a very cool shot
21:41:03Watusimotowhy don;t the turrets aim at the target?
21:41:49raptoruhh weird
21:41:55raptorhere they are: http://sam6.25u.com/upload/3screenshot_4.png
21:42:01raptorthat's the second screenshot glitch
21:43:27raptorsam686: any reason why screenshots would produce these two glitches?:
21:43:29raptor1. http://sam6.25u.com/upload/4screenshot_1.png
21:43:38raptor2. http://sam6.25u.com/upload/8screenshot_2.png
21:43:45raptor#2 look at the turrets not aiming
21:44:47sam686http://sam6.25u.com/upload/4screenshot_1.png - looks like it forgot to clear the screen when rendering a smaller screenshot..
21:47:01sam686umm, my bitfighter 018 freezes trying to switch to fullscreen... on a wrong monitor in a dual monitor screen
21:47:14LoneWolfy_ Quit (Quit: Page closed)
21:47:18raptorha: http://sam6.25u.com/upload/5screenshot_5.png
21:47:25raptorso screenshots from full screen are failing now
21:47:44raptorthat's no good
21:49:50sam686that screenshot problem also happens when your windowed size is quite bigger then 800 x 600
21:52:56Watusimotowow
22:03:08raptorsam686: would you happen to know anything about how that bug may have been introduced? if not, i'll start a bisect...
22:05:56sam686when both monitor is at same resolution, it switches to fullscreen on second monitor ok, but switch to windowed has a missing title bar http://sam6.25u.com/upload/second_monitor_problem.png
22:06:07sam686then trying to switch fullscreen again just freezes
22:06:58sam686another problem, when primary monitor has higher resolution that second monitor can't set to, title bar disappears and freezes too
22:07:13sam686and it always freezes somewhere inside SDL_PullEvent
22:09:35Watusimotoso seekers make sound when firing
22:09:49WatusimotoI am also trying to make ambient sound when they are nearby
22:09:56Watusimotobut I can;t make that work
22:10:19Watusimotonever mind
22:10:25Watusimotothe sound file is corrupt
22:10:31raptorha
22:11:41sam686the same fullscreen multiple monitor freezing problem also happens on windows XP, but also with severe performance problem with multiple graphics card with very slow bandwidth old PCI card when transferring every single pixel from one graphics card to another....
22:12:10raptorhmm... want me to compile a new SDL binary to have you try?
22:13:11sam686actually I don't get any poor performance problem if I keep that opengo windows on the same graphics card (which output 2 monitor) on my windows xp..
22:13:56sam686could try newer SDL...
22:18:20sam686will be back in about 30 minutes..
22:24:04raptorsam686: newest SDL2 debug build: http://sam6.25u.com/upload/SDL-b61791cd595a.7z
22:38:43WatusimotoThanks for fixing the bots, btw
22:38:44BFLogBot Commit: e711d40cf5b1 | Author: watusimoto | Message: New seeker shooting sfx
22:38:45BFLogBot Commit: f4955a3c7392 | Author: watusimoto | Message: Merge
22:39:12raptori have to figure out why two frames are being rendered on the opengl back buffer...
22:39:21raptorI *think* that's what's happening
22:54:55sam686tried that newer SDL2, it seem to fix one of 2 problem I had (multiple monitor using same resolution) no longer freezes, not when I switch from fullscreen to windowed, it freezes the game (stuck inside SDL_PullEvent) till I manually make the window smaller to unfreeze..
22:55:54sam686But, that still doesn't fix a second problem: doing fullscreen on second monitor tried to switch to higher non-existant resolution that the primatr monitor resolution set at, though looks different now..
22:55:58Watusimotook, getting hypnotized watching test items bounce around the screen
22:55:59BFLogBot Commit: 1ddf9e4d2ca4 | Author: buckyballreaction | Message: Fix screenshots taken from fullscreen. Need to clear the buffer now for some reason
22:56:00raptoranother bug bites the dust
22:56:06Watusimotogreat
22:56:10raptoryou're getting sleepy...
22:56:13Watusimotonot sure how to really reproduce
22:56:25WatusimotoI only see weird things when testitems go really slow
22:56:36Watusimotoand i think they;ve always been that way
22:57:34sam686oh and another SDL2 (maybe not) problem, when maximized in windows 7 (with aero off) using 1280 x 1024, it freezes, blinking "Bitfighter 018" on the title bar, but can unfreeze when un-maximizing it..
22:57:49raptorcould it be that it's moving so slowly that server doesn't detect a difference and so doesn't update the position?
22:58:06raptoror some sort of float rounding error client-side
22:58:10WatusimotoI recall from years ago that there's a lower threshold for somethings for some reason
22:58:30Watusimotobut is the bug that was reported only for really slow test items?
22:58:38raptori saw it happen for resources
22:58:46raptorand soccerballs
22:58:48Watusimotobut slow v fast
22:58:56Watusimotoit would be the same for all moveitems, I would expect
22:59:07raptorlet me try to dupe again..
22:59:26Watusimotojust one testitem?
22:59:59sam686I wonder if I should try to make SDL_PullEvent run in a seperate thread from game/render thread... which might fix "Freeze while dragging bitfighter on windows"
23:00:30sam686not only freeze, but also lag all connected players when your client is hosting..
23:00:33raptorso i see the slow motion corrections
23:00:39raptorbut i remember it was more pronounced
23:00:57raptorlike even when moving fast, there'd be large trajectory changes
23:01:15Watusimotoit's possible I fixed it while working on soccer
23:01:22raptorwhoa
23:01:25raptorstill happens
23:02:47raptoruse pulse to hit a resource item or testitem
23:02:56raptorand watch change like there was lag
23:04:43Watusimotoone player?
23:06:19Watusimotomaybe saw it when it was hitting my ship
23:07:00raptormy test level: http://pastie.org/5362799
23:07:23raptoralso a random forcefield showed up: i think it was loaded from a previously loaded editor level that had one
23:09:58raptorsam686: running the render thread separate from the window thread fixes this problem: http://stackoverflow.com/questions/7450156/c-sdl-application-pauses-when-dragged
23:10:18raptorbut i don't think we should multi-thread the application for 018
23:11:08Watusimotoconfirmed the ff bug
23:12:51raptori think the laggy effect always seems to happen shortly after collisions
23:15:04WatusimotoI can only see it rarely when object collides with ship
23:15:43raptori've seen it lots when a resource item and testitem collide, but barely touching
23:15:55raptorlike the resource item flies at it almost on a tangent
23:17:01Watusimotocan you construct a level that directs testitems at a resource in just the right way?
23:17:14raptorgood idea, let me try...
23:17:28WatusimotoI'm trying too
23:18:32WatusimotoI can reliably make it "hop" a tiny bit
23:18:36raptorsegfault!
23:18:49Watusimotoexcellent!
23:20:06raptorselect two barriers, press '-' to shrink their width
23:20:30raptorstack: http://pastie.org/5362839
23:20:33WatusimotoI'll add it to the wiki
23:20:48Watusimotoit crashes for me too
23:21:01raptorerrr, just one barrier does it, too
23:21:44Watusimotook
23:22:01Watusimotobug list is going the wrong way
23:22:05Watusimotobigger
23:22:41Watusimotook, I see the reason for the crash
23:23:17WatusimotoIt's an incorrect comment
23:23:18Watusimoto/ Walls are a subclass of LineItem, so this will work for both
23:23:22Watusimotono longer true
23:23:26Watusimotohence the crash
23:31:33BFLogBot Commit: 62ab3f88d789 | Author: watusimoto | Message: Style
23:31:35BFLogBot Commit: 1e494edb6ffc | Author: watusimoto | Message: Fix wall thickness editing in the editor
23:31:36BFLogBot Commit: 11148a854657 | Author: watusimoto | Message: Merge
23:34:07raptorthis lag problem may be totally unrelated - like could be a frame every-once-in-a-while that takes longer than normal
23:37:10Watusimotowell, I was able to reproduce a very minor variant of the effect reliably
23:37:24raptori'm still working on reliability...
23:37:32Watusimotowith a test item hitting a resource obliquely going very slow
23:37:39Watusimotoit would skip a bit after the collision
23:37:59Watusimotounfortunately, I lost the level testing the wall width bug
23:38:07Watusimotobut I may be able to reproduce it
23:38:26Watusimotoif I don;t fall asleep first, that is
23:43:34raptorsigh - i thought i reproduced, then it failed on the third try...
23:54:34raptorok giving up on this for now
23:54:46raptorhaven't reliably duplicated
23:56:23sam686Some of the problem may be caused by pack/unpackUpdate sending a approximate Position & velocity instead of being exact..

Index Search ←Prev date Next date→

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