#bitfighter IRC Log

Index Search ←Prev date Next date→

IRC Log for 2011-08-09

Timestamps are in GMT/BST.

00:25:00raptor has joined
00:25:00ChanServ sets mode +o raptor
00:25:00raptorwatusimoto: still here?
00:25:00raptore-mail you say"?
00:26:00sam686I see a problem in http://bitfighter.org/gamereports/ it is writing ContestServ stats as server name "TESTING"
00:27:00raptoryes - saw that too - i wonder why that is?
00:27:00sam686also, http://bitfighter.org/bitfighterStatus.json shows a server named "TESTING"
00:28:00sam686Thats probably due to using /setservername and new name is not sent to master..
00:30:00sam686if used /setservername, only the new name will be sent to master after server reconnects to master..
00:42:00raptormakes sense
00:59:00Flynnn Quit (Quit: This computer has gone to sleep)
01:15:00raptorok, i'm going to send watusimoto an e-mail about the Client game stuff
01:31:00raptorrats that doesn't work, i forgot
01:33:00raptorok sent watusimoto an e-mail about the ClientGame stuff, i couldn't find yours so you are not CC'd; sorry
01:36:00sam686couldn't fune my what? email?
01:36:00sam686could not find my email or what?
01:36:00raptoryeah e-mail
02:15:00Flynnn has joined
02:20:00watusimoto Quit (Ping timeout: 258 seconds)
02:24:00raptorwow, he responed in e-mail faster than IRC!
02:37:00sam686 has left
02:38:00sam686 has joined
02:38:00ChanServ sets mode +v sam686
02:52:00Flynnn Quit (Quit: This computer has gone to sleep)
03:37:00Flynnn has joined
04:18:00sam686 Quit (Read error: Connection reset by peer)
04:21:00sam686 has joined
04:21:00ChanServ sets mode +v sam686
04:24:00watusimoto has joined
04:27:00raptorso I understand the idea to keep UI stuff separate from the Game
04:28:00raptorso what to do about those enums that are needed in the headers of both game.h and UIGame.h but can't be forward declared?
04:28:00raptorjust make them U32?
04:54:00raptorsam686: are you around?
04:55:00watusimoto@r did I answer your question?
04:55:00raptorhi
04:55:00raptoryes-ish
04:55:00watusimotoish?
04:55:00raptorI think I agree with your logic
04:55:00raptorbut I don't know how to solve the compiling problem
04:55:00watusimotoyou should -- it's in response to your case!
04:56:00raptorboth game.h and UIGame.h need those enums: VolumeType, UIMode
04:56:00watusimotoI don;t really understand what the issue is
04:56:00watusimotoyes
04:56:00raptorbut we're trying to avoid including the headers in each other
04:56:00raptorand enum cannot be forward declared
04:56:00watusimotoah ok
04:56:00raptor(yes, thanks for considering that issue!)
04:57:00raptorso, answer is put them in a different header?
04:57:00watusimotoso if we define VT in game.h, we can include that in gameui.h right/
04:57:00raptorbut which? (or a new one?)
04:57:00raptoryes, we can do that
04:58:00watusimotoso then VT will be in both game and gameui... so where's the problem?
04:58:00raptorI think the problem was because sam said including game.h there really slowed down compiling
04:58:00raptorbut that is most obvious solution, to me
04:59:00watusimotoit probably triggers a whole chain of compilations
04:59:00watusimotono forward declarations, eh?
04:59:00raptoryeah - it would require a full recompile for you guys. and I know how you loathe to do it...
04:59:00watusimotoyes
04:59:00raptor(hence why have I so many 'fix compiling errors' commits... :)
05:00:00raptorI am so close to being done with ridding ourselves of the object mask...
05:01:00raptori am a little stuck at the moment in this method: ship.cpp:1744
05:01:00raptorLuaShip::getMountedItems
05:02:00raptori'm trying to figure out the general logic of the method
05:02:00raptorso i can adapt
05:04:00watusimotowe could, as you say, stick the enum somewhere else
05:05:00raptoryes
05:05:00watusimotogetMountedItems...
05:05:00watusimotolooks crazy
05:05:00raptorbut I think you will both need a full recompile after this mask removal is complete anyways...
05:05:00watusimotowho wrote that>
05:05:00watusimoto?
05:05:00raptoryeah, i've been a bit confused...
05:06:00watusimotoah, so what's your question?
05:06:00raptorwhat is that method doing?
05:06:00raptoris my question ^^
05:06:00raptor:)
05:06:00watusimotoreturning a list of items mounted to the ship
05:06:00watusimotoflags, balls, resource items
05:06:00raptorand that is for Lua to use, correct?
05:07:00watusimotoyes
05:07:00raptorit is populating the lua state with the request?
05:07:00watusimotoif you have a bot and it wants to know what its carrying... use this method!
05:07:00watusimotoit's creating a lua table, which is a little like an array
05:08:00watusimotothen populating that with carried items
05:08:00raptorok, i still don't understand the purpose of the objectMask in it, though..
05:08:00watusimotoand returning the list
05:08:00raptorespecially if it already knows exactly whats on the ship with: thisShip->mMountedItems
05:08:00watusimotowell... objectType is a bitmask of all the different object types we might want to find
05:09:00watusimotoso
05:09:00raptorlooks redundant
05:09:00watusimotoif you want to find all the flags you are carrying, you pass in flagType
05:09:00watusimotoI think
05:09:00watusimotoI think you call this function by doing a getMountedItems(type1, type2, type3,...)
05:10:00raptorahhh
05:10:00watusimotoand you get a list of all moutned items of those types
05:10:00raptorok
05:10:00watusimotoyou should check the wiki
05:10:00watusimotoit might be documented there
05:10:00watusimotoif you pass no args, you get back a list of everything you are carrying
05:11:00watusimotoLua has no bitwise operators
05:11:00raptorhaha, only reference in the wiki is in 1st para here: http://bitfighter.org/wiki/index.php?title=Programming_robots
05:11:00raptoryes, we don't need bitwise anymore, either
05:11:00watusimotowhich is why we have to pass all the types to be or'ed together by C++
05:11:00raptorso with type numbers as U8, we'll have to do a findObjects on each one now
05:12:00watusimotowe need better docs
05:12:00watusimotoyes
05:12:00watusimotobut in reality, you'll probably specify 0 or 1 type
05:12:00raptorisn't there something like javadoc generation for c++? then we could actually comment the code and generate on the fly...
05:12:00watusimotoI can't see much applicatino for "give me a list of all the soccer balls and flags I'm carrying, but no resource items, please!"
05:13:00raptorhaha
05:13:00watusimotoi hate javadocs
05:13:00watusimotohate hate hate hate
05:13:00watusimotohate
05:13:00raptorhahaha
05:13:00watusimotojavadpocs
05:13:00watusimotohate them
05:13:00watusimotobut yes, there are comparable tools
05:13:00raptorhate writing them? reading them? or the fact they take up 3/4 of the code?
05:14:00watusimotohate reading them, hate that the stucture is terrible
05:14:00watusimotohate the stupid font
05:14:00watusimotobut yes, is we had a system like that for the lua code, that would be great
05:14:00raptori'll be honest - i don't know of any better alternative - and they have been helpful to me in the past (as well as misdirecting me...)
05:15:00raptori do not like how they are presented though..
05:15:00watusimotoI'm a tech crumedgeon
05:15:00raptorseems like they try an gouge your eyes out
05:16:00watusimotoif you can find a system we could use, I'd keep them update
05:16:00watusimotod
05:16:00watusimotodyoxygen
05:16:00watusimotoI looked at that
05:16:00watusimotothat's similar
05:16:00watusimotobut we don;t need to document our whole api, just the lua bits
05:16:00raptoryes, i agree
05:17:00raptorok, let me rework this method...
05:19:00watusimotohttp://www.doxygen.org/http://www.google.com/url?sa=t&source=web&cd=1&ved=0CBwQFjAA&url=http%3A%2F%2Fwww.doxygen.org%2F&ei=KMNATpKXHNPQiALJzNybBQ&usg=AFQjCNGA9Lo4ZIKrHHGjIDEkdj4xC96cGQ&sig2=8sBB12j36efzdsZ5z9VM
05:19:00watusimotohttp://www.doxygen.org/
05:19:00watusimotoignore the first link
05:19:00watusimotofeel free to add some notes about what it does
05:20:00raptorthis is how the naev folks use it: http://wiki.naev.org/wiki/Coding_Style#Doxygen
05:21:00raptorthey even use three spaces instead of tabs!
05:21:00raptorha!
05:29:00raptorok, almost have a method rewrite i'd like you to take a look at to keep me on the right track...
05:30:00watusimotook
05:31:00raptorhere is the new beast: http://pastie.org/2343341
05:34:00watusimotofalling asleep...
05:34:00watusimotolooks ok; what if the user passes no values?
05:35:00raptornot sure, did the old handle it?
05:35:00watusimotodoes the outer look get run?
05:35:00watusimotowell, to get all objects, you'd pass no args, not 0
05:35:00raptorohhh
05:36:00raptori'll have to pull that out of the loop, then...
05:36:00raptorone moment..
05:36:00watusimotoyes, the old one handled it
05:37:00watusimotoif no args were passed, typemask statyed at intiialized value, which was 0
05:37:00Flynnn Quit (Quit: This computer has gone to sleep)
05:39:00raptorok, reload the pastie
05:41:00raptornot super clean, but similar to the old, i guess
05:43:00watusimotook
05:43:00watusimotowhat if I pass the args (1,2,0)
05:43:00watusimotowill I get dupes?
05:44:00watusimoto:-)
05:44:00watusimotoand how many dupes will I get?
05:45:00raptormaking me think late i night...
05:45:00raptorif (objectType == 0 && tempVector.size() == 0)
05:45:00raptoryes?
05:45:00watusimotowhy not
05:45:00watusimotowhy not
05:46:00watusimotowhy not iterate over the mounted items in the outer loop
05:46:00watusimotoand check them against teh requested types in the inner loop
05:46:00watusimotothat way each mounted item can, at most, get added to the list once
05:46:00watusimotono dupes
05:46:00raptoryes, that is good idea
05:47:00watusimotobecause your fix will only reduce the number of type1s from 3 to 2
05:47:00watusimotoyou;ll still have dupes
05:48:00Flynnn has joined
05:48:00raptori guess i do not understand the output of this method...
05:48:00watusimotowe're retirning a list of items that meet the requested criteria, no?
05:48:00raptoryes, but what if i am carrying 10 nexus flags?
05:48:00raptorwill it return 1 or 10?
05:49:00watusimoto20
05:49:00watusimoto10
05:49:00watusimotosorry
05:49:00raptorhahaha
05:49:00watusimotoif you request a list of all nexus flags you;re carryihg, you shoudl get 10
05:49:00raptorok
05:50:00watusimotoand if you speciyf (nexusflagtypenumber, nexusflagtypenumber,nexusflagtypenumber, 0)
05:50:00watusimotoyou should still get 10
05:51:00raptorah, and that's where my loop breaks down...
05:51:00raptorok
05:51:00watusimotoyes
05:51:00raptori get it
05:51:00watusimotogood
05:51:00watusimotoonce a flag meets one criteria, it's found; move on to next
05:55:00watusimotohey, here's an idea
05:55:00watusimotowhat if we made our composite types have negative type numbers
05:55:00Flynnn Quit (Quit: This computer has gone to sleep)
05:56:00raptorum... U8?
05:56:00watusimotoso engineeredtypenumber might be -1
05:56:00watusimotoU8, did you say?
05:56:00raptorthey's still have to map to the normal types somehow
05:56:00watusimotobecause it would be nice to be able to use composites with the bots
05:57:00watusimotoI suppose we could just reserve a few numbers for that puroise
05:58:00raptorso how would you map, logically or mathematically, both engineered types to the new number without using a mask or crazy boolean functions
05:58:00raptor?
06:03:00raptorstill awake? I have a revised method: http://pastie.org/2343341
06:03:00raptori'm fading fast, too...
06:05:00raptoroh my goodness it compiled
06:12:00raptorsome goofy errors - let me see if i introduced them..
06:20:00raptoryep, i sure did
06:20:00raptorgreat
06:20:00raptorwell, i have to get to bed
06:20:00raptorgood night
06:23:00watusimotojnighy
06:23:00watusimotofaded\
06:23:00raptoryeah, you're tired too
06:27:00watusimoto Quit (Ping timeout: 255 seconds)
06:34:00raptor Quit (Remote host closed the connection)
06:48:00LordDVG has joined
06:48:00LordDVG Quit (Changing host)
06:48:00LordDVG has joined
06:58:00LordDVG Quit (Read error: No route to host)
06:59:00LordDVG has joined
06:59:00LordDVG Quit (Changing host)
06:59:00LordDVG has joined
08:08:00LordDVG Quit (Remote host closed the connection)
08:08:00LordDVG has joined
08:08:00LordDVG Quit (Changing host)
08:08:00LordDVG has joined
08:34:00LordDVG Quit (Ping timeout: 250 seconds)
09:06:00LordDVG has joined
11:03:00LordDVG Quit (Ping timeout: 252 seconds)
11:13:00LordDVG has joined
12:30:00LordDVG Quit (Remote host closed the connection)
13:03:00LordDVG has joined
14:34:00LordDVG Quit (Read error: Connection reset by peer)
14:35:00LordDVG has joined
15:18:00raptor has joined
15:18:00ChanServ sets mode +o raptor
15:25:00raptorgood morning!
15:48:00watusimoto has joined
15:55:00Zooooooomber has joined
15:56:00ZooooooomberWhy hello there Pointblank
15:56:00raptorhi Zooooooomber
15:56:00Zooooooomberok raptor, here i a pastebin of my script
15:56:00Zooooooomberhttp://pastebin.com/1Vbr3Rxb
15:59:00raptorok looking...
15:59:00raptorok
15:59:00raptorhuh, i don't see a problem...
15:59:00guest_ has joined
16:01:00Zooooooomberhi guest_
16:02:00raptorZooooooomber:
16:02:00raptorhehe
16:02:00raptorline 43
16:02:00raptorrandom quotation mark at the end :)
16:02:00raptorremove it
16:03:00Zooooooomberaaaeeee ok
16:04:00raptorguest_ is from DC area...
16:04:00raptormust be Little Apple? or BlackBird?
16:04:00Zooooooomberraptor, testing
16:04:00Zooooooomberok
16:04:00Zooooooombernow no errors, but it just says, TNL Service is DOWN.[Screen is terminating]
16:05:00raptoryup that looks like what it does
16:05:00raptorso no syntax errors
16:06:00Zooooooomberok
16:06:00Zooooooomberso i got the script right
16:06:00Zooooooomberbut screen terminates
16:07:00Zooooooombermaybe im having my "starrt up bitfighter
16:07:00Zooooooombercommand in the wrong spot?
16:07:00Zooooooombermy apoligies for pressing enter furiously between sentences
16:07:00raptoryeah could be
16:07:00Zooooooomberok, let me try moving it
16:07:00raptorwait
16:07:00raptorlooks ok
16:08:00Zooooooomberright now its if [ $? == 1 ]; then # only mail if i haven't already screen /root/sam/exe/bitfighterd -IP:ANY:28000 if [ ! -f "$sendoncefile" ]; then echo "$message" | mail -s "$subject" "
16:08:00raptordoes screen automatically terminate when you launch a command?
16:08:00Zooooooomberit only terminates for this command
16:08:00Zooooooomberprobably i need to change directories to do this?
16:08:00Zooooooombermaybe this:
16:08:00Zooooooombercommand cd /root/sam/exe screen ./bitfighterd ......
16:09:00Zooooooombermaybe when i try to run bitfighterd with screen, its searchiing for levels in the current directory i am in?
16:09:00raptorif you ran this exact command on the command line, what happens?: screen /root/sam/exe/bitfighterd -IP:ANY:28000
16:09:00Zooooooomberahh, good idea
16:10:00Zooooooomberah, it terminates
16:10:00Zooooooomberso the script is fine, that command just doesnt work
16:10:00raptormaybe you need quotes around the command: "/root/sam/exe/bitfighterd -IP:ANY:28000"
16:10:00Zooooooomberok
16:11:00Zooooooomberah, probably sohuld close quotes after bitfighterd though
16:11:00Zooooooomberok, when i do this
16:11:00Zooooooomber "/root/sam/exe/bitfighterd" -IP:ANY:28000
16:12:00Zooooooomberit must be searching for levels in the current directory im in
16:12:00raptortry: screen -dmS sessionName /root/sam/exe/bitfighterd -IP:ANY:28000
16:12:00Zooooooomberecuase it says "could not read any levels from file folder "" "
16:12:00raptorno, it is a problem with screen
16:12:00raptoroh yes
16:12:00raptorand that
16:12:00raptoryou might have to add -rootdatadir
16:12:00Zooooooomberill try that
16:12:00raptorhave to focus on work for a bit...
16:13:00Zooooooomberok
16:13:00Zooooooombermight go soon, if i dont catch you when i leave, see you alter
16:13:00Zooooooomberok rootadatadir works thanks
16:15:00Zooooooomberok, now the script is good
16:15:00raptoryou might want to add the kill command right about the screen command just in case
16:15:00raptorright above
16:16:00raptorso if the ping fails, and the server is still running
16:16:00Zooooooomberyou mean, kill a pervious instane of screen?
16:16:00raptorit is killed first
16:16:00Zooooooomberoh
16:16:00Zooooooomberi just tried ./check.sh and it started up the server fine
16:16:00raptorok
16:16:00raptorjust some thoughts..
16:16:00Zooooooomberim trying to see if cron works now
16:16:00Zooooooomberill make sure i note that though
16:17:00Zooooooombernow i need to know what the deal with cron is
16:17:00Zooooooomberim goign to try crontab -l to ese
16:19:00Zooooooomberis sam686 awake?
16:19:00sam686hi
16:20:00Zooooooomberhi
16:22:00Zooooooomberok cron is doing something
16:24:00Zooooooomberit seems crontab isnt working right, but thi is what i have in crontab: */2 * * * * "/root/sam/tnlping/check.sh"
16:25:00raptorrunnign the script every two minutes
16:25:00raptorremove the quotes
16:25:00Zooooooomberok, i tried that exact command from my Root folder directory and it works, but with cron, it is not working
16:25:00Zooooooomberi can remove the quotes, just tried adding them becuase it wasnt working without them either
16:26:00raptorok
16:26:00raptorit shouldn't matter
16:26:00Zooooooomberok ive gottena nother mail message
16:26:00raptortake a look
16:26:00raptorso it is runnign ok
16:26:00raptorbut maybe the environment isn't set up properly
16:26:00Zooooooomberit mails me this: http://pastebin.com/hz7YpEK4
16:27:00Zooooooomberbut the server doesnt go online
16:27:00Zooooooomberok
16:27:00Zooooooomberenviornment
16:27:00raptorhaha
16:27:00raptorbecause of screen
16:27:00raptoryou can't run screen from cron, I guess...
16:27:00Zooooooombernoooooooooooooooooooooooo
16:27:00raptorit needs a full fledged shell
16:27:00Zooooooomberbut
16:28:00Zooooooombermaybe i dont even need to use Screen command?
16:28:00raptorcorrect
16:28:00Zooooooomberpossible for it to stay up without it?
16:28:00raptorjust add an '&' at the end of the line instead - to run in background
16:28:00raptorso ./bitfighterd -yadaydada &
16:28:00Zooooooomberso instead of screen "/root/sam/exe/bitfighterd" -IP:ANY:28000 -rootdatadir "/root/sa$
16:28:00Zooooooomber "/root/sam/exe/bitfighterd" -IP:ANY:28000 -rootdatadir "/root/sa$&
16:29:00Zooooooomberoh sorry, the money sign just means it goes on
16:29:00raptorok
16:29:00raptoradd a sapce before teh &
16:29:00raptorspace
16:29:00Zooooooomberok
16:29:00raptoractually that might not matter, but makes it more readable
16:30:00Zooooooomberdone
16:30:00Zooooooomberlets test!
16:31:00Zooooooomberaah!
16:31:00Zooooooomberit worked!
16:31:00Zooooooombernow i will try to crash it
16:31:00sam686or use a /shutdown in a server
16:31:00Zooooooomberactually, thats good, now i must just make sure it stays up
16:32:00Zooooooomberyes, /shutdown would be a better diea
16:32:00Zooooooomberidea
16:32:00Zooooooomberoh no 666 messages
16:33:00raptorso the way to get rid of mail messages is not let a cron script output anything
16:33:00raptori.e. get rid of any 'echo....' lines
16:33:00Zooooooomberok
16:34:00Zooooooomberthat i can do!
16:34:00Zooooooomberor i could add MAILTO= in my crontab?
16:36:00Zooooooombertrying without echo lines
16:39:00Zooooooomberdidnt work.
16:39:00Zooooooombergives me syntax errors without echo commands
16:48:00Zooooooomberok, i think i got it fixed
16:50:00Zooooooomberoh my this isnt good
16:54:00Zooooooomberhelp! my server is corrupted
16:54:00Zooooooomberevery two minutes, a new Bitfighter Main(); pops up
16:57:00Zooooooomberi cant stop it for some reason
16:58:00Zooooooomber@anyone, will having two ./check.sh scripts in two completley different folders affect each-other?
16:58:00Zooooooomberi made the necessary changes to the new ./check.sh script
17:00:00watusimoto Quit (Ping timeout: 263 seconds)
17:01:00raptorZooooooomber: tell me your woes
17:02:00Zooooooomberraptor : i created a second shell script and told it to tnlping port 28020 instead of 28000
17:02:00Zooooooomberi also tried all the same commands but changed the /hg/ directory from /sam/ to /sam2/
17:02:00Zooooooomberhowever, it seems whenever i start up this shell script, it doesnt bind bitfighter Main(); to port 28020
17:02:00Zooooooomberit binds it to some random port
17:03:00raptorsounds like the -IP:ANY:????? isn't working
17:03:00raptornot sure why
17:03:00Zooooooomberunfortonatley, TNLping is still searching for 28020, so it keeps running the script over and over
17:03:00Zooooooomberyes
17:03:00Zooooooomberithats what it is
17:03:00Zooooooomberbut i dont knwo why
17:03:00Zooooooombermaybe -rootdatadir is contradicting this?
17:03:00raptorcan you start the server manuall with that and it is still random?
17:04:00Zooooooomberi start the server with ./check.sh or cron and it says it conects to bad port
17:04:00Zooooooomberi will try it manually
17:05:00Zooooooomberok im trying it manually
17:05:00Zooooooomberroot@bitfigher:~/sam2/tnlping# /root/sam2/exe/bitfighterd -IP:ANY:28001 -rootdatadir "/root/sam2/exe" Welcome to Bitfighter! No OpenAL support on this platform. Server connecting to master [IP:67.18.11.66:25955]
17:06:00Zooooooombermaybe ill try -ip after -rootdatadir
17:06:00raptorhow do you check the port?
17:07:00Zooooooomberthats not doing it
17:07:00Flynnn has joined
17:07:00Zooooooomberi check what port it is by going on bitfighter and looking at the port
17:07:00LordDVG Quit (Ping timeout: 260 seconds)
17:07:00Zooooooomberits always some random port
17:07:00Zooooooomberhi flynnn
17:09:00raptorsam686 might know more about this problem than I
17:09:00Zooooooombersam686, it seems if i open up a second server from the same IP: it binds it to any old port than the IP:ANY:PORT i specify it to bind to
17:10:00sam686it starts dedicated as a random port number if a port is taken and can't be used
17:11:00LordDVG has joined
17:11:00sam686by default will use port 28000, but adding -dedicated "IP:any:28003" will set the port number and try to user it..
17:11:00Zooooooomberok
17:12:00Zooooooomberi tried -IP:any:28010 and it bound the socket to some random port
17:13:00sam686if it did bing to 28000, then you didn't set the command right
17:13:00Flynnn Quit (Ping timeout: 252 seconds)
17:13:00sam686should be ./bitfighter -dedicated "IP:any:28010" (yours might be ./bitfighterd)
17:13:00Zooooooomberright now i have "/root/sam2/exe/bitfighterd" -IP:ANY:28020 -rootdatadir "/root/sam2/exe/" &
17:14:00sam686you need -dedicated part
17:14:00Zooooooomberbitfighterd dedicated already?
17:14:00Zooooooomberor does the argument only work with -dedicated?
17:15:00sam686quote problem?
17:15:00Zooooooomberoooooh
17:15:00Zooooooomberi see
17:15:00Zooooooomberi thought -ip was its own argument
17:15:00Zooooooomberthanks!
17:15:00sam686it needs -dedicated, then a space, then quote the IP:any:28010
17:16:00Zooooooombertesting
17:16:00Zooooooomberthere we go
17:17:00sam686try "/root/sam2/exe/bitfighterd" -dedicated "IP:ANY:28020" -rootdatadir "/root/sam2/exe/"
17:18:00LordDVG Quit (Ping timeout: 252 seconds)
17:18:00Zooooooombervery interesting
17:19:00Zooooooomberraptor or sam, i have tnl ping set to ping server="IP:74.122.199.250:28020"
17:19:00Zooooooomberhowever, i am getting mail that that is up when it is really down..
17:19:00Zooooooomberwait
17:19:00Zooooooomberoh
17:19:00Zooooooombertnevermind
17:19:00raptor:)
17:19:00Zooooooomberthats for my first ./check.sh script
17:20:00Zooooooomberok, so i wonder why my second ./check.sh script isn't working
17:20:00Zooooooomberit works when I go manual
17:20:00Zooooooomberbut not in cron for some reason
17:20:00Zooooooomberwait
17:20:00Zooooooomberoh, my mistake
17:21:00LordDVG has joined
17:21:00Zooooooomberworks now
17:22:00Zooooooomberill be back
17:22:00Zooooooomberlater
17:22:00Zooooooomber Quit (Quit: Page closed)
17:25:00raptorsam686: are you working on something specific?
17:26:00sam686nothing much..
17:26:00raptorbecause I got my mask removal code to compile
17:26:00raptorand I see two bugs I need to track down...
17:27:00raptorbut I am super busy for the next couple days and can't spend much time on it
17:27:00raptorwant to take a look?
17:27:00sam686yes..
17:27:00raptorok creating a server-side clone...
17:30:00raptorthre bugs...
17:30:00raptorthree
17:33:00raptorok sam686: http://code.google.com/r/buckyballreaction-maskremoval/source/checkout
17:33:00raptorpushed changes to there
17:33:00raptori made a note about the three bugs in the latest commit message
17:34:00LordDVG Quit (Ping timeout: 276 seconds)
17:35:00sam686ok, got the repository, now i will be testing it..
17:38:00sam686rendering items in editor is missing except polywall, but is still selectable (can't see it)
17:38:00raptoroh yeah, that's four...
17:39:00sam686editor fails to save anything except polywall, as it probably didn't find anything except polwyall
17:40:00LordDVG has joined
17:40:00sam686there is a problem with GrenadeProjectile and too many disabled collisions..
17:42:00raptoryes, that is what I fell asleep on
17:42:00sam686http://96.2.123.136/bitfighter/GrenadeProjectile.txt looks like it is trying to call grenadeProjectile over and over again, if it wasn't the TNLAssert, it will stack overflow
17:43:00iKoda has joined
17:43:00raptorwell that's weird
17:45:00raptorthat loop was there before
17:45:00raptori think loops like for(;;) are evil
17:46:00raptorprojectile.cpp:229
17:49:00sam686Might be something to do with setMaskBits(ExplodedMask) to avoid searching database with exploaded burst..
17:49:00raptorohhh... i didn't even see that
17:50:00sam686oh, setMaskBits is for net mask, not database mask..
17:50:00sam686but, deleteObject is what sets the mask to DeletedTypeMask
17:51:00sam686and clears all other masks.
17:51:00sam686so, it will not search the database for anything in delete list
17:52:00sam686the reason for delay delete is for netObject to update, to avoid disappearing burst with no explostion
17:52:00raptordeleted is a boolean now
17:53:00raptorin gridDB.h: bool deleted;
17:53:00LordDVG Quit (Ping timeout: 258 seconds)
17:53:00sam686probably should be mObjectTypeNumber = DeletedTypeNumber
17:53:00raptorwe can't do that
17:53:00raptorwell actually
17:53:00raptori don't know
17:54:00sam686before it was mObjectTypeMask = DeletedType, which will clear all other bit mask
17:54:00raptorohhh
17:54:00raptori didn't know it cleared it all
17:54:00raptori thought that that information was kept
17:55:00sam686if it is deleted, why have it in part of database search? was that way before typeMask to typenumber conversion
17:56:00raptoryes, you are right
17:56:00LordDVG has joined
17:59:00sam686what is: phantom outlines of barriers on some levels, which level have the problem?
18:00:00raptorlet me make one for you...
18:01:00raptoractually do the following: host a game normally
18:01:00raptorthen go to the editor
18:01:00raptoradd a few invisible barriers
18:01:00raptorand test the level
18:02:00sam686i don't see anything at all, nothing, no barriers, no blue lines
18:02:00sam686when testing from editor
18:02:00raptorok, i make screenshot
18:03:00raptorhttp://96.2.123.136/upload/screenshot_0.png
18:05:00sam686ok... don't know why i don't get that lines, maybe you didn't do full rebuild?
18:05:00raptori alwys do full rebuild
18:06:00raptorall i haad to do was host a game first, then go to the editor and test an empty level
18:07:00sam686comes up no ghosting blue lines on my end.., is it any map or only CTF1?
18:08:00guest_ Quit (Ping timeout: 252 seconds)
18:08:00guest has joined
18:09:00raptoronly ctfg1
18:09:00raptorreally weird
18:11:00raptorlet's ignor that one fo rnow
18:12:00sam686is Barrier::mRenderLineSegments being cleared at all?
18:13:00raptordidn;t think i touched that...
18:13:00sam686in GameType.cpp line 2379, that is where it clears old BarrierLines, is it getting to that line or not?
18:14:00sam686during runtime
18:14:00iKoda Quit (Read error: Connection reset by peer)
18:14:00iKoda has joined
18:16:00sam686unless there a problem in Barrier::unionBarriers and zero barriers..
18:17:00raptorproblem might be that BarrierType was used for lots of things, now it is only for one thing...
18:18:00sam686in barrier.cpp line 379, try adding if(barrierList.size() != 0)
18:18:00sam686 for clipRenderLinesToPoly and see if it is still there..
18:20:00raptoractually, i think it needs to be: game->getGameObjDatabase()->findObjects((TestFunc)isWallType, barrierList);
18:21:00sam686polywall is barrierTypeNumber during game play, but maybe not editor...
18:22:00raptoryes, I am thinking that the editor used the BarrierType mask for multiple things... but we can't do that now..
18:24:00raptorok my change did nothing
18:24:00raptoralso, i added polywalls to the editor
18:24:00raptorthen removed them, and now they leave outline, too
18:24:00raptori bet it has to do with that 'deleted' problem...
18:26:00sam686i will try in my linux, will take some time to compile..
18:38:00raptor has left
18:40:00sam686looks like the problem happens in linux only, it happen on my linux..
18:41:00sam686if you start editor, add polywall and test, then move polywall and test again, the old one shows up there..
18:43:00raptor has joined
18:43:00ChanServ sets mode +o raptor
18:43:00raptoryep, that's the problem
18:44:00raptori'm actually suprised that there aren't more bugs...
18:45:00raptorso you think the grenade issue is with deleted?
18:46:00sam686yes, in GameObject::deleteObject
18:46:00sam686setting mObjectTypeNumber = 252; (or some unused number) fixes the problem
18:50:00raptorok, let me try this out...
18:50:00raptorsorry that I am sporadic in my focus - lots to do at work
18:51:00sam686another problem, editor saves 2 same polywalls of what should be only 1 polywall, on top of eachother
18:52:00raptorit works!!
18:52:00raptorlet me update the clone
18:55:00sam686in lbarrier.cpp line 365, i stick in a printf("BarrierCount: %i\n", barrierList.size()); and it prints 4 when there is only 1 polywall
18:56:00sam686in windows, it prints only 2, for 1 polywall
18:57:00raptorafter unpackPolygons?
18:57:00sam686before clipRenderLinesToPoly, after findObjects
18:58:00raptor365?
18:58:00sam686in Barrier::prepareRenderingGeometry(
18:58:00raptorok
18:58:00sam686line 378..
18:59:00sam686ops, i insert the printf at the wrong line for windows..
18:59:00raptori get zero consistently from editor, but a correct number when hosting
18:59:00raptorfor barriers
19:00:00sam686in editor, if you do only 1 polywall, does it show up as 1, or 4?
19:00:00sam686when testing
19:01:00raptorshows as 2
19:01:00raptorhaha
19:01:00sam686test again, now does it show 4?
19:01:00raptorstill 2, three times in a row
19:01:00sam686the reason for 2, is editor is saving the same polywall twice
19:02:00raptoradded two polywalls...
19:02:00sam686move polywall, then test again?
19:02:00raptornow says 4
19:02:00raptorstill two
19:02:00raptor2
19:02:00raptormoved and moved
19:02:00raptorstill two
19:02:00raptori pushed another change for deleted object
19:03:00raptorif you want to update from my clone
19:03:00raptori think the problem is that in many cases PolyWall was considered a BarrierType
19:04:00raptoralso WallSegmentType was BarrierType
19:04:00raptorgreat
19:12:00zoomber_laptop has joined
19:12:00zoomber_laptophi ikoda, guest, karamazovapy, lordDVG
19:12:00raptornow i'm searching for all uses of BarrierType and altering them to include polywalls or wallsegment
19:14:00zoomber_laptopok
19:14:00raptorhi zoomber_laptop, was talking to myself, i think
19:15:00zoomber_laptopactually,
19:15:00zoomber_laptopwhen client hosts a game from "host game" menu, they create their own mServerGame?
19:16:00sam686yes
19:16:00sam686but it is mostly gServerGame..
19:17:00zoomber_laptoppossible to control a remote gServerGame of another server?
19:19:00LordDVGhi zoomber_laptop
19:19:00sam686looks like problem with phantom outlines of barriers on some levels is gone after fixing deleteObject problem
19:20:00raptoryes, i noticed that too..
19:20:00raptornow to missing objects in editor...
19:22:00raptorgood catch sam686
19:33:00zoomber_laptop Quit (Quit: zoomber_laptop)
19:34:00sam686with the lack of TypeMask, you can remove typedef U64 BITMASK;
19:34:00sam686 in gridDB.h
19:34:00raptorthanks
19:35:00raptori'm getting BarrierCount: 4 with one Polywall now
19:35:00sam686is the editor loaded with 1 or 2 polywall (move polywall to find out)
19:35:00raptor1
19:35:00raptori just added another polywall: now count is 6
19:36:00sam686delete all polywall and test?
19:36:00raptor0
19:36:00raptoradded one back: 2
19:36:00sam686something seems strange to why it is 4 for only 1 polywall...
19:37:00sam6861. fix editor saving polywall as 2 lines in level file
19:37:00sam6862. fix something that is creating another doubles..
19:37:00raptori think it has to do with BarrierType being added to so many things before...
19:38:00watusimoto has joined
19:38:00raptorin gametype.cpp:2402
19:38:00raptorthere is getGame()->deleteObjects(BarrierTypeNumber);
19:38:00raptorshould that be more than just that one?
19:39:00sam686server first send s2cAddBarriers(verts.size()==0, ..., ...)
19:40:00sam686then it sends all non-zero size verts of all barrers
19:40:00raptori mean should is be WallType now? since WallType = BarrierType + PolyWallType
19:41:00sam686i guess yes, but not sure if that will create problems..
19:41:00raptorin a way, it is good we are removing mask, because it is forcing us to clean up some of this hierarchy mess
19:49:00raptorok, so I need a name for a category that both polywall and barrier can fit in: How about 'Wall'?
19:49:00sam686i guess that sounds about right: WallType
19:50:00raptorproblem is, a Barrier is not a subclass of WallItem in barrier.cpp
19:50:00raptorso should I rename WallItem to something?
19:50:00sam686not sure on wallitem
19:52:00raptori'm goign to ignore wall item for now...
20:01:00raptorfound another bug: can't pick up health
20:06:00Flynnn has joined
20:10:00raptorok sam686 fixed barriers in editor
20:10:00raptorif you pull from my clone again...
20:11:00raptorthat leaves three bugs that I know:
20:11:00iKoda Quit (Ping timeout: 250 seconds)
20:11:00raptor1. duplicate walls in editor sometimes
20:11:00raptor2. ship cannot pick up health/energy item
20:11:00iKoda has joined
20:11:00raptor3. many objects disappear in editor
20:12:00raptorI don't know if #1 is my fault, but 2 and 3 definitely are
20:23:00karamazovapyCHIMICHANGA!
20:23:00guest Quit (Ping timeout: 252 seconds)
20:24:00karamazovapyPointblank
20:25:00karamazovapyHello Pointblank!
20:31:00raptorPointblank is just learning and learning, never speaking
20:50:00iKodaMust be the perfect guy to on holiday with
20:51:00iKoda+go
20:51:00iKodaSomewhere in that phrase...
20:52:00raptorone day he will reach sentience
21:01:00iKodaWhat are you up to guys
21:01:00raptorbreaking stuff
21:04:00iKodaCmon give more details
21:07:00raptorok
21:08:00raptormy current project has to do with eliminating an 32 bit object mask and turning it into an 8bit object type number
21:08:00raptorand all the fun that comes with come parts of the code using the masks as flags instead of descriptions
21:09:00raptorbut, i think our SDL migration is pretty complete now
21:09:00raptorno going back...
21:09:00raptoroh and we implement PNG screenshots
21:10:00raptorbut really we are just doign lots of refactor and trying to reduce the issue list on google code...
21:17:00karamazovapyyour description of pointblank creeps me out a little bit
21:17:00raptorit's the quiet ones you need to worry about
21:19:00karamazovapy...so what is it actually?
21:19:00raptorsome experiment of Zoomber's
21:19:00raptorit actually has language analysis built it
21:20:00raptorso Zoomber can switch it from learning to speaking one of these days, and it should have a higher success of intelligent response
21:21:00karamazovapyI don't buy it
21:21:00raptorit can take subphrases of sentences we say here and piece them together
21:21:00karamazovapynotifitcan'ttellwhatwe'reactuallysaying!hahahahaha
21:22:00karamazovapyYOUWILLNEVERTAKEMEALIVEPOINTBLANK.
21:22:00raptorhaha
21:22:00LordDVG Quit (Remote host closed the connection)
21:22:00raptorand that's why AI should fear man - chaos will win
21:31:00iKoda Quit (Ping timeout: 255 seconds)
21:38:00iKoda has joined
21:44:00raptorrefactor refactor refactor refactor
21:45:00iKodaRaptor refactoring iz good
21:46:00raptorit gives the illusion I making progress...
21:46:00iKodaYes
21:47:00iKodaThen you realize you don't have any new feature for the users
21:47:00iKodaSo you feel bad
21:47:00iKodaBut satisfied
21:47:00iKodaAt the same time
21:47:00raptorhaha, so true...
21:47:00raptorhow are you hedgewars guys doing?
21:48:00iKodaWell for the iOS port I'm doing a lot of refactoring
21:48:00iKodaLike you are doing
21:48:00iKoda:p
21:48:00raptoris most of your work just Mac compatibility?
21:48:00iKodaNaw
21:49:00raptori mean - the handling of the Mac OS in relation to the project
21:49:00iKodaiOS mostly, but I try to do also some GUI improvments
21:49:00raptorApple OS
21:49:00raptorwhatever
21:49:00iKodaI'm not into algorithms or weapon development
21:50:00raptorah ok
21:51:00raptordoes QT run on iOS?
21:51:00iKodaNope
21:51:00raptorso you had to do a completely different front-end?
21:51:00iKodaYep
21:51:00raptorwith SDL?
21:51:00iKodaNo plain objc
21:51:00raptorah
21:51:00raptorugh
21:52:00iKodaCould have been worse..
21:53:00iKodaQt non on iOS is silly though
21:53:00raptorsoo... QT is silly?
21:53:00iKodaTis happens when an open framework is driven by a company
21:54:00raptordo you think it was a wise decision to go with QT for the desktop OS front-end of hedgewars?
21:56:00raptoror is that a touchy subject...
22:05:00iKodaN it is not touchy
22:05:00iKodaWhat other options are there anyways?
22:05:00raptorGLUT :)
22:05:00iKodaGtk is not so cross platform and wxwidgets is ugly
22:05:00raptorok not really
22:06:00raptorpure SDL wasn't good enough?
22:06:00raptori guess SDL uses widget bindings for wx
22:06:00raptorwhich is ugly
22:14:00guest has joined
22:26:00iKodaThe only thing I dislike about our setup
22:26:00iKodaBesides using freepascal....
22:26:00raptor:)
22:27:00iKodaIs that the front end calls the engine in a separate process
22:27:00raptoris that avoidable with the various languages you use?
22:28:00iKodaYeah but requires some work that nobody wants to do
22:28:00iKoda:p
22:28:00raptorahh.. one of those bugs...
22:29:00raptorpure tedium I bet
22:30:00iKoda:D
23:02:00raptorsam686: i foudn that when you load a file into the editor (in my clone) it duplicates all the walls
23:02:00raptorI have been going through my changes and cannot find what I did to make that happen...
23:07:00iKoda Quit (Remote host closed the connection)
23:08:00raptorit's saving duplicate objects to file some how...
23:10:00iKoda has joined
23:11:00iKodaYeah, plus gsoc is keeping everyone pretty busy
23:12:00raptori can only imagine what it takes to get your game on android...
23:13:00iKodaOur game is the living proof that technology CAN be interoperable!
23:13:00raptorhaha
23:14:00iKodaI wonder if you guys are interested in joining gsoc next year...
23:14:00raptorwe are probably way too small...
23:15:00raptor'not notable' according to wikipedia overlords
23:15:00raptorthey even deleted the history and talk page after deleting our game page..
23:16:00iKodaYeah we risked that too
23:16:00iKodaBuuut being small is not a real problem
23:16:00raptorhow did you convince the wiki-gods?
23:17:00iKodaI wasnt yet really involved in the proj back then
23:18:00raptorsam686: fixed duplicate wall problem
23:18:00iKodaBut being small shouldn't be a showstopper
23:19:00raptori keep joking that someday we'll port bitfighter to java...
23:19:00raptormaybe that could be a GSOC project? :)
23:21:00iKodaJava is so old fashioned
23:22:00raptor.NET?
23:22:00iKodaPort it to objc!!!
23:22:00raptorwait what
23:22:00iKodaYup
23:22:00raptori have never heard someone EVER say to port something to Obj-C
23:28:00raptorwell gotta go. see you all later
23:29:00raptor Quit (Remote host closed the connection)
23:58:00iKoda Quit (Quit: K thx bai)

Index Search ←Prev date Next date→

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