#bitfighter IRC Log

Index Search ←Prev date Next date→

IRC Log for 2013-05-09

Timestamps are in GMT/BST.

00:54:27Watusimoto has joined
02:14:57LordDVG has joined
04:06:49Platskies has joined
04:38:24Platskies Quit (Remote host closed the connection)
06:20:19Watusimoto Quit (Ping timeout: 276 seconds)
08:00:06raptor has joined
08:00:06ChanServ sets mode +o raptor
08:01:48bobdaduck has joined
08:02:04bobdaducksup bros
08:07:25raptorgood morning!
08:14:33LordDVG Quit (Remote host closed the connection)
08:31:54Watusimoto has joined
08:36:29Watusimotohello
08:40:14thread_ has joined
08:40:56thread_Ohayou
08:42:15bobdaduckSorry about last night thread_
08:42:25bobdaduckI ended up not being able to get on until like 10:30
08:42:28thread_its fine
08:53:49raptorhello
08:54:48Watusimotohey, I used testing to fix a real bug!
08:54:55raptorOOooo
08:54:57raptorreally?
08:55:01Watusimotoreally!
08:55:09raptorcan't be!
08:55:23Watusimotoa loadout related bug -- if you changed your loadout you didn't retain your selected weapon
08:55:36Watusimototo be fair, I found the bug before I wrote the test
08:55:38Watusimotobut still
08:55:51raptorhah
08:55:56Watusimotowe now have test coverage for it!
08:56:01Watusimotoup to .0001%!
08:56:09raptoryessss
08:56:43raptoroh
08:56:45WatusimotoI also discovered a potential bug that could be exposed if we changed the way things were arranged slightly
08:56:53raptorso, i fixed the last of the cmake on Windows bugs (I think)
08:56:56Watusimotoand wrote test coverage for that
08:56:58Watusimotogood
08:57:05WatusimotoI fell asleep after dinner last night
08:57:17raptoryeah i figured :)
08:57:27WatusimotoI am getting endless loops in the database now
08:57:37raptor?
08:57:43Watusimotoeach "bucket" is a linked list
08:57:50sam686 Quit (Ping timeout: 245 seconds)
08:57:51Watusimotoand items are pointing back at themselves
08:58:06Watusimotoso traversing a database list takes... a while
08:58:12raptor is not following
08:58:39Watusimotook, the grid database consists of 16x16 "buckets" or lists of items that are in each cell on the map
08:58:51raptorahh... the *grid* database
08:58:56Watusimotoyes
08:58:59Watusimotosorry
08:59:02raptori follow now
08:59:25WatusimotoI cannot figure out how this could happen, and why if it is happening now it never happened before
08:59:40Watusimotobecause we've done nothing (I think) to any of the related code/classes
08:59:44Watusimotoand yet... it is
08:59:55Watusimotoso I'm doing a full rebuild
09:00:15raptoryou're saying a database search is looping too much?
09:00:38WatusimotoI'm saying that we have a linked list that has an entry that points to itself
09:00:51Watusimotoso we usually traverse a list until we hit a NULL
09:01:08Watusimotothis will never happen when the list points back on itself... it will just traverse forever
09:01:22Watusimotoassuming you know how a linked list works
09:01:25raptorok, i understand that
09:01:27raptoryes
09:01:54raptordoubly-linked?
09:01:58Watusimotono
09:02:25Watusimotoanyway... the error has to be in the memory allocator db uses
09:02:37WatusimotoHere, the code is simple:
09:02:39WatusimotoGridDatabase::BucketEntry *be = gridDB->mChunker->alloc();
09:02:39Watusimoto be->theObject = this;
09:02:39Watusimoto be->nextInBucket = gridDB->mBuckets[x & gridDB->BucketMask][y & gridDB->BucketMask];
09:02:39Watusimoto TNLAssert(be != be->nextInBucket, "creating endless loop!");
09:02:52Watusimotowe allocate a new be pointer
09:03:17Watusimotothe bucketEntry is a struct with two items: theObject and nextInBucket, both pointers
09:03:26raptorok
09:03:32Watusimotoso we assign thObject to the thing we're inserting into the database
09:03:40raptorthere should be a 'key' right? where both are NULL
09:03:46Watusimotothe assign the next pointer to whatever is in the database
09:04:04Watusimotothus inserting be as the first item on the linked list
09:04:53Watusimotowhen there are no items in a particular bucket, gridDB->mBuckets[x & gridDB->BucketMask][y & gridDB->BucketMask] should be NULL
09:05:16Watusimotoso be->nextInBucket = gridDB->mBuckets[x & gridDB->BucketMask][y & gridDB->BucketMask] should be be->nextInBucket = NULL
09:05:52raptoryes
09:05:58Watusimotosince we just allocated be, that chunk of memory should not be allocated to anything yet... clear and virginal
09:06:06raptorha yes
09:06:12Watusimotoso that assert (which I just added) should never trip
09:06:26raptorwhat if both are NULL?
09:06:26Watusimotobecause be->next is null
09:06:45Watusimotoif be were null (which it isn't), then be->theObject = this; would break
09:07:20Watusimotothe only way I can see this failing is if *be = gridDB->mChunker->alloc(); is handing out already used pointers
09:07:48Watusimotoand... well, why would that suddenly start happening?
09:08:02Watusimotoso I am hoping I just misunderstand something
09:08:20raptorfull recompile done?
09:08:25Watusimotoyes
09:08:29Watusimotodidn't fix
09:08:35WatusimotoI'm adding a be != NULL assert
09:08:54Watusimotomaybe I'll start printing out the allocated chunk addrs so I can see what's happening
09:09:01Watusimotoweird stuff
09:09:15raptorbut it works in-game! and I haven't changed anything..
09:09:49Watusimotoexactly
09:16:35Watusimotowow
09:16:36WatusimotoGridDatabase::BucketEntry *be = gridDB->mChunker->alloc();
09:16:36Watusimoto TNLAssert(be != NULL, "be is NULL!");
09:16:36Watusimoto printf("%p, %p\n", be, gridDB->mBuckets[x & gridDB->BucketMask][y & gridDB->BucketMask]);
09:16:52Watusimotoprint's the same addr twice
09:17:07Watusimotothat's not right
09:17:30raptorhmm
09:18:45Watusimotobad news, raptor
09:18:49Watusimotoyou touched that file last!
09:18:56WatusimototnlDataChunker
09:18:56raptordid I?
09:19:01Watusimotoyup
09:19:10Watusimotothe good news is it was in Dec 2010
09:19:15bobdaducklol
09:19:15raptorhaha, ok phoew
09:20:32Watusimotoand I've checked my local changes, and I haven't done anythign either. So this is really weird
09:20:33WatusimotoI'm going to try running from a game and see if I hit any of my asserts
09:20:33Watusimotoit is possible there is some initiailizatin that needs to happen
09:20:50raptorlooking through the datachunker class... it's crazy!
09:21:16Watusimotook, the game does not trip the asserts
09:21:51Watusimotoand I can see it hitting them like crazy
09:21:54Watusimotoso
09:22:11Watusimotothat explains why the game hasn't been crashing since 12-2010
09:25:15bobdaducklol
09:26:47raptorha!
09:26:52raptorguess what my change was
09:26:57raptorconverting line endings
09:27:11raptorno differences when whitespace is excluded
09:28:42bobdaduckIf I have an array of playersInfo, and I do something like local players["coolnessFactor"], what happens?
09:28:56Watusimotoyou'll get a 0
09:29:03Watusimotoha ha
09:29:08bobdaduckxD
09:29:36WatusimotoI actually don't understand your question
09:29:44Watusimotoare you asking what happens if you use a string as an array index?
09:29:48bobdaduckI don't understand arrays+dictionaries+keys
09:30:09Watusimotoin lua, arrays and dictionaries are essentially the same thing
09:30:19Watusimotoarrays are just dictionaries with numeric keys
09:30:22bobdaduckOkay, but I still have no idea how to actually use them
09:31:06bobdaduckHow would I register a specific players coolness factor into the array that is already storing player info? And how do I pull that back out?
09:31:12Watusimotohow to use them, or why to use them? (they're called tables in lua, btw)
09:31:27bobdaduckI understand what they're for
09:31:31Watusimotoyou want to assign a coolness factor to each player?
09:31:46bobdaduckSure
09:31:54Watusimotoyou could create a cf table indexed against player names
09:32:04Watusimotocf["watusimimot"] = 100
09:32:04bobdaduckcompletely lost.
09:32:16Watusimotocf["bobdaduck"] = ??
09:32:22Watusimoto(it's your script after all)
09:32:29Watusimotothen something like
09:32:35bobdaduckI'm not actually making this, I'm just trying to understand tables.
09:32:51Watusimotoplayer = "watusimoto"; print cf[player]
09:33:00Watusimotoshould print 100
09:33:11bobdaduckI already have an array of playerInfos, how do I add another... dimension, I guess? to the table.
09:35:18WatusimotoI'd suggest a second table with the same index as the playerInfos array
09:35:29Watusimotonot the cleanest, but probably the easiest
09:35:41Watusimotothen add and remove items from both tables at once
09:37:31bobdaduckand then accessing that I'd have to use ipairs... for i v ipairs(players) v[coolness] = random(10)
09:37:31bobdaduck?
09:37:58bobdaduckI SWEAR I WILL UNDERSTAND TABLES.
09:38:37WatusimotoI think so
09:38:45Watusimotohttp://www.phailed.me/2011/02/learn-lua-the-hard-way-tables/
09:41:38bobdaduck20 dollars says the jokes in that tutorial will help me understand way better
09:42:19WatusimotoI didn't do more than skim it, but I did like the title
09:42:34bobdaducklol
09:43:01WatusimotoI found one that didn't look as good, but you might like the picture
09:43:02Watusimotohttp://lua.gts-stolberg.de/en/Tables.php
09:45:14Watusimotook, I fixed it, sort of
09:45:36Watusimotoit looks like the problem happens when I create two serverGames
09:45:44raptorha
09:45:50Watusimotoeven when I create one, then delete it, then create a second
09:46:08Watusimotothis will be a problem with our testing, for sure
09:46:33raptorthat phailed.me site is good!
09:46:56WatusimotoI'm guessing there's a static that isn't being reset
09:47:08raptormakes sense
09:50:16Watusimotojust creating and deleting a servergame is enough to trigger the crash
09:50:39raptorhidden bug!
09:51:08bobdaduckYou found the hidden bug!
09:51:14bobdaduck*achievement unlocked*
09:53:18bobdaduckOkay so I still don't understand what everyone keeps doing in the code they make me with dictionaries and stuff
09:53:23bobdaduckBut that was a good tutorial.
09:54:23bobdaduckMight have gotten me far enough that analyzing the code will work now
09:56:40bobdaduckRequest for Lua API: onShipKilled(ship, killer)
09:57:08LordDVG has joined
09:57:35LordDVG Quit (Remote host closed the connection)
09:57:49Watusimotodon't we have that already?
09:57:53LordDVG has joined
09:58:07bobdaduckwe have onShipKilled(ship)
09:58:19bobdaduckWe don't detect who killed who though, and that's important.
09:58:30Watusimotook; killer seems appropriate
09:58:48Watusimotowould be nil if player not killed by another player, I suppose
09:59:01raptorcould be any BfObject?
09:59:59bobdaduckif killer.ObjType == asteroid then logprint("loser")
10:01:50raptorif ship == killer then logprint("haha raptor")
10:02:59bobdaduckrofl
10:04:32Watusimotoyes, good point
10:15:33Watusimotook, it's the deleting and recreating of the DataChunker that is the problem
10:15:53Watusimotothis does not happen in a normal game -- it is usually deleted on the way out of the game
10:18:26Watusimotoso I can either hunt down the problem, or just never delete it
10:18:51Watusimotoit's already static, but is reference counted and dleted when we delete the final database
10:19:05Watusimotowhich only happens when we delete both gservergame and gclientgame
10:19:13Watusimotowhich only happens when we're exiting
10:24:43bobdaduck"do it fast or do it right"?
10:26:32Watusimotoit's really a question of faith about whether this is a sign of a more serious bug
10:29:36raptordo it right
10:40:12Watusimotoof course :-)
10:40:27WatusimotoI figured out the allocator is being asked for 0 byte blocks
10:40:38Watusimotowhich is why it keeps handing out the same address over and over
10:41:08Watusimotobecause it hands out a block, then advances an internal pointer by the size of the block to get the addr of the next one
10:41:18Watusimotoso when size == 0... no advancement!
10:56:01bobdaduckWhat's kaen up to lately?
11:00:34raptortaking a break I suppose
11:23:59Watusimoto Quit (Ping timeout: 256 seconds)
11:35:58bobdaduck Quit (Remote host closed the connection)
11:54:30bobdaduck has joined
11:55:57raptor Quit ()
12:08:57traxxasme has joined
12:09:06traxxasmehey
12:09:30traxxasmeanyone online?
12:10:28traxxasmeno?
12:10:40traxxasmeEh, I'll just post to the forums :D
12:10:45traxxasme Quit (Client Quit)
12:20:50bobdaduckxD
12:21:36bobdaduckI was eating!
12:36:16raptor has joined
12:36:16ChanServ sets mode +o raptor
12:38:30raptorwe had a drive by!
12:39:39bobdaduckDrive by ircing
12:42:58raptorwhat makes music on your head?
12:43:07raptora head band!
12:43:24thread_more laffy taffy raptor?
12:43:31raptormaybe :)
12:46:38raptorsooo
12:46:49raptorno one has played on the sMegabyte server for at least a month
12:46:53raptorprobably never?
12:49:58bobdaduckIs that the weird locked one?
12:50:22raptoryes
12:50:23bobdaduckYeah whose even is that xD
12:51:21Watusimoto has joined
13:03:25WatusimotoI think I fixed it!
13:03:45raptorhello
13:04:05raptorthe TNL code that no one has ever looked at before?
13:04:39Watusimotoit's not a tnl problem
13:05:03Watusimoto if(mCountGridDatabase == 0)
13:05:03Watusimoto {
13:05:03Watusimoto delete mChunker;
13:05:03Watusimoto mChunker = NULL;
13:05:03Watusimoto }
13:05:09WatusimotoI added the = NULL line
13:05:21raptorthat's it?
13:05:39Watusimotoit was rusing the old chunker, which still apparently kind of worked, even though it was deleted
13:05:47WatusimotoI think so
13:05:55raptorhuh
13:06:20Watusimotowhy couldn't c++ just set a pointer to NULL after it deletes it? would that be too much to ask?
13:06:37Watusimotowould it be too hard?
13:11:40raptoryeah... just when I think I understand how c++ works..
13:12:34raptorhttp://stackoverflow.com/questions/704466/why-doesnt-delete-set-the-pointer-to-null
13:17:50raptorsome interesting responses here: http://stackoverflow.com/questions/1931126/is-it-good-practice-to-null-a-pointer-after-deleting-it
13:22:06Watusimotook, finally, all tests pass, no crashes
13:22:27raptoryay!
13:25:54WatusimotoI find most of those SO answers unconvincing
13:26:05Watusimotobut what the heck, it's not my language
13:26:13raptori feel the same
13:26:26WatusimotoI'll make my own that nulls after delete, called Chris++
13:26:44raptorbitfighter can be the premier platform that sells it!
13:27:52WatusimotoB++
13:28:03raptordoesn't that exist?
13:28:09WatusimotoB exists, I think
13:28:10WatusimotoB+++
13:28:34bobdaduckrofl
13:28:55raptorheh, the example here is funny: http://www.roesler-ac.de/wolfram/hello.htm
13:29:06raptorclick B
13:32:53Watusimotolovely!
13:39:01BFLogBot Commit: 4a4125256e29 | Author: watusimoto | Message: Rename IdlePath names to make them clearer, slight clarifications to logic, comments, other related items
13:39:02BFLogBot Commit: 749254225e5f | Author: watusimoto | Message: Merge
13:39:04BFLogBot Commit: 77315bb29b01 | Author: watusimoto | Message: Comments
13:39:07BFLogBot Commit: db07d2973b79 | Author: watusimoto | Message: Rename method
13:39:08BFLogBot Commit: 59717e6a2985 | Author: watusimoto | Message: Parallel variable rename
13:39:10BFLogBot Commit: 80b20ce122cf | Author: watusimoto | Message: Used testing to fix a loadout setting error! Also fixed problem with serial creation of multiple ServerGames. Not likely to happen in the game, but certain to occur during testing.
13:40:22raptordoes this mean you're ready to try CMake!
13:41:25Watusimotoin a minute... one more thing I want to fix before I forget. But you can send me the instructions to get started
13:41:37raptorok
13:41:40raptorinstructions:
13:41:45raptor1. cd build
13:42:29raptor0. install cmake: http://www.cmake.org/files/v2.8/cmake-2.8.10.2-win32-x86.exe
13:42:40raptor1. cd build
13:43:01raptor2. run: cmake -G "Visual Studio 10" ..
13:43:17raptor3. open the solution that it generated for you in the 'build' directory
13:43:21raptor4. compile!
13:47:05raptorstep #2 has two dots at the end
13:50:06bobdaduckstep #4 has an exclamation point on the end
13:51:29sam686 has joined
13:51:30ChanServ sets mode +v sam686
13:53:02raptorhi sam686
13:54:18raptorhey sam686, if you ever get the chance, could you test out cmake on windows?
13:54:45raptoryou install the latest cmake, then in the 'build' directory (at bitfighter's root) you run: cmake -G "Visual Studio 9 2008" ..
13:54:57raptorI think that's the right IDE you use...
13:55:06raptorthen open the solution that was created in the build directory
13:56:02sam686I have some kind of a new network setup now. My old pentium D computer is linux debian running it as a gateway NAT. (that computer 3 ethernet ports, 1 on-board, 2 is a card)
13:56:41sam686I still use my dd-wrt wrt54g as just a wireless switch access point though..
13:56:49bobdaduckI could maybe try it?
13:56:49raptorthat old pentium still runs, huh?
13:57:03sam686yes
13:57:08raptorbobdaduck: which visual studio do you use?
13:57:09Watusimotoyou can use it for heating in the cold south dakota winters!
13:57:45bobdaduck2010
13:57:51sam686except winter is over, it isn't cold anymore outside (in my location)
13:58:02raptorok, then the steps are the same as I posted for Watusimoto above
13:58:16bobdaduckWhen I get home we'll see
13:58:25bobdaduckI might get preoccupied hosting gladiator server
14:02:35Watusimotoso I get two projects: zero-check and all-build
14:02:47raptoropen the sln
14:02:56raptorbitfighter.sln should have been created
14:03:05Little_Apple has joined
14:03:12Watusimotothere it is
14:03:13Watusimotoyes
14:03:13Little_Applebonjour
14:03:17raptorhi Little_Apple
14:03:43Little_Applehello
14:03:48Little_Applejust stopping in to say hi
14:04:31raptorWatusimoto: also, if you add to the cmake lists, visual studio knows to run cmake and update itself
14:04:39raptorCMakeLists.txt
14:05:00Watusimotoit will do that automatically?
14:05:05raptoryep
14:05:14raptorat least with vc++ 2010
14:05:23raptorcmake adds in some sort of hook to check itself
14:06:04Little_Appleaaaaand im gone
14:06:07Little_Apple Quit (Client Quit)
14:07:39Watusimotono optimization in the build config?
14:08:02raptorit should add it by default if you select 'Release'
14:09:23raptoryou should never have to mess with the project settings..
14:09:26WatusimotoI usually use optimiazations during development... don't know if that's good or not, but it seems to most closely mirror the release code...
14:09:26raptor*should*
14:09:44Watusimotoespecially as we do almost no release testing
14:09:47raptormaybe I don't understand what you mean by optimizations
14:09:56raptorlike -O2 ?
14:10:04raptorduring development? that's crazy!
14:10:08Watusimotoyou know, funroll loops!
14:10:16raptorhahaha
14:10:17Watusimotowhy is that crazy?
14:10:27Watusimoto(maybe it is??)
14:10:33raptorbecause... debugging
14:10:47raptori mean, if you still compile with the symbols in
14:10:53Watusimotovc++ debugging seems to work fine
14:10:58Watusimotolet me check my other project
14:10:59raptoryou'll still have problems with line-numbers, etc...
14:11:15raptorat least that's my experience on linux + mac
14:11:22raptoralthough mac seems to get closer..
14:12:45Watusimotook, optimizations are off in my project as well.
14:12:49Watusimotoso cancel that
14:15:09thread_ Quit (Remote host closed the connection)
14:17:06raptorok good
14:17:20raptorthen that explains why debugging has also worked for you :)
14:18:37Watusimotoyes!
14:19:03WatusimotoI assume that some flags can be configured in the cmake project... for example, I prefer minimal rebuild, as compiling is slow on this machine
14:19:27Watusimotowhich conflicts with multi-processor compilation
14:19:41raptoryes... somehow...
14:19:46Watusimotook
14:19:48raptorlooking...
14:19:59Watusimotootherwise, the project looks pretty good
14:20:01raptoractually, you should just test that
14:20:14raptorlike change a lesser used header and see what happens
14:20:31raptorbut I know multiprocess build is enabled..
14:20:51Watusimotooh, like maybe I change it in vC++ and it gets written back to cmake somehow??
14:21:01raptorno no
14:21:09raptoradd some whitespace to a header file
14:21:13raptorand see how much it recompiles
14:21:20raptorlike change helperMenu.h
14:21:33Watusimotoah, ok
14:22:47Watusimotoon other (very) minor annoyance is that all the source files end up in a source files folder in the project, which indents them a quarter inch to the right, which reduces my code window soomewhat... this is, I assume just the way cmake structures things
14:22:58Watusimoto(and not an illogical way to do it)
14:23:04raptorheh
14:23:51Watusimotoit's nice that the precompiled files seem to be shared amongs the various projects
14:23:57Watusimotothat will speed things a bit
14:24:22Watusimotoodd
14:24:25Watusimotothere was a compile bug
14:24:33raptoroh really?
14:24:47raptori haven't tested the 'test' project..
14:24:50Watusimotowait, maybe not that odd
14:24:55Watusimotoit was a legit bug
14:25:03Watusimotobut it looked like some profiles built anyway
14:25:13Watusimotoonly one failed
14:25:53BFLogBot Commit: 7ceb78e99e85 | Author: watusimoto | Message: Add an assert to avoid an error I made today
14:25:55BFLogBot Commit: 19f6c631f750 | Author: watusimoto | Message: Need to return something here!
14:26:27Watusimotowell, I'll work with this for a while
14:26:31raptorolk
14:26:33raptorok
14:26:34raptorbecause
14:26:37Watusimotoit looks fine
14:26:56raptorI think it'll be easier writing a cmake module for a new dependency
14:27:15raptorthan it is to write a cmake module for a new dependency AND add it to a vc++ project
14:27:21raptorwith the latter taking 3 times as long
14:27:48Watusimotoyes
14:27:51Watusimotoagreed
14:28:12WatusimotoI may have been wrong about the shared compilation units
14:28:13sam686CMake Error at CMakeLists.txt:101 (message): You must run "cmake .." from the "build" directory (my full output: http://sam6.25u.com/upload/text1305/130509_15-05-37.txt )
14:28:23sam686don't know what I should do..
14:28:26raptorsam686:
14:28:29raptorcd build
14:28:30raptorthen
14:28:35raptorcmake -G "Visual Studio 9 2008" ..
14:28:41raptorwith those two dots
14:28:49raptoryou need to run it not in the project root
14:28:50Watusimotooh, we should have the master project in their too
14:28:56raptorah the master!
14:29:00raptori forgot about that...
14:29:07Watusimotomaybe not built by default, but good to have
14:29:10raptorno wait
14:29:12raptorit's there
14:29:27Watusimotowhere?
14:29:28raptoron linux at least you can do 'make master'
14:29:47Watusimotoyes, on linux you can
14:29:53sam686CMake Error: The source directory "C:/OTHER/Bitfighter/HG019/build" does not appear to contain CMakeLists.txt. when running the same cmake command inside a /build folder
14:29:56Watusimotobut if I work on the master, i need to be able to compile it
14:30:01raptorsam686: add the two dots
14:30:05raptorat the end of the command
14:30:10raptorcmake -G "Visual Studio 9 2008" ..
14:30:34raptorWatusimoto: did the master compile for windows before?
14:30:38Watusimotoyes
14:30:47raptorok, i'll work on that
14:30:53Watusimotoit's been a while since I actually did it though
14:31:09sam686there I see Build files have been written to: C:/OTHER/Bitfighter/HG019/build
14:31:15raptorsam686: yay!
14:31:16Watusimotoalso getting an error about missing httpRequest.h header
14:31:38raptorWatusimoto: not sure about that, is that the test project?
14:31:45Watusimotoyes
14:31:53Watusimotomaybe kaen forgot to check it in?
14:32:07Watusimotoit's in his test, which I've never built before
14:33:17Watusimotothe header is there.
14:33:53WatusimotoI need to include it as #include "../zap/HttpRequest.h"
14:34:51raptorah
14:34:58raptorso FYI
14:35:07sam686so where the master build? (I have my own master build on my own master/bitfighter_master.vcproj )
14:35:16raptorsam686: I need to add it
14:35:37raptorWatusimoto: the CMake layout has one master CMakeLists.txt in the root of our repository
14:35:49Watusimotook
14:35:50raptorthen it has another CMakeLists.txt in each of the compilable projects
14:35:54raptorso zap has one
14:35:57raptortnl has one
14:36:08raptoretc
14:36:13Watusimotoso to add a file to tnl, I edit the tnl list?
14:36:23raptoryes in tnl/CMakeLists.txt
14:36:43raptorlooks like kaen's test project was added to the zap one
14:36:57Watusimotook
14:37:05raptorit should be in its own folder
14:37:10raptor'bitfighter_test' ?
14:37:58WatusimotoI think this will work out fine
14:38:05raptorsam686: it looks like master/CMakeLists.txt exists, but a target isn't produced because it doesn't find mysql
14:38:14raptorhow do you guys find mysql on windows?
14:38:24raptori mean, how do you install it?
14:38:32Watusimotoso what is all_build and zero_check?
14:38:57raptorWatusimoto: I have no idea, actually, but I think they help with the cmake interface with vc++
14:39:04Watusimotook
14:39:41raptorwell, if you want to work on more tests, you can just use your old project until I fix them..
14:39:49bobdaduck!
14:40:24bobdaduckIs there a way to, using your sword script raptor, get who's sword you collided with when you get bursted?
14:41:53raptoruhh
14:42:25raptorcan you do zone:getTeam() ?
14:42:50raptorsorry, have to step out for a bit...
14:42:59bobdaduckWell yeah I think so but
14:43:03bobdaduckI need the exact player
14:43:32bobdaduckseeya
14:45:16bobdaduckI mean the zoneSword is registered to the player already, I think, but I think I remember somethinging about onShipEnteredZone(zone) not working properly
14:46:49sam686I will be back in about 1 hour..
14:47:07bobdaduckseeya
14:59:32LordDVG Quit (Remote host closed the connection)
15:03:26Watusimotoraptor: trying to replicate your concentric circles editor crash
15:03:40Watusimotomy machine starts going into overdrive and never comes back
15:03:50WatusimotoI think I am getting burns on my leg
15:06:44raptoroh, haha
15:06:48raptoryeah, that...
15:07:17raptorclipper + triangulation in a debug build is like 10x slower
15:09:05bobdaduckWhat if we made testitems perfecty round
15:09:42raptorthey are already!
15:09:49raptor:)
15:10:04raptorat least the server thinks they are circles
15:10:04bobdaduckIn more than just a technical sense, I mean.
15:10:06bobdaduckxD
15:39:06fordcars has joined
15:39:24fordcars Quit (Client Quit)
15:39:46fordcars has joined
15:44:24fordcarsbobdaduck: I am nearly finished your bot, I just was reeeeealy occupied this week.
15:47:42bobdaduckCool
15:47:45bobdaduckThat's fine
15:48:06fordcarsby the way, the bot might try to put the flag on it's sword
15:48:44fordcarsso I may need to code if goalZone id is not equal to 2...
15:48:53fordcarsgo to goal zone
15:48:54bobdaduckI changed swords
15:48:56bobdaduckto be loadoutzones
15:49:01fordcarsahhhh ok
15:49:20bobdaduckand for loadoutchanges make it find a neutral loadoutzone to make it flow best
15:49:32fordcarsyeah ok
15:49:51fordcarsbut do I make the bot change the shield module too?
15:50:02bobdaduckmake it switch all the modules
15:50:06fordcarsok
15:50:19bobdaduckYou can try to code intelligent module use if you want
15:50:27fordcarsyeah sure
15:50:43bobdaducklike, cloak when there's nobody nearby if it has cloak
15:50:58bobdaduckor shield if its close enough to a person
15:51:20fordcarswell I am using s_bot's shield :P
15:51:23bobdaduckOr repair neutral items if it has repair
15:56:34bobdaduck Quit (Remote host closed the connection)
16:06:50BFLogBot Commit: 35043e222cb1 | Author: watusimoto | Message: Suppress rebuilding zones when quitting testing from the editor
16:07:45raptorWatusimoto: that zone rebuild happens when leaving 'host game' too
16:07:51raptordoes your fix fix that?
16:07:58Watusimotono
16:10:11WatusimotoI'll see if there is a good way to do that; if not, it seems less serious
16:10:25Watusimotowhen editing, you go in and out of the game a lot
16:10:32Watusimotowith hosting, less so
16:10:46raptorok
16:11:05raptoractually - i was planning to work on it because I think it's my fault (i changed teh suspend code last)
16:12:48Watusimotothe fix might be easy
16:13:06Watusimotoif the game is not a dedicaetd server, and all players are gone... game must be ending, right?
16:13:35Watusimotothe only way a server can have 0 players and not be shutting down is if it is dedicated, right?
16:13:49raptorsounds right to me
16:14:04Watusimotoif so, the fix is simple
16:14:13Watusimototesting now
16:19:25Watusimotoah wrongish
16:19:34Watusimotocurrently even a local server can susend
16:19:37Watusimotosuspend
16:19:42Watusimotoif all players are inactive
16:20:02Watusimotomy fix makes that not happen
16:20:08Watusimotobut maybe that is ok?
16:25:18fordcarsbrb guys
16:25:20fordcars Quit (Quit: Page closed)
16:27:55raptorwhat kind of teeth can you buy with a dollar?
16:28:03raptorbuck teeth
16:43:29fordcars has joined
16:53:28Watusimotogood one, man!
16:54:14raptorwhy can't a bank keep a secret?
16:54:24fordcars Quit (Ping timeout: 245 seconds)
16:54:26raptorbecause there are too many tellers
16:55:12Watusimotoawesome!
16:55:36raptorthese jokes should be funny to you about now, right? is it late enough?
16:56:59Watusimotogive me 15 minutes
17:00:16Watusimotohey -- gaygayfaggot69 is playing!
17:00:31WatusimotoI love our community!
17:00:38raptoryay!
17:01:12raptorso a semi-troll was telling me how back I sucked at the game a couple days ago
17:01:38raptorand I responded by saying "tell me something I don't already know"
17:01:52raptorwhich seemed to quiet him/her...
17:02:05Watusimotogood job
17:05:09WatusimotoI'm posting a bug for you
17:05:14raptoroh goodie
17:06:25Watusimototriangulation fails with empty levels
17:06:34raptoroooo
17:06:36raptorreally??
17:06:40Watusimotoyup
17:06:52raptorwell, that's definitely my bug!
17:06:56Watusimoto:-)
17:07:36raptoralso, if you didn't see, I can not dupe this issue: http://code.google.com/p/bitfighter/issues/detail?id=155#c3
17:09:48Watusimotook; I'll try to verify
17:21:42raptorWatusimoto: here are the zones in 018a in an empty level: http://sam6.25u.com/upload/6screenshot_3.png
17:21:53raptorseem about right?
17:22:53BFLogBot Commit: 046175b495aa | Author: watusimoto | Message: Now also avoids zone building when quitting a locally hosted game
17:23:25WatusimotoI think so
17:28:44raptorok heading home!
17:33:29raptor Quit ()
17:40:24BFLogBot Commit: 0c7ea516a65e | Author: watusimoto | Message: Whitespace
17:40:26BFLogBot Commit: 69cffe78a489 | Author: watusimoto | Message: More ideas for help messages
18:21:55fordcars has joined
18:22:03kaen has joined
18:25:37raptor has joined
18:25:37ChanServ sets mode +o raptor
18:27:29raptorwhat did the judge say to the racket?
18:27:42raptor"we will send you to court"
18:30:41fordcarsthat's horrible
18:30:53fordcars:P
18:30:56raptori know! poor racket...
18:31:41fordcars:P
18:31:54kaenLol raptor your puns made the irclog more palatable
18:32:17raptorthank laffy taffy :)
18:32:22kaenmy internet's been down for a few days
18:32:28kaenso I wrote a half dozen editor plugins
18:32:33raptorOOoooo
18:32:38raptorwelcome back kaen!
18:32:42kaenthanks :)
18:32:51fordcarshaha it has been a long time since my last laffy taffy
18:33:03fordcarsreally? editor plug-ins nice!
18:33:14kaenbezier curve fitting, polygon edge subdivision, align, distribute, convert, and polyline modulation
18:33:22fordcarsooOOOooooo
18:33:45kaenI'm pretty proud :)
18:34:06raptorsweet!
18:34:07kaenI also found about a half dozen bugs and crashes when executing editor plugins...
18:34:21raptorall work with 019 API?
18:34:24kaenthere's definitely some memory corruption going on, as reported by glibc
18:34:25kaenyes
18:34:30raptorand yes
18:34:33raptorthe memory corruptions
18:35:05raptorif you look at the commit log from last night, I added loads of missing virtual destructors with the UI menus
18:35:05kaenvery excited to build the new tests, btw :)
18:35:12kaenah, perfect!
18:35:17raptor(hoping they'd fix some of the crashes)
18:35:19raptoroh
18:35:21kaenyou must have received my telepathic signals
18:35:24raptorheh
18:35:29raptoroh
18:35:34raptoruh Watusimoto is on cmake!
18:35:42kaenoh, I didn't see that!
18:35:44kaensweet
18:35:50raptorand... I had a question about the cmake stuff for the test project
18:36:05raptordid you want to put a separate CMakeLists.xtx in the bitfighter_test directory?
18:36:17raptorinstead of merging the logic with the zap/ one
18:36:25raptortxt not xtx
18:36:41bobdaduck has joined
18:36:49kaenthat sounds like a good idea, but I'm not sure how to share the file lists between the two
18:37:01kaenwhich is what prompted me to combine them in the first place
18:38:20Watusimoto Quit (Ping timeout: 255 seconds)
18:38:43raptorahh...
18:38:47raptormaybe..
18:38:48raptorhmm
18:38:55bobdaduckSo uh
18:39:09raptordo all the zap resources need to be recompiled with the test project? or can they just be linked against?
18:39:19bobdaduckI've been using pleiades to transfer levels to thread_ because I'm too lazy to zip files. I told him its a secret is that k?
18:39:53raptorit's like an open secret!
18:40:00raptorshame shame SHAME
18:40:21bobdaduckITS JUST SO CONVENIENT
18:43:21kaenawesome :)
18:43:37kaenI need a bigger test database for profiling, so use it a bunch, please
18:43:48bobdaduckIn other news, I've re-begun work on DnD
18:44:01kaenbobdaduck you should build 019 so you can look at my WIP
18:44:05raptoryou've decided to learn c++??
18:44:09bobdaduckTo be clear, nobody is going to make me a whip weapon, right?
18:44:19raptorsorry bobdaduck... I have moved on
18:44:21kaennot in my plans...
18:44:24raptorto.. dinner!
18:44:27bobdaduckk just making sure
18:44:28bobdaduckWIP?
18:44:31kaenI made you a bunch of editor plugins though!
18:44:34kaenwork in progress
18:44:35bobdaduckxD
18:44:42bobdaduckk
18:44:47bobdaduckjust as soon as I finish adding halberd
18:45:04kaenI need advice because I am teh noobzorz
18:47:00bobdaduckk
18:47:26fordcarswow
18:48:51kaenok just let me know when~
18:50:00bobdaduckwhat?
18:50:28kaener, when you're ready
18:50:29kaensilly
18:51:07bobdaduckno I mean at fordcars
18:51:29fordcarswassup
18:51:45fordcarssorry just reading your conversation
18:56:48kaenhrm... seems to crashing latest 019
18:57:29bobdaduckI'm building it
18:57:35bobdaduckgimme like a half hour or so
18:58:01kaenumm nevermind dude
18:58:13kaenit won't load until I fix the tree....
18:58:34bobdaduckcool
19:01:03bobdaduckback to DnD
19:05:17raptorhello
19:05:30raptormmm home-made whole-wheat biscuits!
19:05:39raptorlike rocks in gut!
19:05:43raptor*the gut
19:06:29SolumnMushroom has joined
19:06:46bobdaducksup mush
19:06:51SolumnMushroomHello all!
19:07:03bobdaduckokay guys
19:07:13bobdaduckwhich of the following loadouts best fits a warhammer, thematically:
19:07:17bobdaducksensor + armor
19:07:19bobdaduckshield + engineer
19:07:21bobdaduckboost + engineer
19:07:26raptorarmor + armor
19:07:27bobdaduckboost + repair
19:07:43bobdaduckor sensor + shield
19:07:44SolumnMushroomI'm navigating this on my Wii...
19:07:57bobdaduck(that's all)
19:09:07raptorthe one with armor
19:09:17raptoris my opinion
19:09:20SolumnMushroomN6t easy t6 d6 when y64r w6rds c60e 64t 352e th5s...
19:09:26raptori can say no more on the subject
19:09:57SolumnMushroomWith a keyboard
19:09:59kaenbobdaduck, I fixed the level
19:10:02kaenstill care to see?
19:10:10raptor(can I kaen?)
19:10:11bobdaducksure
19:10:15kaensure thing :)
19:10:24kaenI think it only works in 019
19:10:29raptorok
19:10:32kaenalthough I may have just fixed that
19:10:37bobdaduckI need to update right?
19:10:58kaenuh, let's just go to kaen test and try it
19:11:02kaenhosting segfaults for me in 01
19:11:04kaen019
19:11:26bobdaduckso 018a then?
19:11:50bobdaduck Quit (Remote host closed the connection)
19:11:51raptorpoor poor 019
19:12:02kaen018a indeed
19:13:51SolumnMushroomIn the one with the numbers, replace 6 with o and 0 with m...
19:14:47SolumnMushroom352e = like
19:15:42SolumnMushroom now hates num lock
19:19:49SolumnMushroom Quit (Ping timeout: 245 seconds)
19:21:05SolumnMushroom has joined
19:21:40SolumnMushroomI'm back and on Android this time!
19:23:37SolumnMushroomI'm installing Mac OS on my HP Mini
19:23:51raptordon't do it!
19:24:00raptorHackintoshes are crashintoshes!
19:24:28raptorunless if it's for research, then go ahead :)
19:27:53kaenSolumnMushroom, tried debian stable?
19:28:43SolumnMushroomNo, I don't have an intel Mac, So I'm making one. WITH STYLE!!!
19:31:23SolumnMushroomI have Winderp, and Linux on one computer. Mac OS will complete it.
19:31:32raptorbug slain!
19:31:56raptori said... bug slain!
19:32:15SolumnMushroomThat, and I overwrote my Mac computer with Linux...
19:32:33raptor!bot
19:32:33BFLogBotI'm a real boy!
19:32:47SolumnMushroomLolwut
19:34:49SolumnMushroomCan you commit for the good of the project? :-)
19:35:56-moorcock.freenode.net- *** Looking up your hostname...
19:35:56-moorcock.freenode.net- *** Checking Ident
19:35:56-moorcock.freenode.net- *** Couldn't look up your hostname
19:36:05-moorcock.freenode.net- *** No Ident response
19:36:12BFLogBot has joined
19:36:12Topic is 'Bitfighter 018a released! | http://www.bitfighter.org/downloads | Join us in the forums! at http://bitfighter.org/forums/'
19:36:13Set by watusimoto!~eykamp@SU115.tudor.lu on Wed Mar 06 15:04:30 UTC 2013
19:36:13-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.
19:36:53SolumnMushroomWelcome back!
19:36:54kaensometimes when I look at `top` on bitfighter.org, the java process has 1G of virtual memory in use
19:37:14raptori know
19:37:21raptori need to rewrite the logbot in python
19:37:28kaenI can help!
19:37:29raptorso it'll use 768MB of virtual memory!
19:37:33kaenheh
19:37:53raptorit's been a longtime goal of mine
19:38:44raptorand to base it off of something like this: https://github.com/xrogaan/apib
19:38:50raptorit's the bot on #naev
19:39:02raptorbut i'd need to add a socket listener, etc..
19:39:30SolumnMushroomSo, how 'bout Cards Against Humanity?
20:00:23BFLogBot Commit: 41b127e2dcc9 | Author: buckyballreaction | Message: Fix crash on building bot zones for an empty level
20:09:48raptorthat bot looks like it uses a really good irc python lib..
20:28:41kaenregarding your question earlier, I believe the main objects could be linked to by the test project (in theory) if they use exactly the same defines and compiler flags
20:28:51kaenagain, the problem is that I don't know how to tell cmake to do that
20:29:00raptorhmmm
20:29:20raptorok, i'm trying to set up the 'master' project in windows - i'll take a look at that afterwards
20:31:04kaenmaybe we could make a libbitfighter target, have the main bitfighter target only contain main.cpp and then the test and main projects could have the lib as a dependency
20:32:14raptorthat's what I had to do with android
20:36:31kaenweird
20:36:35SolumnMushroomI feel like an idiot. I dechecked "Enable Wireless" on my Ubuntu partition, and now wireless on the entire computer is nonfunctional... :-P
20:37:11kaenhuh. most wifi cards aren't that finicky
20:37:17SolumnMushroom slaps his face with the force of 10000000 suns
20:37:30raptorunless they somehow adjust a BIOS setting
20:37:39kaendoes your laptop have a fn + F8 combo?
20:37:43kaento toggle your card?
20:38:12SolumnMushroomThe wireless toggle is f11
20:39:17SolumnMushroomTo know what computer I'm talking about, look up HP Mini 110-3130nr
20:48:47fordcarslater guys
20:48:51fordcars Quit (Quit: Page closed)
20:49:52SolumnMushroomI think I just got it to work again.
20:50:19SolumnMushroomYES!
20:50:34SolumnMushroomWith that,
20:50:48BFLogBot Commit: f7edb61df47f | Author: buckyballreaction | Message: master target now compiles in Visual Studio. Not sure if the MySQL variant works yet, but this is a first step
20:51:12SolumnMushroom Quit (Quit: AndroIRC - Android IRC Client ( http://www.androirc.com ))
20:52:48raptordid he just have a one way conversation with himself for almost 2 hours?
20:52:53SolumnMushroom has joined
20:53:58SolumnMushroomOn wireless WinXP again!!!
20:54:53SolumnMushroomI had to reenable the Broadcom driver in Linux...
21:09:09SolumnMushroomhttp://pyz.socialgamer.net/game.jsp
21:09:32SolumnMushroomGo here now
21:09:47raptorsorry, busy coding..
21:11:08SolumnMushroomkaen up for a round of Cards Against Humanity?
22:58:49BFLogBot Commit: d624a7824031 | Author: buckyballreaction | Message: Fix missing ARRAYSIZE definition when building release build on non-windows platforms
22:58:51BFLogBot Commit: 9e05bac6156f | Author: buckyballreaction | Message: Organize 'test' target in CMake a little more. Also fix compiling main_test on Linux
23:04:17SolumnMushroom Quit (Quit: Leaving)
23:46:07BFLogBot Commit: ef05fd193414 | Author: buckyballreaction | Message: Organize CMakeLists.txt for test target a little better
23:48:43raptornight
23:48:46raptor Quit ()

Index Search ←Prev date Next date→

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