#bitfighter IRC Log

Index Search ←Prev date Next date→

IRC Log for 2012-12-28

Timestamps are in GMT/BST.

00:06:25YoshiSmb_ has joined
00:08:02YoshiSmb Quit (Ping timeout: 244 seconds)
00:17:24YoshiSmb_ is now known as YoshiSmb
00:19:29raptorbobdaduck, wh bug with burst did you say you found?
00:20:29bobdaduckTestitems and bursts
00:20:32bobdaduckmake an empty level
00:20:34bobdaducktestitem
00:20:39bobdaduckburst it
00:20:44bobdaduckit only gets x velocity
00:20:50bobdaduck(or something like that.)
00:21:33raptorha!
00:21:37raptoryou're right!
00:21:41raptorok writing down..
00:24:51bobdaduckIts been that way forever
00:24:56bobdaducklike, literally, forever.
00:25:00bobdaduckeven Zap! had that.
00:25:07raptorwait what??
00:25:11raptorreally?
00:25:44bobdaduckyeah
00:27:52bobdaduckokay
00:37:37bobdaduckI set a healthitem to "no regen"
00:37:41bobdaduckand its still regening.
00:41:11raptoryou mean it keeps coming back?
00:41:17bobdaduckyeah.
00:42:34bobdaduckat "no regen" it spawns faster than one set at 100 regen.
00:43:32raptorha! confirmed
00:43:36raptorwriting down...
00:44:05bobdaduckWant to find bugs? Make a dungeon.
00:44:38raptorheh
00:44:39raptorso true
01:08:44YoshiSmb Quit (Ping timeout: 252 seconds)
01:36:13YoshiSmb has joined
02:08:22YoshiSmb Quit ()
02:09:33sam686 Quit (Ping timeout: 245 seconds)
02:35:47bobdaduck Quit (Quit: Page closed)
02:48:07sam686 has joined
02:48:10ChanServ sets mode +v sam686
03:05:32CrazyLinuxNerd has joined
03:26:32raptorok, i have a few min for bugfixing...
03:42:21LoneWolfy has joined
03:42:27LoneWolfyraptor...
03:42:44LoneWolfyi saw you were just on and now you're not >.>
03:43:14raptorhi
03:43:19raptori'm fixing bugs, sorry
03:43:21LoneWolfyoh
03:43:23LoneWolfy:P
03:43:29LoneWolfyi found one i think
03:43:52LoneWolfywith fullscreen streched, i enter level editor and exit, and the view is warped
03:44:31raptorah yes
03:44:35raptorthat is fixed for 018a
03:45:45raptorsam686: are you working on any bugs?
03:45:49LoneWolfygood
03:46:04LoneWolfyanyway, just checking in
03:47:00raptorgreat!
03:47:24raptorsay, did you enter in a level for the contest?
03:50:17LoneWolfy Quit (Ping timeout: 245 seconds)
03:59:32raptorsam686: do bursts push other ships for you?
03:59:59raptoroh, maybe it is just eliza...
04:01:33raptorsingle player retrieve doesn't allow team damaging?
04:01:37raptorhas that always been so?
04:39:51CrazyLinuxNerd Quit (Quit: Leaving)
05:43:00raptor Quit ()
05:43:55kodaws has joined
05:49:58kodaws Quit (Ping timeout: 244 seconds)
06:20:49bobdaduck has joined
06:43:52fordcars has joined
06:44:28fordcarsNice, I got internet working
06:44:46fordcarsmy internet
06:49:03fordcars Quit (Ping timeout: 245 seconds)
07:38:56sam686 Quit (Quit: :))
07:51:04bobdaduck Quit (Quit: Page closed)
08:08:49kodaws has joined
08:14:46Watusimoto has joined
08:30:24CrazyLinuxNerd has joined
08:45:51BFLogBot Commit: 8a80f7569664 | Author: watusimoto | Message: Make ids more visible. Fully working now except for walls/polywalls, which will have to wait
08:45:52BFLogBot Commit: daab170a05fa | Author: watusimoto | Message: Merge
09:04:20BFLogBot Commit: aae96b83dcab | Author: watusimoto | Message: Get rid of dynamic cast
09:04:22BFLogBot Commit: 10021c37c4e3 | Author: watusimoto | Message: Fix flag "Left Behind" issue
09:08:58Watusimoto Quit (Ping timeout: 248 seconds)
09:44:11watusimoto has joined
09:44:12ChanServ sets mode +o watusimoto
09:46:10CrazyLinuxNerd Quit (Quit: Leaving)
09:49:20LordDVG has joined
10:46:44CrazyLinuxNerd has joined
10:55:29Invisible has joined
10:56:04CrazyLinuxNerd Quit (Quit: Leaving)
10:57:05Invisiblewatusimoto: when will you be able to play on the contest server??
10:57:27watusimotouh, hi
10:57:33Invisiblehi
10:57:39watusimotojust a tip: type w and hit <tab>
10:57:55Invisiblewatusimoto: great
10:57:56watusimotoin any event, I can't play from work
10:58:08Invisiblei know
10:58:18watusimotoso... tonight?
10:58:37watusimotoor are you asking when the contest server will be up?
10:58:37Invisibleyeah
10:58:47Invisibleboth
10:58:56watusimotodon't know about when it will be up
10:59:04Invisibleok
10:59:14watusimotolater today... it needs to be running by 8PM EST tonight for sure, so before then
10:59:35Invisiblegonna work on the game
10:59:44watusimotoexcellent!
10:59:46Invisiblebye
10:59:50watusimototschuss
11:10:14CrazyLinuxNerd has joined
11:12:19Invisiblehi
11:13:20watusimotohi
11:17:27watusimotoso Invisible: often people will sign in when their computer turns on; doesn't mean there is someone there who wants to chat
11:20:10LordDVG Quit (Remote host closed the connection)
11:22:08Invisibleoh
11:22:38Invisiblei'm still workin' on the game...
11:24:44Invisiblesome of the code is cut off, so i'm going to need help with that
11:25:58watusimotowhich lines do you need (paste me what you have)
11:26:09watusimoto(or at least enough of what you have that I can find it)
11:26:54Invisibleself.image = self.font.render("%d %d"%(self.leftscore, self.r))
11:27:09Invisibleself.image = pygame.image.load(os.path.join('images','pong_ball.g'))
11:27:46watusimotoself.image = self.font.render("%d %d"%(self.leftscore, self.rightscore), True, (0,0,0))
11:28:57watusimotoself.image = pygame.image.load(os.path.join('images','pong_ball.gif'))
11:28:59watusimotoprobably
11:29:08Invisiblethanks
11:29:24watusimotoyou'll also need:
11:29:25watusimotocollided = pygame.sprite.spritecollide(self.ball, [self.leftpaddle, self.rightpaddle], dokill=False)
11:30:13watusimotoyes, that 2nd line was correct
11:30:30Invisibleok
11:30:43Invisiblebrb
11:30:52Invisible(soon)
12:04:44Invisibleit runs
12:05:23Invisiblebut the ball will not appear
12:05:46watusimotowhy not?
12:06:09Invisiblewatusimoto...
12:06:24Invisibledont know
12:06:56Invisiblethere are no errors
12:07:21Invisiblei finished step 2
12:07:25watusimotodo you have to serve it?
12:07:28Invisible...
12:07:46Invisibledont know
12:07:55watusimotoread the text; it explains how to play
12:08:08Invisiblespace maybe
12:08:10watusimotoyou need to serve the ball for it to appear
12:08:17watusimotolook at your code -- it will tell you
12:08:29Invisiblewhich part??
12:08:51watusimoto # serve with space if the ball isn't moving if event.key == K_SPACE: if self.ball.velx == 0 and self.ball.vely == 0: self.ball.serve()
12:08:54watusimotooops
12:09:16Invisiblethanks
12:09:19watusimotoit checks to see if you pressed space; then if the ball is not moving; if all that is true, then it calls serve
12:10:16Invisiblespace not work...
12:10:35watusimotook, well you need to figure out why not
12:10:56watusimotofirst thing to do is to see if serve is being called
12:11:04watusimotoself.ball.serve()
12:11:45watusimotofind the serve function (it will be in your ball object), and put a line near the top like this:
12:12:25watusimotoprint("Serving the ball!")
12:12:31watusimotoand see if that gets printed
12:12:52watusimoto(it should get printed on the little console at the bottom of the editor)
12:13:06watusimotoor if you are lazy, just print("XXX")
12:14:23Invisiblenope
12:14:35watusimotook, well, that is your problem
12:14:38Invisiblethere is no ball
12:14:42watusimoto(or at least your first problem)
12:14:51watusimotoThere is no ball because you are not serving it
12:14:59watusimotoso now you need to figure out why that is
12:15:01Invisiblethanks
12:15:26Invisibleeverything is wright
12:15:40Invisiblecorrect
12:15:51watusimoto"right"
12:16:05Invisiblethanks...
12:16:10watusimotoso move your print statement to just under where you detect the space being pressed
12:16:20watusimotosee if the system is even responding to that
12:16:38watusimotowe're trying to isolate the problem
12:17:18watusimototell you what -- since this is supposed to be bitfighter related, change your irc room to #pong
12:18:22Invisible has left
12:34:00Invisible_ has joined
12:51:48kaen Quit (Read error: Connection reset by peer)
12:53:34Invisible_bye
12:53:55Watusimoto_ has joined
12:57:47Invisible_ Quit (Ping timeout: 245 seconds)
13:07:58kaen has joined
14:35:41Watusimoto_ Quit (Ping timeout: 265 seconds)
15:18:01raptor has joined
15:18:01ChanServ sets mode +o raptor
15:18:08raptorbuenos
15:21:29raptoroh yay, i wake up to 3 new PMs..
15:23:16raptoranother level! great!
15:23:35raptorwatusimoto: you need to get me one (and I need to get me one)
15:24:03watusimotohi
15:24:16raptorhi
15:24:19watusimotomy levels invariably suck
15:24:47raptoryes, well... it's not a high bar.. :)
15:26:23watusimotoI see... :-)
15:26:53kaenin soccerGame, the home timer is an attribute of SoccerBallItem. If I wanted a timer shared between all instances of an in-game object how might I call its update() method?
15:27:06raptori've spend too much time writing bobdaduck's scripts to be able to do anything complex..
15:27:19kaenin that case putting it in the item's idle() method would over-increment it, right?
15:27:24raptorhi kaen
15:27:28kaenhello :)
15:27:38watusimotohi
15:27:38raptoris there an idle() method there somewhere?
15:27:43kaengood morning, pardon my technical question so early :)
15:27:52raptorbrain... can't... compute...
15:27:55watusimotowhy do you want to share the timer?
15:28:25kaenthere's a synchronized event happening and multiple instances of the item
15:29:07watusimotoso just so I'm clear... when a goal is scored, the timer is fired, right? then all balls go back to the start?
15:29:26kaenit's set and when it fires they go back
15:29:45watusimotoso all balls are kind of sendHoming timer at the same time, so why have multiple timers?
15:29:45kaenso each ball keeps it's own timer... d'oh
15:29:54watusimotois that where you're going?
15:30:02kaenindeed
15:30:28kaenbut I just realized that all I have to do is set the timer for each instance at once :x
15:30:51kaenso early...
15:30:58watusimotoso currently if we have two soccer balls, it looks like only the scoring ball goes home
15:31:05watusimoto(juding by the code, that is)
15:31:25raptorkaen: if i may be so bold to ask - what are you doing? making a level for the contest? hacking you're own server? :)
15:32:00kaenwell I'm playing with a swarm-style game mode
15:32:48raptora new game mode? cool...
15:32:53kaen shrugs
15:33:35watusimotoin soccerGame.cpp, there is a block that begins:
15:33:35watusimoto else if(hitObject->getObjectTypeNumber() == GoalZoneTypeNumber)
15:34:00watusimotoyou could get all the balls and essentially run that block on them
15:34:13kaenindeed
15:34:42kaenI got a lot of clarity out of that conversation, thanks :)
15:34:47watusimotoactually, just set the timer on all of them and you should be good
15:35:41kaenI also wanted to understand TNL and how its used in bitfighter. I've already learned a lot about NetClasses and the RPCs
15:36:48watusimotowell, then you're way ahead of us!
15:37:00kaenlol
16:05:34raptorsooo... did all of our GCI folk take off for vacation?
16:09:00raptorwatusimoto: slight problem - i can't compile dedicated on master server, so where should i host the contest server?
16:10:03watusimoto why can't you compile there?
16:10:09watusimotooh, because of what we discussed
16:10:11watusimotoyesterda
16:10:13watusimotoy
16:10:18raptorthat infuriating linker error
16:10:31raptori'm going to try again now..
16:10:42watusimotowhat about upgrading gcc?
16:10:47raptorhaha
16:10:49raptorupgrading gcc
16:10:51raptorhaha
16:10:52watusimotohaha?
16:10:52raptorhahahaha
16:11:02watusimotosudo apt-get install gcc
16:11:07watusimotoor yum or whatever
16:11:11watusimotothat won't work?
16:11:16raptoryeah, ummm.... it's already at the latest i could find for centos 6
16:11:32raptorand, trying to recompile a newer version is death
16:11:39raptor(already tried)
16:11:40watusimotobut it compiles on your home machine
16:11:43raptoryes
16:11:50raptorhmm... maybe
16:11:58watusimotoso you are convinced it is a gcc version issue?
16:12:00raptormaybe i could transfer a binary
16:12:36watusimotowell, try building again on the master, and if it still doesn't link, I'll take a look. If none of that works, what are our options?
16:12:44raptorit's more like there is probably something we're doing wrong in the code, but new gcc is more forgiving
16:13:00raptoroptions: transfer a binary and hope it runs...
16:13:35watusimotolet me try building here
16:13:45raptorohhh.. it's been a few weeks since 018, maybe some random distro had a smart packager that fixed the linker errors..
16:14:08raptori've had that happen several times before for other projects..
16:18:26watusimotomake dedicated?
16:18:39raptori'm updating the project on master now...
16:18:44raptoryes
16:18:53raptorit could be dedicated doesn't compile... let me check
16:20:03watusimotowon't even get to the linker for me
16:20:11raptoryeah... let me fix..
16:20:17watusimotorabbitGame.cpp:229:60: error: invalid static_cast from type ‘Zap::Game*’ to type ‘Zap::ClientGame*’
16:20:29raptorhaha
16:20:38watusimotothat's a valid cast
16:20:44raptornot for dedicated
16:20:48raptorno ClientGame.h
16:20:58watusimotooh
16:22:42raptorhuh... i'm not sure how to fix this one..
16:22:46raptorifdef somewhere?
16:24:40watusimotowhat line is it? Ship *localShip = static_cast<ClientGame *>(getGame())->getLocalShip(); ?
16:24:52raptoryes
16:25:07watusimotogetTeamColor never need run on server
16:25:23watusimotoso you could ifdef the whole thing out and replace it with return Colors::white or something
16:26:21raptorthat was my dirty fix..
16:26:34watusimotoor, perhaps cleaner would be to ifdef everything out except the final return
16:26:59BFLogBot Commit: 0973caf2a616 | Author: buckyballreaction | Message: Fix compiling dedicated
16:26:59watusimotowhatever
16:27:02raptorthat is exactly it..
16:27:36raptorok compiling on master
16:28:32watusimotobuilds for me here
16:29:26raptorwhat version gcc?
16:29:45watusimotogcc (Ubuntu/Linaro 4.6.3-1ubuntu5) 4.6.3
16:30:01watusimotoprobably newer than whatver is on bf.org
16:30:02raptorit's the entire 4.4 line that seems to fail, maybe earlier
16:30:09raptor4.4.6 is what is on master
16:30:17watusimotostill failing?
16:30:23raptorran out of RAM..
16:30:27watusimotowhoa!
16:30:58watusimotois there a way I can install an older gcc?
16:31:17raptoryes, usually under the package guise of gcc-4.5 or gcc-4.1
16:31:40raptorbut then you'd have to change the Makefile to use gcc-4.5 instead of just gcc
16:32:54-pratchett.freenode.net- *** Looking up your hostname...
16:32:54-pratchett.freenode.net- *** Checking Ident
16:32:54-pratchett.freenode.net- *** No Ident response
16:32:54-pratchett.freenode.net- *** Couldn't look up your hostname
16:32:55[frigg VERSION]
16:33:00BFLogBot has joined
16:33:00Topic is 'Bitfighter 018 released! | http://www.bitfighter.org/downloads | Welcome GCI students! Please introduce yourself and feel free to ask any questions.'
16:33:00Set by raptor!~raptor@unaffiliated/greenmachine on Thu Dec 06 10:11:09 GMT 2012
16:33:00-ChanServ- [#bitfighter] Welcome to #bitfighter. This is an IRC channel, many or all of the users may not be paying attention. Please have patience when waiting for a response.
16:33:14watusimotogcc-4.4-arm-linux-gnueabi - GNU C compiler
16:33:14watusimotogcc-4.4-arm-linux-gnueabi-base - GCC, the GNU Compiler Collection (base package)
16:33:14watusimotogcc-4.4-arm-linux-gnueabihf - GNU C compiler
16:33:14watusimotogcc-4.4-arm-linux-gnueabihf-base - GCC, the GNU Compiler Collection (base package)
16:33:14watusimotogcc-4.4-locales - GCC, the GNU compiler collection (native language support files)
16:33:14watusimotogcc-4.4-multilib - GNU C compiler (multilib files)
16:33:14watusimotogcc-4.4-source - Source of the GNU Compiler Collection
16:33:19watusimotoare my 4.4 options
16:33:25watusimotonone look appealing
16:33:30raptorhuh
16:35:08watusimotowell, despite that, sudo apt-get install gcc-4.4 seems to install something
16:35:21raptorha
16:39:05raptoryep, linker error still: http://pastie.org/5589478
16:39:31raptorit's almost as thought std::vector::end() isn't found...
16:39:42raptori've tried rearranging the compilation units
16:40:02raptorchanging headers (adding lots of them) to the failing classes
16:40:06raptornothing compiles it
16:40:43raptor*links
16:41:55watusimotowhat do you make of this?
16:41:56watusimotogcc-4.4: error trying to exec 'cc1plus': execvp: No such file or directory
16:42:07raptorahh, you need g++-4.4
16:42:21raptorWHAT
16:42:27raptordedicated_debug compiles on master
16:43:00watusimotocan we host on sam's machine?
16:43:09watusimotoor wherever sam hosts?
16:43:17raptormaybe... if he's around..
16:43:38YoshiSmb__ has joined
16:43:45watusimotoor host with the _debug version?
16:43:50watusimotobut what's with that?
16:43:58YoshiSmb__hi.
16:44:02watusimotohi
16:44:17YoshiSmb__ is now known as Yoshismb
16:44:39watusimotoundefined reference to `std::vector<Zap::Point, std::allocator<Zap::Point> >::end() const'
16:44:43watusimotoblech
16:47:01raptorok, compiling with -g works, but -O2 doesn't
16:50:19watusimotointeresting
16:51:52-pratchett.freenode.net- *** Looking up your hostname...
16:51:52-pratchett.freenode.net- *** Checking Ident
16:51:52-pratchett.freenode.net- *** Couldn't look up your hostname
16:51:52-pratchett.freenode.net- *** No Ident response
16:51:53[frigg VERSION]
16:51:58BFLogBot has joined
16:51:58Topic is 'Bitfighter 018 released! | http://www.bitfighter.org/downloads | Welcome GCI students! Please introduce yourself and feel free to ask any questions.'
16:51:58Set by raptor!~raptor@unaffiliated/greenmachine on Thu Dec 06 10:11:09 GMT 2012
16:51:59-ChanServ- [#bitfighter] Welcome to #bitfighter. This is an IRC channel, many or all of the users may not be paying attention. Please have patience when waiting for a response.
16:52:26[Yoshismb VERSION]
16:53:34raptorha, can't compile with both -g and -O2
16:53:38raptorout of memory
16:58:26Yoshismb Quit (Ping timeout: 245 seconds)
16:59:36watusimotowhat is -g?
16:59:43raptordebugging symbols
17:00:00raptorso badically it's -g -O0
17:00:06watusimoto-O1 doesn't work here, nor does -O3
17:01:42raptorsetting the heap size for gcc...
17:03:34raptorso the question is what optimization is it that breaks it...
17:04:48watusimotoI franklly don't even understand the underlying message
17:05:02watusimoto undefined reference to `std::vector<Zap::Point, std::allocator<Zap::Point> >::end() const
17:05:12watusimotolooks like an STL problem
17:05:15raptoryes
17:05:29raptorlike stl::vector isn't properly compiled
17:07:10raptorone of these:
17:07:12raptor-fauto-inc-dec -fcprop-registers -fdce -fdefer-pop -fdelayed-branch -fdse
17:07:13raptor -fguess-branch-probability -fif-conversion2 -fif-conversion -finline-small-functions
17:07:15raptor -fipa-pure-const -fipa-reference -fmerge-constants -fsplit-wide-types -ftree-builtin-call-dce
17:07:16raptor -ftree-ccp -ftree-ch -ftree-copyrename -ftree-dce -ftree-dominator-opts -ftree-dse -ftree-fre
17:07:18raptor -ftree-sra -ftree-ter -funit-at-a-time
17:07:33raptorif you 'man gcc' and search for /-O1, you'll see the optimizations it is using
17:10:07raptorwatusimoto: try -Os
17:12:29watusimotono luck
17:12:32watusimotosame error for me
17:12:41raptoryeah, i can't compile with -g and -O2...
17:12:43watusimotoI repoaced the -Ox flag in the zap/Makefile
17:12:44raptoron master
17:12:47raptornot enough ram
17:12:57raptortry adding -g
17:12:59watusimotocan you kill some proecess to create ram?
17:13:08raptori've done that several times...
17:13:16raptori can do -O2
17:13:17raptoror -g
17:13:20raptorbut not both..
17:13:42watusimotoso -Os -g or -O2 -g?
17:13:48raptoryes
17:13:51watusimotowhich?
17:13:55raptoror even -O1 and -g
17:13:58watusimotoI see
17:13:59raptorstart with -O1
17:14:19raptori want to see if it's optimizations or the lack of debugging symbols that breaks it
17:14:25watusimotook
17:15:09watusimoto-Os -g failed
17:15:22raptorok, looks like it's an optimization..
17:15:44raptorumm, are you doing a full recompile of make dedicated?
17:15:51watusimotono
17:15:54watusimotothis is the problem
17:16:07raptoryou'll need to at a minumum compile one of the classes that is failing
17:16:13raptorlike the ServerGame.cpp
17:16:36watusimotoI'm recompiling only zap
17:16:43raptorthat's fine
17:16:46watusimotois it?
17:16:49raptorbut all classes before linking?
17:16:50raptoryes
17:16:54watusimotomaybe I need to be recompiling tnl
17:16:59watusimoto?
17:17:07watusimotoI'm doing make clean; make dedicated
17:17:13raptorthat's fine
17:17:14watusimotoin the zap folder
17:17:15watusimotook
17:17:19raptorperfect
17:17:27raptorok, so, it's an optimization..
17:17:39raptorargh i hate these problems
17:17:44watusimotoindeed
17:25:23watusimotoso what if we compile unoptimized?
17:25:30raptorwe can do that
17:25:44watusimoto-O1 -g does not work
17:26:21watusimotoI don't think we have another choice at this point
17:29:36raptorok, i just compiled those three failing classes separately with -O2 and linked them in with all the other classes with -g
17:29:40raptorit linked fine
17:31:17watusimotomeaning ship.cpp and two others?
17:31:22raptoryes
17:31:28watusimotothat's interesting
17:31:33raptortesting somethings else..
17:31:38watusimotoso it's not a general failure with vector
17:31:43raptorno
17:31:59watusimotook, I've got to go
17:32:00raptormaybe the issue is with tnl..
17:32:02raptorok
17:32:31raptorthanks
17:32:43watusimotoone other idea might be to take one class, the shortest one perhaps, and start ripping out the guts of the functions until we find one that makes a difference
17:32:51watusimotobut that for a later time
17:33:01raptoryeah... i'm dissecting...
17:33:02watusimotoI think we can fix the compilation in principle
17:34:00watusimoto Quit (Quit: Leaving.)
17:34:01raptormaybe i'll bisect...
17:56:33BFLogBot Commit: 089a8a5304da | Author: buckyballreaction | Message: Fix compiling zap/ folder as debug against tnl optimized
18:06:25kodaws Quit (Read error: Connection reset by peer)
18:20:14kaen[100%] Built target bitfighter
18:20:22kaenso I have a working CMakeLists.txt
18:20:25raptorhi kaen
18:20:28raptorWHAT
18:20:31raptorreally?
18:20:34kaenindeed
18:20:38raptorthat was a goal of mine
18:20:45kaenit's a million hacky right now
18:20:51kaenbut the build runs :)
18:21:04raptorthat's great!
18:21:09raptorsooo...
18:21:12raptorLinux?
18:21:16kaenyep
18:21:18raptor(and what version of gcc?
18:21:21raptor)
18:21:33kaen4.6.1
18:23:06raptorhow are you at solving linker errors?
18:23:25kaennot too shabby, I just rattled through a bunch while working with cmake lol
18:24:02raptorbecause we're current'y stuck - we can't compile on systems with gcc 4.4 (maybe 4.5) and earlier because of a linker error we can't solve..
18:25:28kaenwith the latest mercurial?
18:25:56raptoryes, well with at least anything since the release of 018
18:26:08raptorbut you'd have to have gcc 4.4
18:26:21kaenI just installed it
18:26:25raptorhere is the error: http://pastie.org/5589478
18:26:38raptorit has to do with a global fillVector, i think
18:27:24kaenI'll see what I can do
18:27:29raptorbut i've spent so many hours on it, that i'd appreciate a new set of eyes...
18:27:52raptorso it only happens when you 'make dedicated'
18:27:59raptorand not 'make dedicated_debug'
18:28:17kaenoh wow that's a good one
18:28:17raptorbasically adding optimizations to gcc (-O2, whatever) is what kills it
18:28:23kaensure
18:28:38raptorbut gcc 4.6+ works
18:29:04Watusimoto has joined
18:29:13Watusimotohey raptor
18:29:18raptorhi Watusimoto
18:29:34Watusimotowe're about to eat, but I wanted to quickly check in and make sure everything is under control
18:29:41Watusimotoso... is everything under control?
18:29:43raptoruhh..
18:29:45raptorso
18:29:51raptori can start the contest server in debug if needed
18:29:59raptorbut i'm still trying to solve the linker errors
18:30:06raptorso sure, in control...
18:30:13Watusimotowhat about the hybrid some optimized some not optimized build you made?
18:30:26Watusimotosurely that would be better than debug mode?
18:30:43raptoryeah - that compiles, but i don't know if it would cause problems...
18:30:55raptormaybe i'll make two binaries and keep the debug one around just in case
18:31:41raptorWatusimoto: question - is this the revision we want to use?
18:32:23raptorthe most recent? also kaen has kindly offered to take a look at the linker error, too
18:32:43WatusimotoI have some more stuff to check in; I'll do that after dinner. So... no.
18:32:54raptornothing with TNL i hope?
18:32:56WatusimotoI fixed a couple of the errors on the list
18:32:57Watusimotono
18:33:01Watusimotonothing at all
18:33:02raptorok
18:33:10Watusimotonothing client-server, either
18:33:15Watusimotook, back in a bit
18:33:16raptorexcellent
18:33:17raptorok, thanks
18:33:19Watusimotothanks!!!!
18:36:51raptorok, actually, it looks like TNL_DEBUG is the difference... not -g or -O2
18:37:39Watusimoto Quit (Ping timeout: 252 seconds)
18:38:18raptorkaen: what linux distro do you use?
18:39:33kaen... ubuntu
18:39:38kaenat the moment
18:39:49kaenbecause I haven't fixed my busted arch install yet
18:40:44raptornarrowed it further: TNL_ENABLE_ASSERTS must be defined...
18:43:08raptori wonder...
18:47:14raptoroh my goodness
18:47:34kaen braces himself
18:47:45raptortnl/tnlAssert.h
18:48:08raptorthe #defines in the #3lse
18:48:10raptor#else
18:48:14raptorline 92
18:48:22raptorchanging both of those to:
18:48:23raptor #define TNLAssert(x, y) { }
18:48:25raptor#define TNLAssertV(x, y) { }
18:48:35raptorfixes everything...
18:49:12kaenD:
18:49:18kaenbut also :D
18:50:24kaenthe comment above it is pretty dubious...
18:50:32kaen // Tricky way of making sure asserts are completely compiled out
18:50:34raptoryeah... i made that change
18:50:41kaenoh heh
18:50:46raptori stole it from the SDL library
18:51:09kaenI see
18:51:23raptorbut it doesn't work for older gcc for some reason
18:51:29raptor researches more...
18:52:48kaencool, I'm going back to doing useless things then~
18:52:52raptorheh
18:52:59raptorthanks for your support!
18:53:07raptori think i just need a listening ear sometimes...
18:53:07kaenno sweat :)
18:53:08raptoralso
18:53:14raptorcmake would be handy..
18:53:25kaenI'll iron out the wrinkles
18:54:19kaenmy only motivation is the dependency scanning and selective compilation honestly
18:54:33raptordependency scanning rocks!
18:54:42kaenword
18:55:30kaenit'll save me from wondering if my runtime crash is a bug or a mistake in the linkage
18:55:49kaenwhich is how I got on the subject this morning...
19:01:02raptorcurse the linkage!
19:06:33kaenthat's what I try to tell my computer
19:06:38kaenbut no it still segfaults
19:06:44raptorha
19:09:56Watusimoto has joined
19:09:58kaenso there's no -llua in the current Makefile, does that meant that we always want to build against the lua in the tree?
19:10:19raptorwe link in liblua-vec.a
19:10:30raptorwe have a customized version of lua called lua-vec
19:10:36raptorthat is optimized for vectors
19:10:46raptorerr, the geometric kind
19:11:19kaenI see, and the same goes for tomcrypt, alure and tnl
19:11:34raptorcorrect
19:11:45raptoralure may exist in some distros - but i've customized it, too...
19:15:31kaenso should I link against system versions for the others? or always use the included source where possible?
19:16:40raptoreverything else system, except for libtomcrypt, tnl, lua-vec
19:16:47kaengotcha
19:17:23raptorlooks like alure, too.. for non-dedicated
19:28:17BFLogBot Commit: c580a937b35d | Author: eykamp | Message: Reduce code duplication with new isLocalPlayerShip() function
19:28:18BFLogBot Commit: cd9ffa62a4eb | Author: watusimoto | Message: Comment
19:28:20BFLogBot Commit: 807445b3fb56 | Author: watusimoto | Message: Fire ShipLeftZoneEvent when ship dies in a zone
19:28:21BFLogBot Commit: 43eb708955da | Author: watusimoto | Message: Var name
19:28:23BFLogBot Commit: 145f66d01d52 | Author: watusimoto | Message: Var name
19:28:25BFLogBot Commit: cdec3b016a76 | Author: watusimoto | Message: Comments, fix compile problem
19:28:27BFLogBot Commit: 4d3c6592079f | Author: watusimoto | Message: Comment
19:28:28BFLogBot Commit: 20be8500bd62 | Author: watusimoto | Message: Comment
19:28:30BFLogBot Commit: 67df0e26e508 | Author: watusimoto | Message: Comment
19:28:31BFLogBot Commit: b1eef60e940a | Author: watusimoto | Message: Comment
19:28:33BFLogBot Commit: 3be21c8ad124 | Author: watusimoto | Message: Bursts can no longer damage things through forcefields
19:28:34BFLogBot Commit: 8f949f47de1c | Author: watusimoto | Message: Fix Zap! era bug with bursts only making items move horizontally. Now they can be blasted every which-way!
19:28:36BFLogBot Commit: 398ee95b29b5 | Author: watusimoto | Message: Remove unneeded if
19:28:38BFLogBot Commit: 89cfb2af9019 | Author: watusimoto | Message: Replace counter with a timer, rename some vars. No (intentional) change in functionality.
19:28:39BFLogBot Commit: 54a212b594df | Author: watusimoto | Message: Set the timer period just once during initiailization
19:28:41BFLogBot Commit: 129953afa291 | Author: watusimoto | Message: Kill warning
19:28:42BFLogBot Commit: 592a64640f2b | Author: watusimoto | Message: Kill warnings
19:28:45BFLogBot Commit: 449fadcd8e8b | Author: watusimoto | Message: Merge
19:28:52Watusimotook, those are my changes
19:30:22Watusimotothe only one that is really relevant for tonight is ... hell none of these are relevant
19:30:30WatusimotoI fixed the main thing I wanted to fix this morning
19:30:50raptori've almost fixed the linker error... you're never going to believe this..
19:31:01Watusimotoyou're probably right
19:43:40Watusimotook, stepping out for a bit; back later
19:47:54Watusimoto Quit (Ping timeout: 248 seconds)
19:54:33raptorok, i pushed the bonkers fix
19:54:36BFLogBot Commit: e91190a25a31 | Author: buckyballreaction | Message: Fix really weird linker errors when compiling optimized build on older GCC. This should fix compiling on CentOS and Debian systems as well as any system using GCC 4.5 or older
19:54:43raptornow to business!
20:02:38raptoruhhhh
20:02:47raptorwatusimoto made a serious change in gameplay
20:03:30kaenoh?
20:03:37raptorhttps://code.google.com/p/bitfighter/source/detail?r=3be21c8ad124007d7f5c6cf44bb0a7574bef582c
20:03:46raptori will have to revert for the contest, as some maps require that...
20:04:36kaenoh boy so we're using latest mercurial server against 018 release clients?
20:04:47raptoryeah... this is going to be interesting
20:04:52kaen:)
20:06:05kaenon a positive note, I've got dedicated vs. full build targets working. just need to add debug mode and tell it about the included libraries
20:06:19Watusimoto has joined
20:06:53Watusimotohi\
20:06:57raptoruh
20:07:00raptorhi!
20:07:03raptori fixed the bug
20:07:05Watusimotooh brother.
20:07:05raptorlinker errors
20:07:09raptorsee my last commit...
20:07:25Watusimotomy kids have decided to play in the contest tonight
20:07:41Watusimotoand, by some crazy twist of fate, their mother agreed
20:07:46raptorha!
20:08:01Watusimotoso I think they are waking up at 4AM to catch the last two hours
20:08:02raptorwait, that's.... late...
20:08:31Watusimotoso... starts in 5 hours
20:08:37Watusimotothat's 2AM here
20:08:43Watusimotoruns for 4 hours
20:08:47Watusimotothat's until 6AM here
20:08:54Watusimotoso, yeah, that will work
20:08:59Watusimotoexcept it will suck
20:09:05Watusimotofor them
20:09:10Watusimotoand probably for all of us tomorrow
20:09:17raptorcrazyies
20:09:56Watusimotouh... I don';t understand your commit
20:10:21raptoryeah, me neither...
20:10:42Watusimotoha
20:10:46Watusimotohow did you find it?
20:11:57raptori found by accident that it was the lack of TNL_DEBUG flag that was causing the linker error
20:12:13raptori then went in to see what it triggered, and found it triggered TNL_ENABLE_ASSERTS
20:12:30Watusimotowell, that gets my vote for hack of the day!
20:12:32raptorso i removed my tricky assert code (which I added in 017?)
20:12:36raptorand it compiled!
20:13:01Watusimotowhat was the intent of the assert code you added in the first place?
20:13:22raptorto completely compile out the assert
20:13:28raptorbefore a block { } was being added
20:13:47Watusimotoand you wrapped it in the do ... while?
20:14:14Watusimotowhat do you mean by "completely compile out"?
20:14:20raptoryes, and that is trickery i pulled from SDL to make sure asserts get completely compiled out, method and all
20:14:26raptorin optmized builds..
20:14:40Watusimotocompile out meaning no trace at all?
20:14:46raptorexactly
20:14:59raptorthe old way still leaves a method call with an empty block when compiled
20:15:33raptorso it's a performance gain, especially with all the asserts in tnlVector
20:15:34Watusimotobut do {x} while 0 will still do x, no?
20:15:52raptornope
20:16:04raptorthere is a read of it somewhere, that i'm trying to find again...
20:17:02Watusimotommmm... I think it should
20:17:13Watusimotowhile(0) {x} won't run
20:18:24Watusimotodo { stuff() } while(0);
20:18:24Watusimotois doing the exact same thing as stuff()
20:18:35Watusimotoaccording to the stack overflow article I'm looking at
20:18:49Watusimoto1st line of 1st answer: http://stackoverflow.com/questions/2687569/c-do-while0
20:20:05Watusimotoand explains why you might use the do{} in a macro
20:20:29raptorthe trickery is in the (void) sizeof ((x))
20:20:33raptorargh
20:20:39raptordedicated compile fails
20:21:00Watusimotois this comment correct?
20:21:01WatusimotoTricky way of making sure asserts are completely compiled
20:21:07Watusimotoor is it missing a word at the end?
20:21:10raptorcompiled out...
20:21:11raptoroops
20:21:32WatusimotoI'll fix it, as I want to look at that code in context anyway
20:22:01raptorit has to do with your changes
20:22:16Watusimotothe most recent ones?
20:22:18raptori have to leave in 20 min for about 2 hours
20:22:20raptoryes
20:22:26raptormy change compiles fine
20:22:29Watusimotowant me to work on it?
20:22:34raptori think i got it..
20:23:06raptorhmm.. no i don't...
20:23:44raptorship.cpp:1269:34: error: invalid use of incomplete type ‘class Zap::ClientGame’
20:23:48raptorwith dedicated...
20:24:17Watusimotommm
20:24:25Watusimotoifdef with return false
20:24:42Watusimotooops, won't work
20:24:52Watusimotoas Clientgame is in the args
20:25:02Watusimototry ifdefing the whole fn out
20:25:05Watusimotoit may compile
20:25:12Watusimotoif nothing tries to call it, it may link
20:25:13raptorok
20:25:25Watusimotoit's only called by client game
20:25:30koda has joined
20:25:38Watusimotowe may be able to get by without ifdeffing it in the .h file
20:26:39raptorok, ifdef'd...
20:26:43raptorno compiles
20:27:04raptoralso
20:27:12raptorwe need to revert: https://code.google.com/p/bitfighter/source/detail?r=3be21c8ad124007d7f5c6cf44bb0a7574bef582c
20:27:17BFLogBot Commit: 067b0288d2e7 | Author: buckyballreaction | Message: Fix dedicated compile
20:27:21raptorsome contest levels depend on that
20:27:24raptorwill that be a problem?
20:28:57Watusimotolooking
20:29:06Watusimotoreally?
20:29:14Watusimotono, revert away
20:29:37Watusimotojust change line 916
20:29:59Watusimotoleave the fn isWallOrForcefieldType in place for now
20:30:13Watusimotowe can discuss this issue later
20:35:21raptorok
20:35:25raptorreverting and leaving...
20:35:33Watusimotook, anythink you need me to do?
20:35:41Watusimotoother than make a level?
20:35:54raptornope
20:35:56Watusimotook
20:35:58Watusimotolater!
20:37:50BFLogBot Commit: 865d6f3bcb9e | Author: buckyballreaction | Message: Revert part of 3be21c8ad124, bursts need to damage beyond forcefields for now
20:38:02raptorbye
20:38:04raptor Quit ()
22:20:19LordDVG has joined
22:56:18Fordcars has joined
22:56:48Fordcarsis the party still on tonight?
22:57:22kaenyou bet!
22:57:31kaen\o\
22:57:55Fordcarsnice I had been occupied lately for christmas
22:59:11Fordcarsgreat to see that it's tonight because I just came back home
22:59:42kaennice timing :)
23:00:35Fordcarsyou bet! I was supposed to come back on the 27 but a big snow storm blocked the roads
23:01:55kaenoh wow
23:02:02kaenI do not miss snow at all ^^
23:02:10Fordcarslol
23:15:58LordDVG Quit (Remote host closed the connection)
23:17:34BFLogBot Commit: 949d75820bae | Author: watusimoto | Message: Centralize setting the timer period
23:17:35BFLogBot Commit: 230776e873cd | Author: watusimoto | Message: Stupid variable name
23:17:37BFLogBot Commit: 3d104a90c343 | Author: watusimoto | Message: Replace dynamic cast with static
23:17:40BFLogBot Commit: ebc1f9aa1f47 | Author: watusimoto | Message: Unkill warning -- need to read the data to keep thing srom crashing!
23:17:41BFLogBot Commit: 8466bef4002b | Author: watusimoto | Message: Merge
23:17:43BFLogBot Commit: 9e4c22c0ff41 | Author: watusimoto | Message: Merge
23:17:44BFLogBot Commit: 96d0db0bf864 | Author: watusimoto | Message: Fix zone control capture without flag bug
23:17:46BFLogBot Commit: df9f76649998 | Author: watusimoto | Message: Only send position when it changes
23:17:47BFLogBot Commit: 627ef1d7023e | Author: watusimoto | Message: Add TODO for 019
23:17:50BFLogBot Commit: 763a81cd07cf | Author: watusimoto | Message: Whitespace
23:17:52BFLogBot Commit: 2d542ca1b9fb | Author: watusimoto | Message: Make sure flag goes home after touchdown in ZC
23:46:32raptor has joined
23:46:33ChanServ sets mode +o raptor
23:46:43raptorello@!
23:47:44raptorWatusimoto: new changes for the game server?
23:48:03Watusimotohi
23:48:05Watusimotoyes
23:48:12Watusimotothree major bug fixes
23:48:15raptorok
23:48:37WatusimotoI just spun through all the stock levels using production client and latest server, and only found one bug
23:48:45raptorsweet!
23:48:53Fordcarsby the way I have a tiny bug when you play retreive and you retreive all flags, the last one you bring to goal goes instantly to it's spanw point
23:48:53Watusimotoit's the flag left behind bug
23:49:04WatusimotoI think my zc fix (the last one above) broke it again
23:49:17raptorFordcars: that has been a bug since i've joined bitfighter...
23:49:53Fordcarshaha lol well then it's not important at all
23:50:12Fordcarsis it normal you can't repair teleporters?
23:50:18raptoryes
23:50:26Fordcarsok
23:50:45Fordcarsyou could abuse the teleporter too much huh
23:50:48raptorbecause we weren't sure how game-altering they were going to be..
23:50:57Fordcarsyeah ok
23:51:50raptoroh yay dedicated still works
23:52:15raptor2 hours until party!
23:52:29raptori think...
23:52:34raptorunless i'm mixed up
23:52:37Fordcarsyeah
23:52:52Fordcarsits nearly 7
23:54:25raptortheShip!
23:55:08Watusimotois that a comment on my checkin?
23:55:13raptoryes
23:55:15raptorhaha
23:55:28WatusimototheStupidVariableName
23:55:39raptorthisIsTheVariableToUse
23:55:47kaenahoy!
23:55:50raptorhi kaen
23:55:59kaenhttps://code.google.com/r/bkconrad-kaen-bitfighter/source/detail?r=a9bd18d61d43d91ef48e9e83d62302d606fe08e5
23:56:05kaen^ cmake
23:56:14raptorwhoa
23:56:49kaenonce applied just do `cmake . && make` and watch her go
23:56:56raptorcan't wait!
23:57:05kaen(it'll clobber the Makefile though)
23:57:10raptoroh...
23:57:12raptorhmmm
23:57:14raptorok
23:57:29raptori won't merge it quite yet.. if that's OK - maybe after the contest
23:57:38kaenyou bet
23:58:11kaenwould you mind testing it though?
23:58:14raptorsure!
23:58:22raptorone moment... finishing my other commit reviews..
23:58:32kaenno problem
23:58:51Watusimotohey, so what's the advantage of using cmake?
23:58:56Watusimotoin our case
23:59:09Watusimoto(I already accept that make is horrid)
23:59:10raptorless maintenance of make files
23:59:19kaenuh, nope
23:59:27kaenwe'd have to restructure for that
23:59:32Watusimotothere's not dealing with make!
23:59:37kaenwell, maybe a little bit
23:59:38Watusimotothat's a plus!
23:59:41kaenlol
23:59:42raptorbut also better dependency seearching..
23:59:55kaenselective compilation is my big thing

Index Search ←Prev date Next date→

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