#bitfighter IRC Log

Index Search ←Prev date Next date→

IRC Log for 2011-12-18

Timestamps are in GMT/BST.

00:01:00Watusimoto:-)
00:02:00raptorhi
00:02:00raptorbuggy?
00:02:00raptorhorse-and-
00:26:00raptorWatusimoto, sam686: just curious, what are you guys working on?
00:29:00WatusimotoI'm working on getting despondent over this console of ours
00:30:00raptorare you trying to add a feature
00:30:00Watusimotono, just trying to fix bugs
00:30:00Watusimotomostly to do with the cmd history
00:31:00raptorah the history - i just figured it'd never work right...
00:31:00Watusimotoit will
00:31:00raptorI took a crack a the 'Toggle script results (Ctrl + R)' bug
00:31:00raptoragain
00:32:00Watusimoto:-)
00:32:00raptorbut i can't debug with eclipse like I need to be able to, to solve it
00:32:00raptorso it drove me crazy, again...
00:32:00raptorbut maybe you could clear something up for me...
00:32:00Watusimotomaybe
00:33:00raptorin 015a, what was the getGridDatabase()?
00:33:00raptorin UIEditor, and how does it relate to getEditorDatabase()
00:33:00Watusimotowe didn't use database in the editor
00:34:00Watusimotoexcept possibly, in a limited way
00:34:00Watusimotodon't remember exactly
00:34:00raptorso it was used like this (in 015a): LuaLevelGenerator(name, args, mGridSize, getGridDatabase(), this, gConsole);
00:34:00WatusimotoI think we used it to search efficiently for things engineered items might interact with
00:35:00raptornow it's like this: LuaLevelGenerator levelGen(name, folderManager->luaDir, args, getGame()->getGridSize(), mLoadTarget, getGame(), gConsole);
00:35:00Watusimotothe general idea is that items generated by the levelgen are handled separately so they can be rendered differently and be deleted later
00:35:00Watusimotothat's really all that needs to happen
00:36:00raptoryes - but mLoadTarget is never actually being filled
00:36:00WatusimotoMy thought was they would go into a separate database
00:36:00Watusimotoit could be that it never worked after porting to 016
00:37:00Watusimotobut we could also add a flag to editorObject that marks items as belonging to the levelgen, and keep all objects in a single database
00:37:00Watusimotoexcept we'd need to think about how undo/redo worked in that context
00:37:00raptorwell either way, i am not seeing how anything is being added to any database with LuaLevelGenerator
00:37:00Watusimotomaybe they're not
00:37:00Watusimotodo you see the objects in the editor?
00:37:00raptoronly a straight line
00:38:00raptorso yes-ish
00:38:00Watusimotoworking with mazeracer?
00:38:00Watusimotoah
00:38:00Watusimotoyou are seeing the centerlines
00:38:00raptoryes
00:38:00Watusimotoof the walls
00:38:00Watusimotook
00:38:00raptorwhich i figure is a different bug
00:38:00Watusimotoyes
00:38:00raptori'm first trying to solve the addition/removal part
00:38:00Watusimotoso the wall centerlines are being added to soemthing
00:38:00Watusimotobecause you see them
00:39:00raptoryes, and i cannot seem to track down that 'something' in LuaLevelGenerator
00:39:00Watusimototry commenting chunks out
00:39:00Watusimotoor tell me where in the code you think things are happening
00:39:00Watusimotoand i'll take a quick look
00:42:00Watusimotough, happens in the constructor
00:42:00raptorso in the LuaLevelGenerator constructor is this: mGridDatabase = gridDatabase;
00:42:00Watusimotook, I see why you're confused
00:43:00raptorbut i don't see how mGridDatabase is filled from the runScript() method
00:43:00Watusimotowhere does runScript get called from?
00:44:00Watusimotoah
00:44:00Watusimotosee it
00:44:00raptorUIEditor.cpp:603
00:44:00Watusimotook
00:44:00Watusimotodid you trace to the pcall?
00:44:00Watusimotolua_pcall?
00:44:00Watusimotoline 655
00:44:00raptorin runMain()?
00:45:00Watusimotoof luaObject.cpp
00:45:00Watusimotoyes
00:45:00raptoryes
00:45:00Watusimotoinside there somewhere
00:45:00Watusimotothat is where it happens
00:45:00raptorbut that is where i ended
00:45:00raptorreally?
00:45:00Watusimotoreally
00:45:00Watusimotothat launches the script
00:45:00Watusimotothe script then creates the objects via callbacks to C++
00:45:00Watusimotoand those callbacks insert objects into the database
00:45:00raptorhmmm...
00:46:00raptorok
00:46:00raptorcallbacks
00:46:00Watusimotogood thing you asked; you probably wouldn't have found that
00:46:00raptorgotta love those
00:46:00Watusimoto:-)
00:46:00Watusimotocallbacks maybe not the right word
00:46:00Watusimotothey are the functions that the lua script calls to add an object to a level
00:46:00raptormore like abstracted insanity
00:46:00raptor'hidden method redirection'
00:46:00Watusimotook, this isn;t abstract or hidden
00:47:00Watusimotothis is just a lua script running
00:47:00raptorok
00:47:00Watusimotoand doing some things
00:47:00raptori'm just joking..
00:47:00Watusimotoso the script might run a line that looks like this:
00:47:00Watusimoto(i know)
00:47:00Watusimotolevelgen:add_wall(some args)
00:47:00Watusimotoor something like that
00:47:00raptorah, and it calls the appropriately wrapped c++ method
00:47:00Watusimotothat calls a c++ process called add_wall(dfadsf)
00:48:00raptorok
00:48:00Watusimotoand add_wall creates a wall object and sticks it in the database
00:48:00raptori get it
00:48:00raptornow to find out why the appropriate database isn't used
00:48:00raptorthanks!
00:48:00Watusimotobut I totally see why you didn't see that
00:49:00Watusimotothe lua_getglobal(L, "_main"); finds the global var called _main
00:49:00Watusimotowhich in our case is a function, and puts it on the stack
00:49:00Watusimotothe lua_pcall runs whatever is on the top of the stack (hopefully it's a function)
00:49:00raptorok
00:50:00Watusimotoremember that even functions are variables
00:50:00Watusimotoif you read that link I sent you the other night
00:50:00Watusimotoeverything is a variable
00:50:00Watusimotoin lua
00:50:00raptoryes
00:50:00raptorjust takes a little getting used to..
00:50:00Watusimotoindeed it does
00:51:00Watusimotoso after the lua script gets run, we jump back to the editor to do what we want with our new db full of objects
00:51:00Watusimotoalles klar, herr komisar?
00:52:00Watusimoto(though if you were insulated from popular music in your youth, you might not recognize that line)
00:52:00Watusimoto(and you might be too young anyway...)
00:52:00raptori grew up with mozart
00:54:00WatusimotoHere's what you missed
00:54:00Watusimotohttp://www.youtube.com/watch?v=wbFEaOH31fY
00:54:00Watusimoto1:24
00:55:00raptori think i've heard that riding in my older brother's car
00:56:00raptoroh yeah, i remember this, but i was < 10 years old at the time i think
00:57:00Watusimotothis was jr high time for me, I think
00:57:00Watusimotowell, time for bed. I'll fix this stupid console bug tomorrow
00:57:00Watusimotogot what you need for now?
00:57:00raptornight!
00:57:00raptoryes
00:57:00raptorand thank you
00:58:00Watusimotook good
00:58:00Watusimotono problem
00:58:00WatusimotoI'd rather give you the knowledge to fix it than fix it myself :-)
00:58:00raptorand i'd rather have the knowledge, too...
00:58:00WatusimotoI'm lazy that way
00:58:00Watusimotoin the best tradition of Perl programmers
00:58:00raptorha!
00:59:00Watusimotoyou know the quote?
00:59:00raptorwhich one by Larry Wall
00:59:00WatusimotoLaziness Impatience Hubris
00:59:00Watusimotoyes
00:59:00raptorwhy do things one way when you can do them 9 or 10 different ways?
00:59:00WatusimotoI strive
00:59:00Watusimotowell, those traits are not language specific
01:00:00raptoryes, the number is definitely higher for PERL
01:00:00raptorLaziness Impatience Hubris
01:00:00WatusimotoI like Perl alot, but I understand why many hate it
01:00:00raptorhehe
01:00:00Watusimotoand I won;t (can't?) defend it
01:01:00Watusimotobut anway
01:01:00Watusimotosee you tomorrow
01:01:00raptori was unfortunately in the position of only ever maintaining PERL, not creating anything with it
01:01:00Watusimotothat's awful
01:01:00raptoryeah...
01:01:00raptorwell good night!
01:05:00raptorha! i think the mLoadTarget is being filled - it's just being filled with a WallType
01:05:00Watusimoto Quit (Ping timeout: 248 seconds)
01:07:00raptorNOT being filled with a WallType, I mean
01:22:00koda Quit (Quit: koda)
01:46:00raptor Quit (Ping timeout: 252 seconds)
02:03:00raptor has joined
02:03:00ChanServ sets mode +o raptor
02:03:00raptorsam686: are you around?
02:04:00sam686?
02:04:00sam686hi
02:04:00raptorhi
02:04:00sam686am I around?
02:04:00raptorare you around = paying attention in IRC
02:04:00raptorcould you help me track down a problem?
02:04:00sam686probably yes
02:04:00sam686ok
02:05:00raptorI'm trying to solve the insert level gen items in the editor with Ctrl+R problem
02:05:00raptorso I finally have the items being inserted into tthe correct database - let me get you a patch that allows this
02:07:00raptorhttp://host-78-12-107-208.midco.net/upload/levelgen_database_fix.diff
02:07:00raptor^^ there
02:07:00raptorso now when you do Ctrl+R in the editor, the objects are put into the correct database. However, when pressing Ctrl+R again, it should delete them from the database, but it doesn't
02:08:00raptorthe test is UIEditor.cpp:3575
02:09:00raptorthere it'll call clearLevelGenItems(), but it doesn't actually clear them
02:10:00sam686got done importing patch..
02:10:00raptorsorry it edits headers
02:11:00sam686I am using the faster build trick 1>build7.cpp
02:11:00sam6861>build6.cpp
02:11:00sam6861>build5.cpp
02:11:00sam6861>build4.cpp
02:11:00sam6861>build3.cpp
02:11:00sam6861>build2.cpp
02:11:00raptoroh smart
02:11:00sam686done with build / ling
02:11:00sam686link
02:13:00sam686i wonder if watusimoto want that faster build trick, if he is on a slow computer like mine..
02:13:00sam686ctrl + G is useful to goto line number in visual C++ text editor
02:14:00raptorthat doesn't work in eclipse...
02:14:00sam686well at least you know if you ever need to edit in visual c++ in windows
02:16:00sam686is the new master server not working? I can't ssh and I can't load http://199.192.229.168/
02:16:00raptorhmmm...
02:17:00raptori sometimes wonder if that VM goes to sleep sometimes..
02:17:00sam686i guess the server have either powered "OFF" or is not online..
02:19:00sam686as for editor Ctrl + R, it does insert almost everything in editor, except barriers
02:22:00raptorso the idea is for ctrl + R to insert the items into mLevelGenDatabase
02:22:00raptorwhich that patch finally fixes...
02:22:00raptorbut
02:23:00raptorthey are never removed from mLevelGenDatabase: for some reason, UIEditor.cpp:540 doesn't actually work
02:27:00sam686mLevelGenDatabase.removeEverythingFromDatabase(); does not clear mLevelGenDatabase, it only clears mAllObjects
02:28:00sam686i mean mAllEditorObjects
02:28:00raptorreally??
02:29:00raptorwait
02:29:00sam686mAllEditorObjects are in EditorObjectDatabase, mAllObjects in GridDatabase
02:29:00raptorahhh...
02:29:00sam686class EditorObjectDatabase : public GridDatabase (so EditorObjectDatabase contains GridDatabase)
02:30:00raptorhuh...
02:30:00raptorhmmm
02:30:00sam686i think you are a bit confused...
02:31:00raptorso EditorObjectDatabase needs to implement removeEverythingFromDatabase
02:31:00raptorthat also clears mAllEditorObjects
02:33:00sam686probably yes..
02:34:00raptorok, le tme see if this works
02:46:00raptorok, ready for a new patch?
02:46:00sam686ok
02:47:00raptorI think this works: http://host-78-12-107-208.midco.net/upload/levelgen_database_fix_try2.diff
02:47:00raptorbut i still need to solve inserting with CTRL+I
02:47:00raptorand showing the actual barriers...
02:50:00sam686ok, almost done compiling
02:50:00sam686almost doen compiling new patch
02:50:00sam686its linking
02:50:00sam686almost done
02:50:00sam686now i can run it..
02:51:00sam68630 minutes of typing waiting to compile...
02:51:00raptorok, brb
02:52:00sam686runtime errors? ctrl + R 1 time: wall and spawn appear
02:52:00sam6862 times: only walls disappear, but not spawn
02:52:00sam6863 times: crash
02:53:00raptor3 times crash?
02:54:00raptorhuh
02:56:00sam686that was using "6357.levelgen 4 4 50 2 119 0 1 0 0 0 1 1"
02:57:00raptorcan i get that?
02:57:00sam686it is here http://bitfighter.org/forums/viewtopic.php?t=404
02:57:00raptorthanks!
02:59:00raptoryup crazy crash: is spike my CPU first...
03:00:00sam686another problem, loading in-game makes that levelgen fail to make some flags and goalzones'
03:01:00sam686as in, level gen appear to be unable to add FlagItem or GoalZones
03:01:00raptorhmm...
03:01:00raptoryay for more bugs...
03:03:00raptorwhy woul dmy new EditorObjectDatabase::removeEverythingFromDatabase() spike the CPU like that?
03:04:00sam686i don't know why, i can't ctrl + R anymore...
03:04:00raptorit will be spiking your CPU
03:05:00raptorlet me try a different approach
03:05:00sam686i mean nothing happens, it doesn't freeze, it simply acks like it ignores it
03:05:00raptorweird
03:06:00sam686oops, i was pressing ctrl + F, it is ctrl + R!!!
03:07:00raptorhaha
03:07:00raptornow what happens? CPU spike on the forth one?
03:07:00raptor1st: add, 2nd: remove; 3rd: add; 4th: CPU spike <-- is my behavior
03:07:00sam686though, it doesn't spike my CPU, it just instantly halts in an error, inside > Bitfighter_mysql_d.exe!Zap::GridDatabase::addToDatabase(Zap::DatabaseObject * theObject=0x01ce9858, const Zap::Rect & extents={...}) Line 90 + 0x6 bytes C++
03:07:00sam686 Bitfighter_mysql_d.exe!Zap::EditorObjectDatabase::addToDatabase(Zap::DatabaseObject * object=0x01ce9858, const Zap::Rect & extents={...}) Line 725 C++
03:07:00sam686 Bitfighter_mysql_d.exe!Zap::DatabaseObject::addToDatabase(Zap::GridDatabase * database=0x01c0e358) Line 497 + 0x1a bytes C++
03:07:00sam686 Bitfighter_mysql_d.exe!Zap::BfObject::addToGame(Zap::Game * game=0x01c3a028, Zap::GridDatabase * database=0x01c0e358) Line 245 C++
03:07:00sam686 Bitfighter_mysql_d.exe!Zap::EditorObject::addToEditor(Zap::Game * game=0x01c3a028) Line 112 C++
03:07:00sam686 Quit (Excess Flood)
03:08:00sam686 has joined
03:08:00ChanServ sets mode +v sam686
03:08:00sam686stupid "Quit (Excess Flood)"
03:09:00sam686was only trying to paste a stack dump...
03:10:00raptorhaha
03:11:00raptorok, try3, i think this works: http://host-78-12-107-208.midco.net/upload/levelgen_database_fix_try3.diff
03:11:00raptori mean, doesn't crash
03:11:00sam686http://sam686.maxhushahn.com/upload/ctrl_R_problem.txt
03:11:00sam686there, thats the trace of my crash
03:11:00raptorthat's totally different than my crash...
03:16:00raptordoesn't fully work yet...
03:29:00karamazovapyZoomber:
03:30:00karamazovapy(also, don't turn around, der kommissar's in town)
03:35:00raptorweird song
03:38:00raptorthis is crazy
03:57:00raptoryay sam686, i think i got Ctrl+R working well, now...
04:01:00sam686CTRL + Record?
04:02:00raptornew master is still down...
04:04:00sam686i hope watusimoto didn't forget to pay for that new master server...
04:04:00raptorhehe
04:06:00sam686we are here: http://199.192.229.168/ neighbor server (one IP number apart) at 199.192.229.167 and 199.192.229.169 works fine, so it is limited to only us...
04:08:00raptorbe back a bit later..
04:09:00sam686ok...
04:09:00sam686i will "<@raptor> be back a bit later.." ?
04:10:00sam686or you will "be back a bit later.." ?
04:10:00sam686or who will "be back a bit later.." ?
04:46:00sam686 Quit (Ping timeout: 245 seconds)
05:10:00raptorback
05:10:00raptorand you're gone!
05:14:00sam686 has joined
05:14:00ChanServ sets mode +v sam686
05:17:00raptorhi sam686
05:21:00sam686i will be going to bed in a ew minutes..
05:21:00sam686in afew minutes
05:21:00raptorhehe
05:21:00raptorok quick question
05:21:00sam686?
05:22:00raptornew diff (you don't have to compile): http://host-78-12-107-208.midco.net/upload/levelgen_database_fix_try4.diff
05:22:00raptorCTRL +R works with your levelgen to add and remove items...
05:23:00sam686ok..
05:23:00raptorbut in UIEditor.cpp:554
05:23:00raptorthis happens when you do Ctrl+I
05:24:00raptori get a TNLAssert at gameObject.cpp:240
05:25:00raptorit's essentially trying to move the items from the levelgen database to the editordatabase
05:25:00sam686could try a removeFromGame() first...
05:27:00sam686i just tested that latest patch, CTRL + R doesn't remove from database anymore, and adds what look like zero width barriers
05:27:00raptorwhat
05:28:00raptori wish i could see the guts of the database object
05:28:00raptoreclipse doesn't follow objects so well..
05:28:00sam686actually not, but the barrier walls don't render right until you edit that wall
05:28:00raptori can't even edit the wall...
05:29:00sam686select multaple objects by left click and drag?
05:29:00raptorlet me try that...
05:30:00sam686oh it looks like i can select barriers near the center (0,0)
05:30:00raptori can't select anything..
05:30:00sam686oh, delete or move a spawn point or something, then press ctrl + Z
05:30:00sam686then it goes to correct rendering
05:31:00raptorusing 6357 levelgen
05:31:00raptori cannot select anything even after deleting/moving other object
05:32:00sam686did you use ctrl + Z
05:32:00raptoryes
05:32:00sam686try adding spawn point, then ctrl + Z
05:32:00raptorok
05:32:00sam686is ctrl + Z working? what that does is undo
05:33:00raptorit works, but levelgen stuff is untouchable
05:34:00sam686oh, maybe your changes is now different from the patch you gave me then...
05:34:00raptoryes
05:34:00raptorsorry
05:34:00sam686as ctrl + R inserts into level
05:35:00sam686for me anyways..
05:35:00raptoryeah, that newest patch effectively keep it all separate
05:35:00sam686what CTRL + R should do is toggle script, and ctrl + I inserts as it says in level editor (F1) instruction
05:36:00raptoryes, that is my goal...
05:40:00raptorwell, i won't keep you up anymore...
05:40:00sam686ok, bye..
05:40:00raptornight!
05:40:00sam686 Quit ()
06:14:00raptorkaramazovapy: are you still up?
06:40:00raptor Quit (Remote host closed the connection)
07:43:00LordDVG has joined
08:10:00karamazovapy Quit (Ping timeout: 240 seconds)
08:14:00BFLogBot Quit (Ping timeout: 240 seconds)
08:14:00Disconnected.
08:16:00-anthony.freenode.net- *** Looking up your hostname...
08:16:00-anthony.freenode.net- *** Checking Ident
08:16:00-anthony.freenode.net- *** No Ident response
08:16:00-anthony.freenode.net- *** Couldn't look up your hostname
08:16:00BFLogBot has joined
08:16:00Topic is 'Latest release 015a http://bitfighter.org | Forums: http://bitfighter.org/forums/ | GC Project: http://code.google.com/p/bitfighter/'
08:16:00Set by raptor!~raptor@unaffiliated/greenmachine on Sat Apr 30 22:51:58 PDT 2011
08:16:00-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.
08:23:00LordDVG Quit (Remote host closed the connection)
08:52:00BFLogBot Quit (Ping timeout: 240 seconds)
08:52:00Disconnected.
08:52:00-anthony.freenode.net- *** Looking up your hostname...
08:52:00-anthony.freenode.net- *** Checking Ident
08:53:00-anthony.freenode.net- *** No Ident response
08:53:00-anthony.freenode.net- *** Couldn't look up your hostname
08:53:00Disconnected.
08:53:00-anthony.freenode.net- *** Looking up your hostname...
08:53:00-anthony.freenode.net- *** Checking Ident
08:53:00-anthony.freenode.net- *** No Ident response
08:54:00-anthony.freenode.net- *** Couldn't look up your hostname
08:58:00BFLogBot Quit (Ping timeout: 240 seconds)
08:58:00Disconnected.
08:58:00-anthony.freenode.net- *** Looking up your hostname...
08:58:00-anthony.freenode.net- *** Checking Ident
08:58:00-anthony.freenode.net- *** No Ident response
08:58:00-anthony.freenode.net- *** Couldn't look up your hostname
08:59:00BFLogBot has joined
08:59:00Topic is 'Latest release 015a http://bitfighter.org | Forums: http://bitfighter.org/forums/ | GC Project: http://code.google.com/p/bitfighter/'
08:59:00Set by raptor!~raptor@unaffiliated/greenmachine on Sat Apr 30 22:51:58 PDT 2011
09:03:00-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.
09:03:00BFLogBot Quit (Ping timeout: 240 seconds)
09:03:00Disconnected.
09:03:00-anthony.freenode.net- *** Looking up your hostname...
09:03:00-anthony.freenode.net- *** Checking Ident
09:03:00-anthony.freenode.net- *** No Ident response
09:04:00-anthony.freenode.net- *** Couldn't look up your hostname
09:04:00BFLogBot has joined
09:04:00Topic is 'Latest release 015a http://bitfighter.org | Forums: http://bitfighter.org/forums/ | GC Project: http://code.google.com/p/bitfighter/'
09:04:00Set by raptor!~raptor@unaffiliated/greenmachine on Sat Apr 30 22:51:58 PDT 2011
09:04:00-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.
09:07:00Watusimoto has joined
09:31:00Watusimoto Quit (Ping timeout: 240 seconds)
09:50:00Watusimoto has joined
10:40:00LordDVG has joined
11:20:00Watusimoto Quit (Ping timeout: 252 seconds)
13:11:00koda has joined
13:38:00koda Quit (Quit: koda)
14:03:00LordDVG Quit (Remote host closed the connection)
14:35:00LordDVG has joined
14:53:00raptor has joined
14:53:00ChanServ sets mode +o raptor
15:09:00LordDVG Quit (Ping timeout: 252 seconds)
16:15:00karamazovapy_ is now known as karamazovapy
16:15:00karamazovapyhey raptor
16:29:00sam686 has joined
16:29:00ChanServ sets mode +v sam686
17:20:00raptorhi karamazovapy
17:21:00karamazovapylooking for me last night?
17:21:00raptoryes
17:22:00raptorI was curious if you happen to have a levelgen script that essentially used all available ways to add items, and that also added every type of item
17:22:00raptorbasically something I could use for a test set against 016
17:22:00karamazovapyno, for the most part I just use addlevelline
17:23:00raptorok
17:23:00raptorjust curious, thanks
17:24:00sam686hint, converting a level to levelgen is as simple as for each line, adding "levelgen:addLevelLine("LINE_GOES_HERE")
17:25:00sam686my 6357.levelgen uses only addLevelLine
17:25:00raptorah ok
17:26:00raptorwell, i gotta go - i'll be back in about 3-4 hours or so
17:26:00sam686though, my 6357 level gen use only one addLevelLine, so i can cheaply replace it with logprint
17:27:00raptori need to talk to watusimoto about an architectural problem with using GridDatabases for the levelgen code
17:28:00raptorbut i'll be back later
17:28:00raptorbye
17:28:00raptor Quit (Remote host closed the connection)
17:32:00koda has joined
18:21:00koda Quit (Quit: koda)
18:35:00LordDVG has joined
19:20:00Watusimoto has joined
19:28:00Little_Apple has joined
19:28:00Little_Applehelloo
19:28:00Little_Applecan i have the master server address plz?
19:28:00Little_Applei think bb pooped it up again
19:28:00Little_AppleHELLOOOO
19:29:00sam686can you just remove the "MasterServerAddressList=" line completely, or delete that "bitfighter.ini"
19:29:00sam686then it should go back to defaults
19:29:00Little_Appleoh
19:30:00Little_Appleduh xD
19:30:00Little_Applethanx sam
19:30:00Little_Applei did not think of that
19:30:00Little_Appleaudios
19:30:00Little_Apple Quit (Client Quit)
19:46:00sam686i think http://bitfighter.org forum is going very slow now, and the old master is kind of down as well...
19:55:00sam686oh now bitfighter.org have mysql error in forum: SQL ERROR [ mysqli ] Lost connection to MySQL server at 'reading authorization packet', system error: 104 [2013]
20:13:00WatusimotoI don't see any sql errors
20:14:00LordDVG Quit (Read error: Connection reset by peer)
20:18:00sam686well, i think it went back to working now.... except the master may still be dead
20:19:00sam686bitfighter.org master seems down, because i can't connect to master
20:19:00sam686but the website works fine, was very slow, got a one time error too..
20:22:00sam686also, the new master server, website http://199.192.229.168/ and everything in that address seem to be down as well
20:27:00sam686Watusimoto , reading my messages yet?
20:31:00Watusimotothese messages here?
20:31:00sam686yes, the messages i said here
20:31:00WatusimotoI'll check the master -- I can't connect either
20:32:00Watusimotooh, we have the restart master script!
20:32:00sam686the old bitfighter.org is probably down - BlackBird can't connect to master, but did connect to server
20:32:00Watusimotousing that right now
20:33:00Watusimotothe computer is definitely not down
20:33:00Watusimotobitfighter.org/restartmaster , by they way
20:33:00sam686still can't connect to master... yet...
20:33:00Watusimotome either
20:34:00Watusimotook, time to step it up
20:34:00sam686maybe the "restartmaster" not working right anymore?
20:34:00Watusimotono, I think it works
20:35:00Watusimotoit was definitely down
20:35:00sam686now i think i got connected to master...
20:35:00Watusimotook, I can connect now
20:36:00sam686next problem, this doesn't work http://199.192.229.168/ (where we might move bitfighter.org to)
20:36:00WatusimotoRestarted Sat Apr 2 23:14:01 EDT 2011
20:36:00WatusimotoRestarted Sun Apr 10 16:40:02 EDT 2011
20:36:00WatusimotoRestarted via restart script Sun Dec 18 15:32:45 EST 2011
20:36:00WatusimotoRestarted via restart script Sun Dec 18 15:35:07 EST 2011
20:36:00Watusimoto[eykamp@domo exe]$ date
20:36:00WatusimotoSun Dec 18 15:36:27 EST 2011
20:37:00Watusimotothis was the last few entries from the restart log
20:37:00Watusimotolooks like it tried to restart iteself
20:37:00Watusimotothe second entry is probably my manual restart
20:37:00Watusimotobut there should be a cron script that tests the server every minute and restarts it
20:37:00Watusimotoif it's down
20:38:00sam686umm does it correctly kill the old master? so it frees that UDP port number...
20:38:00Watusimotoit should
20:38:00Watusimotobut the master wasn't running at all
20:38:00Watusimotocron job is still in crontab
20:39:00Watusimotowell, no matter. we'll revisit all this when we move the server
20:39:00sam686ok, i guess it was somewhat a mystery...
20:39:00sam686but on to moving the server, problem is the new server seems to be dead,,
20:40:00Watusimotois it
20:40:00Watusimotothat's because there was a problem with my payment
20:40:00Watusimotobut it's paid, so it should be up
20:42:00Watusimotommmm.... not able to connect
20:42:00sam686it seems down, it was running web page, master, ssh, but everything seems down
20:42:00sam686for new server
20:48:00Watusimotook
20:48:00Watusimotoaccount has been suspended, I'm paid up, I submitted a ticket to get things fixed
20:49:00WatusimotoI'm going to prepay for 12 months when we get this cleared up to avoid problems
20:49:00sam686ok, so that means it is waiting for it for me and raptor....
21:19:00koda has joined
21:31:00raptor has joined
21:31:00ChanServ sets mode +o raptor
21:32:00raptorhello!
21:33:00sam686hi
21:33:00raptorWatusimoto: if you are here i have some deep architectural questions about using the GridDatabase for editor and levelgen objects
21:33:00raptorok, not that deep
21:33:00raptorhi sam686
21:33:00Watusimotohere
21:33:00Watusimotoask away
21:34:00raptorok
21:34:00raptorso
21:34:00raptori have a patch that fixes the ctrl + r problem
21:35:00Watusimotogreat!!
21:35:00Watusimototoggling of levelgen items
21:35:00raptorhowever ctrl + i doesn't work because it triggers the assert at: http://host-78-12-107-208.midco.net/upload/levelgen_database_fix_try5.diff
21:35:00raptoroops
21:35:00raptori mean this is the patch: http://host-78-12-107-208.midco.net/upload/levelgen_database_fix_try5.diff
21:36:00raptorbut
21:36:00raptorit triggers an assert on gameObject.cpp:240
21:36:00raptorbecause
21:36:00WatusimotoI see no assert in that second diff
21:36:00Watusimotoor were you refering to go 240?
21:36:00raptoryes that
21:36:00raptorsorry, too many thoughts at once
21:37:00WatusimotoTNLAssert(mGame == NULL, "Error: Object already in a game in GameObject::addToGame.");
21:37:00Watusimotook
21:37:00raptorso, basically, since we are using EditorObjectDatabase for objects now
21:37:00sam686watusimoto, there is an easy ctrl+G to goto line number in a file, in visual c++
21:37:00raptorwhen calling addObject to add to *any* database, it ties the object to a game object (ClientGame)
21:37:00Watusimotoyes -- I just used it :-)
21:38:00Watusimotook
21:38:00raptorthe problem is, for ctrl + I to work, i need to move the objects from mLevelGenDatabase to the editor object database
21:38:00Watusimotocorrect
21:38:00raptorand that is done by calling 'addObject' again
21:38:00Watusimotowithout triggering the assert
21:38:00Watusimotocorrect
21:38:00raptorwhich triggers the assert
21:38:00Watusimotoright
21:39:00Watusimotowhy? because it is already part of the game
21:39:00Watusimotoright?
21:39:00raptorright
21:39:00Watusimotook
21:39:00Watusimotoso
21:39:00Watusimotowhy not remove it from the game before transferring it to the other db?
21:39:00sam686removeFromGame() or simillar?
21:39:00Watusimotothere's a method to do that somewhere
21:39:00Watusimotoprobably
21:40:00Watusimotoah
21:40:00Watusimotono, that should be ok
21:40:00Watusimotouse removeFromGame()
21:40:00Watusimotoor
21:40:00Watusimotojust set mGame to NULL
21:41:00Watusimotobut I sense that you think this is all horribly hacky
21:41:00Watusimotoand it makes you uneasy for some reason
21:41:00raptorit's already hacky
21:41:00raptorand i've used removeFromGame() and it didn't work:
21:41:00raptor for(S32 i = 0; i < objList->size(); i++)
21:41:00raptor objList->get(i)->removeFromGame();
21:42:00raptorassert still triggered
21:42:00Watusimotothe same one?
21:42:00raptorlet me test that...
21:42:00Watusimotowell, the other one probably won't trigger regardless
21:43:00raptoroh, yes, the same 'type' of assert is triggered (i was checking to see if it was the same object)
21:44:00raptoryeah first object in the list always triggers it
21:44:00Watusimotowell, that's odd
21:44:00Watusimotoso you removeFromGame, which sets mGame to NULL
21:44:00raptoraparently it didn't...
21:44:00Watusimotothen you addToGame, which asserts on mGame not being NULL?
21:44:00raptoryes
21:45:00Watusimotowell, it's not overridden anywhere... how can mGame not be set to NULL?
21:45:00raptoryeah, it's crazy
21:45:00Watusimototry for(S32 i = 0; i < objList->size(); i++)
21:45:00Watusimoto{
21:45:00WatusimotoobjList->get(i)->removeFromGame();
21:46:00WatusimotoTNLAssert(objList->get(i)->mGame == NULL, "C++ is broken!");
21:46:00Watusimoto}
21:46:00raptorwhat the crazy
21:46:00raptorok, so i put them in the same loop and now it works on the first object
21:46:00Watusimotobut what, remove then add?
21:47:00raptoryeah, i had it in separate loops before, now it looks like this:
21:47:00raptor for(S32 i = 0; i < objList->size(); i++)
21:47:00raptor {
21:47:00raptor EditorObject* obj = objList->get(i);
21:47:00raptor obj->removeFromGame();
21:47:00raptor obj->addToEditor(getGame());
21:47:00raptor }
21:47:00raptorhaha
21:47:00Watusimotoand that works
21:47:00raptorwell, not really...
21:47:00raptorloop synchronization issues
21:47:00raptorso half the objects did insert :)
21:48:00Watusimotoloop synchronization issues??
21:48:00Watusimotoare you threading this?!?
21:48:00raptorlike, remove from the loop your looping through...
21:48:00Watusimotoah
21:48:00raptorlet me set up a tempList
21:49:00Watusimotoyes, I see
21:49:00raptorin fact this may have been the problem all along
21:49:00Watusimotothere is something wrong with this design, I think
21:50:00raptorwhat does putting 'const' on the front of this actually do?: const Vector<EditorObject*> *objList = mLevelGenDatabase.getObjectList();
21:50:00WatusimotoI think it makes the Vector immuable ?
21:50:00raptoryes there is something wrong with the design
21:50:00WatusimotoI think to make the ptrs immutabnle you'd do Vec<const EditorObj *>
21:51:00raptorbefore we just used simple vectors for editorobject storage, but now since we're using the GridDatabase we're having to deal with a lot more garbage
21:51:00Watusimotoyes
21:51:00Watusimotobut the griddb brings some big advantages, such as spatial searches
21:51:00raptoryes
21:52:00sam686a crash may happen when deleting DatabaseObject without removing that object from database first.
21:52:00Watusimotoperformance really sucked when I was looping through all items lookig for things that a turret could grab on to
21:52:00sam686it might help to add this: DatabaseObject::~DatabaseObject() { removeFromDatabase();}
21:54:00raptorok let's see if this compiles..
21:56:00raptorit works!!
21:57:00Watusimotogreat
21:57:00sam686grrr not again, patch didn't change any file because i used a "Import from clipboard" and only imported a link into discription, and not the actual patch changes...
21:57:00Watusimotowhat does your loop look like?
21:58:00Watusimotooops
21:58:00Watusimotoon a brighter note, oglConsole now works much better in the history department
21:58:00Watusimotostill not perfect, but good enough for 016
21:58:00Watusimotothe things that REALLY annoyed me are fixed
21:59:00Watusimotoas are several out-and-out bugs
22:00:00raptorit works!!!!!!!!!!!!!!!!!!
22:00:00Watusimotogreat
22:01:00raptorman, what a study this bug has been...
22:04:00Watusimotohave you learned anything we can use to improve things?
22:04:00raptornot really, other than code clean-up
22:05:00Watusimoto:-)
22:05:00raptoroh and documentation
22:05:00Watusimotothe database design is not really that good -- at least not the part where I tried to adapt it to the editor
22:05:00raptori did find another bug
22:06:00raptorcreate a barrier
22:06:00raptormove it twice
22:06:00raptorundo
22:06:00Watusimotogood.... I think
22:06:00Watusimotowhat?!?
22:06:00raptormove it again
22:06:00Watusimotoyou didn't fix that did you?
22:06:00raptorsee the undo shadow as the previos move
22:06:00Watusimotoah, ok
22:06:00Watusimotorats
22:07:00Watusimotoso there's still the larger issue i need to clean up
22:07:00raptorthe goofy undo bugs?
22:07:00Watusimotoyes; items become unselectable sometimes,
22:07:00raptori still can't get that reliably..
22:07:00Watusimotobut maybe it's not always the fault of undo
22:07:00sam686editor crash when pressing ctrl + z while dragging barrier...
22:07:00WatusimotoI can get it often but not reliabley
22:08:00Watusimotoshould be easy to fix -- no undo while dragging
22:08:00WatusimotoI'll fix that
22:08:00Watusimototrivial
22:09:00raptorok, i think i'm ready to commit...
22:09:00Watusimotoactually, raptor, could you fix it?
22:09:00Watusimoto else if(inputString == "Ctrl+Z") // Undo
22:09:00Watusimoto {
22:09:00Watusimoto if(!mCreatingPolyline && !mCreatingPoly && !mDraggingObjects && !mDraggingDockItem)
22:09:00Watusimoto undo(true);
22:09:00Watusimoto }
22:09:00Watusimotothat if() is what needs to change
22:09:00Watusimototo what I just put there
22:09:00raptorok
22:09:00Watusimotothat will reduce checkin conflicts
22:09:00Watusimotobut that should be the fix
22:10:00Watusimotothanks
22:10:00raptorok
22:11:00sam686oh and maybe also ctrl + alt + Z might also crash while dragging..
22:11:00Watusimotoyou mean redo?
22:11:00sam686yes
22:11:00Watusimotoah, yes. raptor, please add those same two conditions to
22:11:00Watusimotoelse if(inputString == "Ctrl+Shift+Z") // Redo
22:11:00Watusimoto {
22:11:00Watusimoto if(!mCreatingPolyline && !mCreatingPoly)
22:11:00Watusimoto redo();
22:11:00Watusimoto }
22:11:00raptorok
22:12:00raptordone
22:12:00Watusimotomaybe we should create a function called undoEnabled() or something that contains those 4 conditions
22:12:00raptori wo nder what a multi-dimensional karnaugh map of all the editor flags would look like for any given operation...
22:12:00Watusimotowhat does that mean?!?
22:13:00raptorjust thinking about an old digital logic class
22:13:00raptorthat made us do karnaugh maps by hand
22:13:00raptorignore me
22:14:00sam686ctrl + C while dragging barrier makes the rendering change (harder to see what you are dragging)
22:14:00sam686i guess redo can't be done while dragging anyway...
22:16:00sam686"Master server - sending request" gets in the way while in editor..
22:16:00sam686can't see words at the bottom left
22:16:00Watusimotothis is what our christmas tree looks liek
22:16:00Watusimotohttp://www.google.com/imgres?hl=en&safe=off&client=firefox-a&hs=C5P&sa=X&rls=org.mozilla:en-US:official&tbm=isch&prmd=imvnsuo&tbnid=6bknOC2D-NQa1M:&imgrefurl=http://www.instructables.com/id/Make-a-Charlie-Brown-Christmas-Tree/&docid=VDp3ldXLJOjTjM&imgurl=http://www.instructables.com/files/deriv/FU7/9HFV/G33OQ7QC/FU79HFVG33OQ7QC.MEDIUM.jpg&w=500&h=315&ei=G2buTsuKJYWV8gOkgpGfCg&zoom=1&biw=1366&bih=613
22:17:00Watusimotosam -- yes, I noticed that myself
22:17:00raptorok, i'm ready to push
22:17:00Watusimotoraptor -- do it
22:17:00raptorhahaha
22:17:00Watusimotomy son thought no one would buy it. he named it "tree"
22:17:00sam686also, editing attribultes while in fullscreen causes everything to change size...
22:18:00Watusimotosam -- what do you think we should do about that text conflict?
22:18:00Watusimotooh brother
22:18:00WatusimotoI'll look at the editing attributes bug
22:19:00sam686umm maybe don't try to connect to master except at main menu, server list, chat lobby
22:19:00Watusimotomy paper list is growing again
22:19:00raptori'm trying!!\
22:19:00Watusimotoor maybe don't display the message in the editor?
22:19:00raptori vote don't display
22:20:00WatusimotoI agree, at least for 016
22:20:00sam686not having it display might be better
22:20:00Watusimotoit's on my paper list
22:21:00BFLogBot - Commit 72134159d588 | Author: buckyballreaction | Log: Fix CTRL + R/I levelgen preview/insertion
22:21:00BFLogBot - Commit 4ef461a58b07 | Author: buckyballreaction | Log: No crashing when undo/redo is triggered whil dragging an object
22:21:00raptorWatusimoto: i ask that you look at that fix
22:21:00raptor72134159d588
22:22:00raptorjust to make sure i really didn't mess something up...
22:22:00Watusimotois it checked in?
22:22:00raptory
22:23:00Watusimotogot it
22:23:00Watusimotook, added database to processing psuedo items... looks good
22:24:00raptoryeah, that was the sneaky problem
22:25:00Watusimotodoesn't look like there's much to go wrong
22:25:00Watusimotopretty straightforward
22:26:00WatusimotoIt did occur to me that the removeEverythingFromDatabase() loop can be done without a temp vector
22:26:00Watusimotoby doing an i-- after deleting each item
22:26:00raptoroh yeah duh...
22:26:00Watusimotomore efficient, but hardly critical
22:27:00Watusimotoso yeah, I think it looks good
22:29:00raptori *always* forget about decrementing because at work i use hashmaps everywhere
22:30:00raptorwith java
22:31:00BFLogBot - Commit 691b4e019dc5 | Author: watusim...@bitfighter.org | Log: Strip out debugging code
22:31:00BFLogBot - Commit 572662b60ef9 | Author: watusim...@bitfighter.org | Log: Show cursor when scrolling back in history in oglConsole
22:31:00BFLogBot - Commit b58c86680110 | Author: watusim...@bitfighter.org | Log: Make oglConsole cursor blink, fix small history bug
22:31:00BFLogBot - Commit 7b055e11c193 | Author: watusim...@bitfighter.org | Log: Possibly more efficient?
22:31:00BFLogBot - Commit 85e35e62d184 | Author: watusim...@bitfighter.org | Log: Fix yet another self-inflicted cmd history bug
22:31:00BFLogBot - Commit 41333270f7c1 | Author: watusim...@bitfighter.org | Log: In oglConsole, can now type backspace to start editing currently selected history line
22:31:00BFLogBot - Commit 114cef6edc03 | Author: watusim...@bitfighter.org | Log: Remove unneeded variable, fix yet another self-inflicted injury, everything seems to be working as intended at the moment.
22:31:00BFLogBot - Commit ac826fefcc6e | Author: watusim...@bitfighter.org | Log: Remove spurious calls to wrap
22:31:00BFLogBot - Commit 78d784346126 | Author: watusim...@bitfighter.org | Log: Merge
22:31:00raptorWOW
22:31:00raptortime to test..
22:31:00raptoroh, Watusimoto, did you update to the latest mercurial?
22:33:00raptorplease say you modeled after linux bash instead of mac bash
22:34:00Watusimotoyes
22:34:00Watusimotobut not completely
22:34:00Watusimotobut mostly
22:34:00raptoryeah, har dto do completely without including bash itself :)
22:34:00WatusimotoI'd liek to do it more, but that will take more work, and it is good enough for now
22:34:00raptoroh yay, backspace works
22:34:00sam686well... there is my bug list, with some changes I made. http://sam686.maxhushahn.com/wiki/index.php/Bitfighter016_bug
22:35:00Watusimoto21 is fixed, right raptor?
22:35:00raptoryes!!
22:35:00raptorhowever
22:35:00sam686well it may be just been fixed
22:35:00sam686but i haven't fully checked that yet
22:36:00Watusimoto34 and 35 are on my paper list
22:36:00raptorthere is still a render bug
22:36:00sam686what render bug?
22:36:00raptorfor some reason walls in the levelgen DB are not rendered properly
22:36:00raptorshould be simple compared to what i had to go through to get the DB working
22:39:00raptorWatusimoto: the console is much more operational now
22:40:00Watusimotoyes
22:40:00sam686maybe because levelgen needs some way to calculate clipper in alternative database...
22:40:00Watusimotonot perfect, but almost usable
22:41:00Watusimotowe should probably pass a db to clipper
22:41:00raptorsam686: that would probably be done at the bottom of EditorUserInterface::runScript
22:42:00raptori think that last for loop on line UIEditor.cpp:617 is trying to make sure clipper can run...
22:42:00sam686also, Barriers may actually be split into individual wall segments, maybe that part is missing when adding walls into alternative database
22:42:00Watusimotofrom oglConsole...
22:42:00Watusimotodouble d; /* bra size */
22:42:00WatusimotoI just love that comment
22:43:00Watusimotothe sheer raw youthful vigor!
22:43:00raptorhaha
22:43:00Watusimotoso completely pointless
22:43:00Watusimotoand yet...
22:43:00Watusimotoit's just brash
22:45:00raptorso here: http://pastie.org/3017178
22:45:00raptorcan be altered, yes?
22:45:00raptori mentioned this one to karamazovapy: 'make cmds for inserting items in editor all ctrl-xxx based'
22:45:00raptorhe said it was unnecessary
22:45:00raptorand that he used the quick keys exclusively when building maps
22:52:00Watusimotoyes, sure
22:52:00Watusimotoalter away
22:53:00raptori cannot alter for some reason...
22:53:00Watusimotothen copy/paste into a new pastie!
22:53:00Watusimotothey're free!
22:53:00raptorhaha, can you tell i'm of a sound mind right now?
22:53:00Watusimotolike GUIDs... once I learned how they worked, I wanted to waste them!
22:54:00sam686i don't think there is an edit in pastie...
22:54:00Watusimotofixed another oglconsole bug
22:54:00WatusimotoI can edit it
22:54:00Watusimotorefresh the pastie
22:55:00sam686haha, first line: HI RAPTOR!!!
22:55:00Watusimoto:-)
22:55:00WatusimotoI have an edit link off to the right
22:55:00Watusimotoabout 25% way down the page
22:55:00Watusimotojust below the embed box
22:56:00raptorthat means you save your cookies
22:56:00raptori tend to toss mine...
22:56:00sam686i don't see no edit button for me, i will show you a screenshow of what i see
22:57:00raptorok updated it: http://pastie.org/3038147
22:57:00sam686http://sam686.maxhushahn.com/upload/pastie_no_edit_button.PNG
22:58:00sam686but thats different then http://pastie.org/3017178
22:58:00sam686different link
22:59:00sam686wiki would have been better then pastie for all the editing needs, even bitfighter.org wiki might work... we can edit together...
23:00:00raptoryes, i agree, but pasties are so fast!
23:01:00Watusimotofixed my next big annoyance with oglcrapsle, whcih is the mangled characters that randomly appear
23:01:00BFLogBot - Commit 3bca3dc02bff | Author: watusim...@bitfighter.org | Log: Make console more opaque to reduce visual conflict with editor
23:01:00BFLogBot - Commit edec49ab45f3 | Author: watusim...@bitfighter.org | Log: Fix deformed oglConsole characters
23:01:00BFLogBot - Commit cbf6c5170a15 | Author: watusim...@bitfighter.org | Log: Make sure colored console bakground fills entire screen. Lame way to do it, but easier than fixing it properly.
23:03:00sam686or maybe a bug list could be in an HG file...
23:03:00sam686to read while offline.
23:05:00raptorthe wiki is definitely best...
23:07:00sam686I wonder if we can stick all the CPP files into wiki, hahaha...
23:09:00raptorbe back in about an hour...

Index Search ←Prev date Next date→

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