#bitfighter IRC Log

Index Search ←Prev date Next date→

IRC Log for 2013-11-29

Timestamps are in GMT/BST.

00:25:18Skybax_ has joined
00:27:43Skybax Quit (Ping timeout: 260 seconds)
00:37:26Darrel Quit (Quit: IRC client killed.)
00:49:24Skybax has joined
00:51:58Skybax_ Quit (Ping timeout: 240 seconds)
01:30:45BFLogBot Commit: 12488260fc0d | Author: sam8641 | Message: Better prediction of client side energy meter
01:30:47BFLogBot Commit: 6bf4c5e8b408 | Author: sam8641 | Message: Remove no longer needed RPC c2sDeploySpybug, Spybug can still be deployed.
01:45:00phillreeder Quit (Ping timeout: 245 seconds)
01:50:29Skybax_ has joined
01:52:37Skybax Quit (Ping timeout: 248 seconds)
02:08:08Skybax has joined
02:11:01Skybax_ Quit (Ping timeout: 272 seconds)
03:32:45raptor has joined
03:32:45ChanServ sets mode +o
03:37:17raptoron to fixing stuff!
03:43:56BFLogBot Commit: 62f6d8859069 | Author: buckyballreaction | Message: Fix returning bad index with lua_getWeapon
04:00:27BFLogBot Commit: 70d96c61b011 | Author: buckyballreaction | Message: Fix projectileshooter bot
04:07:45raptorkaen: sam686: are tests failing for you?
04:10:00sam686test passes for me, oh i haven't pulled raptor's changes after mine..
04:10:17raptorlet me make clean
04:10:56sam686still passes for me after the "Fix projectileshooter bot"
04:11:55raptorlet me run again in gdb... it's throwing an assert
04:12:08raptorafter recompile
04:12:37sam686Something new I see on the test, I see multiple "debugger.lua loaded for Lua 5.1"
04:13:18raptorfailed BfTest.LevelReadingAndItemPropagation
04:15:39sam686mine just say [ OK ] BfTest.LevelReadingAndItemPropagation (95 ms)
04:19:47raptorhttp://pastie.org/pastes/8515994/text
04:21:07raptorlooks like float rounding problems
04:21:15raptorbut why is it so much bigger on my system?
04:22:18sam686there don't appear to be no compression of position on Teleporter, like there is on MoveObject, so that probably isn't a problem i don't think
04:23:52sam686teleporter->getOutline().x don't match teleporter->getOrigin().x?
04:24:20sam686oh i mean there something like Point centroid = findCentroid(*teleporter->getOutline());
04:24:45sam686more like findCentroid(*teleporter->getOutline()).x don't match teleporter->getOrigin().x ?
04:25:03raptoryes
04:25:29raptorcentroid calculation does do float division
04:26:54sam686maybe you could play around with Teleporter::generateOutlinePoints() in teleporter.cpp
04:27:33Platskies has joined
04:28:00raptorgood idea, i'll optimize it away from division
04:29:57raptorhmm... nope, that didn't work
04:32:15Platskies Quit (Ping timeout: 272 seconds)
04:33:19Quartzy has joined
04:33:59QuartzyAlright so I'm looking for opinions.
04:34:02sam686how about trying something like this? http://sam6.25u.com/upload/text1311/131129_04-33-39.txt
04:34:07QuartzySkybax asked a good question: "So, when it comes to voting, are we supposed to vote based off which one we hate the most, or which one is the most playable despite being idiotic?"
04:34:37QuartzyI'm wondering what sort of solution people are thinking. I sort of think the serious maps that trickled their way into the contest should be disqualified from voting, but that's me.
04:35:08QuartzyAnd then of course, one would instinctively vote for Skybax' latter idea, "the one that is most playable despite being idiotic."
04:35:30raptorin the past, when i've done contests, i always did all the maps because I couldn't decide
04:35:42QuartzyFair enough.
04:35:45Quartzyraptor have you seen the maps?
04:36:15raptoryes
04:36:31QuartzySo you personally wouldn't disqualify any of them from voting?
04:36:45QuartzyBecause if I don't, there is indeed a problem that Skybax was wise to point out.
04:36:58sam686I would vote on what is good, mostly if it is playable, like in a past.
04:37:08Quartzyyes of course.
04:37:25QuartzyProblem is, if we do that as it stands? Let's be honest, Retention Loss wins.
04:37:27raptori would say that despite the theme, people should not vote the ones they hate the most
04:37:32QuartzyAnd Retention Loss didn't even ATTEMPT to stick to the theme.
04:37:35sam686maybe don't name the contest "stupid"? though its a bit too late to change title of contest..
04:37:36QuartzyI love it, but I'm just being honest.
04:37:58QuartzyWow sam, that's so helpful. /massive_sarcasm
04:38:41sam686one of level contest map is mostly blank, only a few line item and testitem, but no walls.
04:38:56QuartzyYes I'm with you raptor. But *again* ... if that is the case, then everyone will just vote for the maps that are indeed excellent, but didn't even try to fit the theme.
04:39:07QuartzyThere's no point in even doing themes if people are going to break them and then reap the benefits.
04:39:08raptorhmmm... hard one
04:40:26raptormaybe theme-breakers should probably be allowed to vote first round, but not make it to the second round (and make sure that's explicit in the voting post)
04:40:33raptoralso contact the map makers and tell them
04:40:34sam686parhaps you could have just set some easier simple rules..
04:40:57raptorthere are 2 maps that break the theme, i think - wasn't there a bitmatch with nice lineart...
04:41:17QuartzyYes. Line Match and Retention Loss.
04:41:39QuartzyLine Match is a fairly average, solid Bitmatch, while Retention Loss is borderline brilliant.
04:42:04QuartzyYou may be right, I should probably just take it up with the authors.
04:42:25raptormaybe let the map makers try to convince you that they're 'stupid', if they fail, let them be voted on in the first round only - that way the authors still get some recognition
04:43:02QuartzyI like it, except only being allowed to be voted in the first round is a little ... weeeird.
04:43:09QuartzyBut yes, I should allow them to defend themselves.
04:43:42raptorit would be an experiment, i suppose - or don't allow voting (i've done neither option before so I don't know how it'll work out)
04:44:32sam686I don't know if the rule "only one map per person in voting list" still applies, the way _k did it.
04:44:56raptorsam686: we stopped doing that when _k left - it allowed more submissions
04:45:01sam686ok
04:45:09raptormore, higher quality ones, actually (I think)
04:45:58sam686I only made one map to go into this level contest though..
04:46:16QuartzyYes well. 21 maps 9 authors.
04:46:18sam686though probably shouldn't tell which one yet..
04:46:19QuartzyIt'd be a bit cruel.
04:46:47Quartzybtw sam if you want to host the next contest feel free.
04:46:47BFLogBot Commit: f7d1b87ce12d | Author: buckyballreaction | Message: Minor performance improvement in findCentroid()
04:47:11QuartzyAnd I don't mean that in a snobby way
04:47:23QuartzyYou have ideas floating around and it'd be interesting to see it run by someone other than raptor or I.
04:48:00raptorsam686: did you just basically use pythagorean triangles for the outline points in that test one you gave me?
04:48:19QuartzyOr _k. I should give him credit even if he was an elitist dick.
04:48:25sam686ok, probably after the current level contest..
04:48:45sam686no just come out with some numbers..
04:49:26sam686its a good test if the problem is really the generateOutlinePoints
04:50:09sam686except it probably be too tiny of outline
04:50:30QuartzyFudge it. I'm probably worrying too much. We'll just see how the voting goes I guess.
04:50:31QuartzyThanks.
04:50:38raptorlooks like it passes with those sam686
04:50:48raptorthank you Quartzy, for hosting this!
04:51:04QuartzyYou're welcome and thanks for hosting the server itself.
04:53:20Quartzy Quit (Quit: Page closed)
04:53:43sam686ok, here is with the TELEPORTER_RADIUS http://sam6.25u.com/upload/text1311/131129_04-53-21.txt
04:54:02sam686oops, multiple + + +
04:54:58sam686http://sam6.25u.com/upload/text1311/131129_04-54-39.txt (fixed multiple + + +(
04:56:09raptorthat one failed - 1275.0006 to 1275
04:56:19raptorinteresting...
04:57:40sam686Back to original code, Did it in a slightly different direction http://sam6.25u.com/upload/text1311/131129_04-57-12.txt
04:58:52raptorfails with original error
04:59:02raptori'm not too worried, honestly
04:59:22raptorit's within 0.002 of a Point
04:59:24sam686maybe the problem is inaccurate F32 calculation..
05:00:28sam686I think only the outline is used only for... maybe the help screen outline, and maybe the editor teleporter outline.
05:00:41raptoryeah
05:00:46raptorlet's ignore it, i think
05:03:26sam686printf("%f %f \n", centroid.x, teleporter->getOrigin().x); gives me 1274.999634 1275.000000
05:03:36sam686yet it passes the EXPECT_FLOAT_EQ ?
05:03:55raptornow that's odd
05:04:12raptorseems like a weirdness in GTEST, then
05:08:15Platskies has joined
05:09:12sam686umm what linux command do I use to build a test?
05:10:06raptormake test
05:10:11sam686my first try: cd bitfighter_test; make "No makefile found"
05:10:23raptoroh, in the build dir
05:10:35sam686ok
05:10:40raptorit's part of cmake, too
05:13:03Platskies Quit (Ping timeout: 252 seconds)
05:15:37sam686well I got some errors http://sam6.25u.com/upload/text1311/131129_05-15-19.txt
05:16:18raptordid you have code changes?
05:20:15sam686it doesn't say nothing on the code changes
05:20:51sam686hg commit "nothing changed"
05:22:34sam686trying make clean then make test
05:23:06raptorwhat system are you using?
05:23:26sam686its a debian 64 bit, running as my router too..
05:24:06sam686multiple ethernet cards in the machine, all working and bridged except "eth0" for the internet
05:24:28sam686same build error
05:24:33sam686on make test
05:24:48raptorneat setujp
05:24:51raptor*setup
05:24:55sam686main_test.cpp:(.text+0x2957): undefined reference to `Zap::Teleporter::getOrigin() const'
05:25:06sam686yes
05:25:18raptorwell that's odd - are we mixing 'const' somewhere?
05:25:29sam686who needs extremely slow linksys router?
05:26:04sam686and my linksys wireless was dying too. I instead use hardware wireless case as hostapd "wlan0" access point now
05:27:09sam686this is my linux hardware http://sam6.25u.com/upload/text1311/131129_05-26-47.txt
05:28:10sam686I still use my linksys as a 5-port switch as well... I have too many ethernet ports it seems..
05:28:40raptorneat!
05:28:52raptorI might have to do something like that soon - our linksys is dying as well, i think
05:30:18sam686well only my wireless part of my linksys was dying, so I shut off that wireless because of my other wireless (mini pci-express to pci-express adapter + some old wireless G card that works with hostapd)
05:33:20sam686If I want, I could even remove my graphics card and linux will just boot just fine as a server/router.
05:33:54sam686with of course a few one-time beeps of missing graphics card..
05:34:36raptorhaha
05:36:36Platskies has joined
05:36:42sam686teleporter.cpp line 704, perhaps removing "inline" fixes link errors..
05:43:09sam686Well At lease I got the test compiled...
05:43:16sam686... with this: [ FAILED ] BfTest.LevelReadingAndItemPropagation
05:43:20BFLogBot Commit: d79977231f84 | Author: sam8641 | Message: Fix linux linker error
05:43:29raptorha!
05:53:18BFLogBot Commit: b456cbaf5afc | Author: buckyballreaction | Message: Don't set bot default loadout in initialize(), it will overwrite anything you set it its main() function
05:54:42kumul Quit (Quit: Leaving)
05:56:53Platskies Quit (Ping timeout: 248 seconds)
05:56:59raptorok, going to bed
05:57:06raptorgood night!
05:58:51raptor Quit ()
06:02:43Nothing_MuchAnybody here?
06:06:37Platskies has joined
06:06:42sam686hi
06:26:00Nothing_Muchwhat's goin' on?
06:28:39sam686im mostly just doing C++ coding
06:30:12Nothing_Muchoh cool, where?
06:30:58Platskies Quit (Ping timeout: 265 seconds)
06:31:25BFLogBot Commit: beabd5e07842 | Author: sam8641 | Message: Fix failing test on linux.
06:33:57Nothing_Muchnice
06:34:01sam686I may be going to bed soon, so probably no more of me C++ bitfighter tonight..
06:35:40Nothing_MuchHave a g'night then sir
06:35:55sam686see you later..
06:53:15Platskies has joined
07:18:41Kiryx has joined
07:21:15Flynnn has joined
07:39:49Platskies Quit (Ping timeout: 252 seconds)
07:45:17Darriel has joined
07:45:53Darriel is now known as Darrel
07:55:11Platskies has joined
08:20:49Platskies Quit (Quit: Gotta go)
08:21:07Platskies has joined
08:26:25Platskies Quit (Quit: Gotta go)
08:30:32destroyerimo has joined
08:36:26Invisible has joined
08:47:59watusimoto has joined
08:47:59ChanServ sets mode +o
09:27:32Platskies has joined
09:28:39Kiryx Quit (Ping timeout: 272 seconds)
09:57:09Invisible Quit (Ping timeout: 272 seconds)
10:33:05kumul has joined
11:15:27Kiryx has joined
11:33:49LordDVG has joined
11:48:41Canseco has joined
12:00:23Platskies Quit (Read error: Connection reset by peer)
12:10:38Invisible has joined
12:23:24Flynnn Quit (Quit: This computer has gone to sleep)
12:31:29LordDVG Quit (Remote host closed the connection)
12:37:45Canseco Quit (Remote host closed the connection)
12:45:08Canseco has joined
12:49:00watusimoto Quit (Ping timeout: 252 seconds)
12:51:03Invisible Quit (Ping timeout: 272 seconds)
12:53:28watusimoto has joined
12:53:28ChanServ sets mode +o
12:56:06Canseco Quit (Remote host closed the connection)
13:05:50Canseco has joined
13:23:21kumul Quit (Ping timeout: 272 seconds)
13:24:38kumul has joined
13:29:58phillreeder has joined
13:34:31phillreeder Quit (Ping timeout: 260 seconds)
13:46:08Canseco Quit (Quit: http://quassel-irc.org - Bye)
13:46:46Canseco has joined
13:48:47Canseco Quit (Remote host closed the connection)
13:50:29Canseco has joined
13:55:03Darrel Quit (Ping timeout: 260 seconds)
14:02:01Invisible has joined
14:04:34Darrel has joined
14:43:26alex_velea has joined
14:44:08alex_veleaHello everyone :)
14:44:53alex_veleaI have a little problem ...
14:45:29alex_veleaI've created my own class in a new file with .h and .cpp form like in the other files .. in the .h only the definitions and i've used ifndef
14:46:08alex_veleaI get an error on building :( .. after it says linking CXX executable .. bla bla
14:46:45alex_veleain function .. bla bla .. undefined reference to my constructor .. even if i defined it
14:50:20watusimotois your cpp file included in the project?
14:51:06watusimotothe way to get it into the project (properly) is to edit cmakesources.txt (or something similar) in the zap folder
14:51:34watusimotoand then change to the build folder and run cmake ..
14:52:17alex_veleaOhh
14:52:21alex_veleaI did not do that
14:52:38alex_veleaI just saw that if i put my class in anothe .h class it works
14:52:40alex_velea:D
14:52:41alex_veleaThanks
14:56:34Invisible1 has joined
14:58:06Invisible Quit (Ping timeout: 245 seconds)
15:11:51alex_veleaWhat do i need to include to use your types? like U32 F32 ..
15:17:20alex_veleanvm .. solved it
15:29:05alex_veleaCan i write a function in GridDatabase to return the private vector<DatabaseObject *> mAllObjects?
15:34:55LordDVG has joined
15:42:03Invisible1 Quit (Ping timeout: 272 seconds)
15:49:43raptor has joined
15:49:43ChanServ sets mode +o
15:49:51raptorTODAY IS RELEASE DAY
15:50:01raptornow's your last chance to get bugs fixed!
15:50:40raptorunless, of course, someone objects to it being release day...
15:51:31raptorand good morning!
15:52:08Invisible has joined
15:58:58alex_velea Quit (Quit: Page closed)
16:05:31raptorsam686 or kaen, LogNetBase is on for dedicated build, and I can't find how to turn it off. Would you know how?
16:09:22kumul Quit (Quit: Leaving)
16:10:27Invisible Quit (Ping timeout: 260 seconds)
16:12:14Invisible has joined
16:22:27raptori mean, the extra TNL network logging
16:22:38raptorlogs every single packet
16:27:55Canseco Quit (Remote host closed the connection)
16:32:53kaenOHBOYOHBOYOHBOY
16:33:03kaenI'll take a look in just a second
16:33:18kaenI object to today being release day!
16:33:26raptoroh?
16:33:29kaenon the grounds that release day should have been months ago :)
16:33:34raptorhaha
16:33:43kaenand good morning
16:34:27raptorhello!
16:34:35raptori'm verifying osx still compiles...
16:36:08kaenI'm getting a ton of these: warning: right operand of comma operator has no effect
16:36:23kaenon the dedicated build
16:36:44raptoryes, that happens because we have weird typedef replacements with ZAP_DEDICATED, i think
16:36:58kaenohh that's right
16:37:05kaenit's mostly render type functions
16:37:19raptorI haven't been too concerned about it because we don't ship bitfighterd
16:37:39kaenah, interesting
16:37:39raptorbut it sure indicates we have a messy project...
16:38:00kaenI'd like to try some architectural things for 020 ...
16:41:40raptorI still haven't wrapped my head around *proper* client-server architecture - it seems we mix stuff waaaay too much
16:44:01kaenI think we just lack clear separation between the shared stuff and the client/server specific stuff
16:44:39kaenA lot of it has to do with TNL
16:45:10kaenTNL creates its own remote class for us, whereas other network engines let you define a shared base class, and a derived server and client class
16:45:30kaenor alternatively, use the server as the base class, and subclassing it to make the client version
16:45:46kaenwhich gives you only two classes per object type and mixes in CSP for free
16:46:16kaenthe hard part about those architectures is that you need virtual methods for the divergent behaviors
16:46:26kaenand to wire those into the shared behaviors as needed
16:47:30kaenessentially what have right now is *as* complicated as a having two different classes for each object, but we don't even get the organizational benefit of actually splitting the divergent code away from the shared code
17:04:17raptorsigh, yes
17:04:26Invisible Quit (Ping timeout: 264 seconds)
17:04:31raptori think i'd like to see a good example of organized client-server code...
17:04:36raptorif it exists
17:06:19watusimotoraptor: I've been thinking about the notifier... I think we should include it with the game, and just plunk it into the executable folder
17:06:32raptorwatusimoto: you sure¿
17:07:05raptorwatusimoto: I can probably do that, but the exe would be 3-4MB larger...
17:07:13raptorthe installer exe, i mean
17:11:12kaenI think we should, too
17:11:36watusimotoI think so. First off, I think many players will want it; second, a few more MB won't matter; and third, it needs to be in the same folder, and better just to put it where it needs to be than try to explain to people how to manually install it
17:11:37kaenit seems like it would do much more than 3-4MB worth of good
17:12:10watusimotounless we have an automatic installer that figures out which path it goes in, but still who wants to do that?
17:13:39watusimotoActually, I need to check the windows installer -- I think we should make a shortcut to it on the start menu
17:14:10raptorI gotta stop thinking I only have a 50MB hard drive with windows 3.11
17:14:34watusimotoyes -- drives are up to several hundred mb these days
17:14:46raptorwatusimoto: one of the GCI tasks made it intelligently look for the exe in various program files folders
17:14:46watusimotoand vista works great!
17:15:03watusimotoI didn't see the details of what he did
17:15:14raptori did, it looks good
17:15:24watusimotobut I know he made it better than only looking in the local folder
17:15:28raptori'll compile that in, but having the exe in the same folder will be fine anyways
17:15:34watusimotoindeed
17:16:04watusimotoI also think we should include at least the s2c change that passes music from server to client, even if we don't expose the functionality until 019a
17:19:07watusimotoso my two tasks for release are 1) making sure the windows installer handles the notifier properly; and 2) changing the s2c as mentioned above
17:19:16watusimotoI can get those done for tonight
17:21:05watusimotobtw, on another topic, splitting each class into a c and an s will be more confusing that what we have, I think
17:21:43watusimotoin any case TNL expects classes to be the same on the client and server
17:21:51watusimotoso unless we hack that, we are stuck
17:22:19watusimotowhat we can do is split out (some) functionality with specially named functions, as I have done in some places
17:22:25watusimotolike idle_server and idle_client
17:24:01watusimotoactually, I am not sure I stand behind my statment above about confusion
17:24:54watusimotobut the inheritance chain will be a mess for things like repairItems
17:25:47watusimotorepairItem_server would inherit from repairItem or pickupItem_server or pickup_item or...???
17:26:26raptorwatusimoto: i'm not sure I like the idea of the s2c change
17:26:34watusimotoI think it's safe
17:26:45raptorhmmm...
17:26:47watusimotoit's just adding a string param
17:26:58watusimotowe don't need to do anything with it on either end
17:27:00raptorok, well... make sure the signature is 'future-proof'
17:27:25watusimoto019a proof, or soem future we have not yet contemplatd proof?
17:27:53raptor019x proof i suppose - I've recently been thinking that every s2c should just implement a bstream
17:28:20raptorthat way we always have the same sig and just need to make sure it's mirror from client-server
17:28:21watusimototell you what -- I'll make the change I think we need; it will be simple, and I'll check it in early (i.e. I'll do that task first) and they we can discuss
17:28:28raptorok
17:28:36watusimotoand revert if that's what we want
17:28:42raptori'll... go seal another window...
17:28:47watusimotohave fun!
17:28:55raptoroh hey, osx release bundle still copmiles
17:28:58raptorthat's good news
17:33:36myxabrik1 has joined
17:34:27watusimoto Quit (Ping timeout: 252 seconds)
17:47:21phillreeder has joined
17:53:15kumul has joined
17:54:54myxabrik1Hi everyone! I'm new Code-In participant. I'm trying to build Bitfighter, but after running make I can't find any executable.
17:56:14kaenhi myxabrik1 :)
17:56:17kaenwhat OS are you running?
17:56:31kaenthe exe should be in bitfighter/exe
17:56:43myxabrik1I'm running Ubuntu Linux
17:56:45kaenok
17:56:57kaendid you use cmake first?
17:57:12kaen!gci
17:57:12BFLogBotGCI Students Welcome! Please see: http://bitfighter.org/wiki/index.php/GCI_Student_Bootstrap for getting started.
17:58:19Watusimoto has joined
17:59:17myxabrik1Yep. Though, targets like 'dedicated' and 'bitfighterd' are absent.
18:00:49kaenthe targets are absent? or were not built?
18:01:13myxabrik1Ou, I've found some errors in dependencies.
18:01:18kaenah ok
18:01:44kaensudo apt-get install libsdl1.2-dev zlib1g-dev libpng12-dev libopenal-dev libvorbis-dev libspeex-dev libmodplug-dev cmake build-essential
18:02:16kaenand then you'll probably have to run cmake .. again
18:12:14phillreederWatusimoto: we need to sort out some stuff with the seeker targeting
18:12:31Watusimotook\
18:12:35raptorhi phillreeder, did you see my response on the GCI task?
18:12:40phillreederoh yea, saw that
18:12:50phillreederalso, gonna need more time with this
18:12:59raptorso loadout bugs were fixed... and thanks for finding them :)
18:13:27phillreederawesome! I just worked around it by using Timer to fire the loadout change after 0ms
18:14:00destroyerimo_ has joined
18:14:36destroyerimo Quit (Ping timeout: 252 seconds)
18:15:36phillreederso Watusimoto: let's lay out how we want to get this done. Right now I have Robot::getSeekerTarget() and Robot::lua_getSeekerTarget(). I tried to make the getSeekerTarget() outside Robot, but I figured that would be too much of a pain and just settled for putting it in Robot for now.
18:15:37raptorheh, ok
18:15:54Watusimotoraptor: my promised update is in
18:16:06Watusimotothe "contraversial" one
18:16:12BFLogBot Commit: b4fb37586bcb | Author: watusimoto | Message: Add dummy string parameter in preparation for having level-sepcified music in 019a, changed nearby parameter from STE to string to save STE slots.
18:17:11Watusimotophillreeder: ok for now; I think it will make sense to mix it into getFiringSolution when we get things working
18:17:38Watusimotothen that could pick a method based on either selected weapon, or a passed parameter, whichever makes the most sense
18:18:39phillreederthat's exactly what I had in mind, thought it might be better to keep the logic seperated for now
18:18:44phillreederone more thing
18:18:50Watusimotofor the moment, if that makes things easier, yes
18:20:17Quartzy has joined
18:20:38QuartzyStupid Contest voting is up! http://www.bitfighter.org/forums/viewtopic.php?f=13&t=2152
18:20:59phillreederdo you want lua_getSeekerTarget() to return a target angle or ship?
18:22:30phillreederas in, should calling it give the .bot a target to cache and fire at or the angle at which to fire at the optimal target
18:23:02Watusimotowell...
18:23:32WatusimotoI think the way it works with bots now is that they seek their target and getfiringsolution returns a shooting angle, or a nil if there are no good ones
18:24:04Watusimotoso in terms of picking a target, I thik we either create a new function called lua_getSeekerTarget() or implement in lua. A new fn is my choice
18:24:26Watusimotoso... a target, I think
18:24:37phillreedermakes sense
18:26:22Quartzykaen +sam686 @raptor LordDVG Skybax Stupid Contest voting is up! Stupid Contest voting is up! http://www.bitfighter.org/forums/viewtopic.php?f=13&t=2152
18:26:31Quartzy Quit (Quit: Page closed)
18:31:50kaenQuartzy is probably going to have Bad Day�
18:36:01Watusimotook, good, the relocation of bf.org didn;t change our ip address
18:37:39kumul Quit (Read error: Connection reset by peer)
18:42:01raptorhi again
18:42:47raptorI don't even have a clue what I shoudl do to add the notifier for linux/osx
18:42:50raptoresp. osx
18:44:09Watusimotohow will we distrubute it for osx? as a .py?
18:44:40raptoryeah, but because of the .app bundling, there will be no way to launch it other than having the user go and find it themselves
18:44:51raptori mean, osx/linux will just ue the .py
18:44:54raptor*use
18:45:11Watusimotoso we can just stick it in the install folder, and let people find it
18:45:28WatusimotoI mean we can advertise it and all, but there's not really a good alternative
18:46:27raptorthe alternative is to create a new bundle just for it (on osx) but there's some weird stuff with bundling python things and I don't have this knowledge (or am particularly motivated to figure it out because... osx)
18:46:29Watusimotothough osx must keep a list of executables
18:46:58raptorosx keeps a list of .app
18:47:01raptor.apps
18:47:12Watusimotowell, we can upgrade our distribution over time; we might be able to get some help from one of our mac players
18:48:52Watusimotowe could (maybe) use distutils for linux
18:50:28raptori am unfamiliar with that...
18:51:02Watusimotohttp://svn.pythonmac.org/py2app/py2app/trunk/doc/index.html
18:51:15Watusimoto(this is something different, for osx)
18:51:28LordDVG Quit (Remote host closed the connection)
18:51:36raptori spent a couple of hours trying to get that to work once
18:51:42raptorpy2app
18:51:47raptorI left angry
18:52:00raptorbut that was a year ago..
18:52:39Watusimototrying to see if it has been updated at all
18:54:16phillreederhey quick question. I'm splitting lua_getSeekerTarget and getSeekerTarget, should Robot::getSeekerTarget return a BfObject or take a reference to one in the args and set that
18:55:11kaenit'll have to return an object
18:55:27kaenoh, pardon me
18:55:31kaenyou mean in c++ world
18:55:36phillreederhehe yes
18:55:51kaenI'd say just return a pointer
18:56:07phillreederI'll go with that
18:57:14Kiryx Quit (Ping timeout: 265 seconds)
18:58:13Skybax Quit (Ping timeout: 248 seconds)
18:58:39Watusimotonow reading about this
18:58:40Watusimotohttp://www.pyinstaller.org/
19:01:32Watusimotothe project still active, with commits on the 27th
19:08:38sam686raptor, don't know how to turn off network logging? Check bitfighter.ini [Diagnostics] LogNetBase=No
19:10:39kaenit's not respecting the INI option
19:10:51kaenI tried to trace it in GDB but got bogged down
19:11:34kaennow I'm fixing my last few release-critical bugs in pleiades :P
19:15:00LordDVG has joined
19:15:05LordDVG Quit (Changing host)
19:15:05LordDVG has joined
19:15:27kumul has joined
19:19:30Watusimotopyinstaller seems to work on windows
19:19:51Watusimotoneed to work a bit on dinner, will continue experimenting after dinner
19:22:24LordDVG Quit (Remote host closed the connection)
19:24:32Watusimoto Quit (Ping timeout: 260 seconds)
19:26:51raptorsam686: the INI isn't working for the extra logging - I had assumed you or kaen had added some define or other flag in one of your TNL commits, but I was unable to find it
19:27:17sam686it works, but only changes whats logged for bitfighter.log
19:28:13raptorah
19:28:22raptorso then the change was for stdout
19:28:26raptorsomewhere
19:28:56sam686see setupLogging, I don't see any such network logging there
19:29:07sam686in main.cpp
19:29:30sam686theres 2 setup logging functions with different parameters
19:31:31sam686what kind of output you getting? A bunch of TNL logprintf's?
19:36:35raptorsam686: http://pastie.org/pastes/8517548/text
19:36:45raptorthat's being printed on the console
19:37:55kaensame here
19:42:30sam686is that as a dedicated build or just a regular full client/server build?
19:43:40kaenonly for dedicated
19:45:58Kiryx has joined
19:46:53sam686i see the output logging problem too on dedicated, need to find where it is coming from though..
19:49:50kaenanything I can do to help the packaging, raptor ?
19:50:59raptorkaen: do you think we shoudl distribute the notifier in the linux packages somehow?
19:52:00kaenyes
19:52:14raptorinstalled in /usr/bin or /usr/share/bitfighter... ?
19:52:23kaenyou should just be able to add it as a +x file (with hashband) in /usr/bin
19:52:33kaenhashbang*
19:52:41raptorinstalling in a bin dir is fraught with packaging convention problems
19:52:42kaen#!/usr/bin/env python
19:52:50kaenreally?
19:53:00kaenthey only let you specify one exe?
19:53:27raptorno, you can have multiple, but then they go through extra LINT if it's like python or perl
19:53:50raptorwell, i'm not at that point anyways..
19:55:11raptormy 5 month old just ate pumpkin pie and is now beserk
19:57:24kaenI would go beserk if I had to eat pumpkin pie, too
19:57:39kaen:)
19:57:53kaenI'm more of a cheesecake guy myself
19:59:18raptorthe wife gave her more because she's 'happy crazy' not 'grumpy crazy'
20:01:12BFLogBot Commit: fe9baa7566cc | Author: sam8641 | Message: dedicated build: It will help if oglConsole doesn't print EVERYTHING to console
20:02:16raptorwas that really the issue?
20:03:43BFLogBot Commit: 818930a680b0 | Author: sam8641 | Message: Oops, still want to output to gConsole, but not printf (stdout)
20:03:44sam686made a little error on my commit
20:04:18sam686yes that was a problem, a PRINTF in oglConsole
20:04:40sam686even better, no OglConsoleLogConsumer gOglConsoleLog; in dedicated build
20:07:29raptorgood find!
20:07:32raptorthanks
20:20:56phillreederraptor the seeker should only target enemy ships correct?
20:22:51myxabrik1 Quit (Remote host closed the connection)
20:27:37raptorphillreeder: yes
20:28:20raptorsorry, crises at the moment...
20:52:27Watusimoto has joined
20:59:28Watusimotoraptor: did you merge the gci notifier code into the repo?
20:59:44raptorWatusimoto: not yet, am planning on it in a few min
20:59:49Watusimotook, great
21:00:00WatusimotoI like this new python build tool
21:00:05Watusimotopyinstaller
21:00:15Watusimotoworks great on windows, is supposed to be cross platform
21:00:17raptoryou mean the one i'm already using?
21:00:24Watusimotois that the one you are using?
21:00:36Watusimotoif so... then yes!
21:00:36raptoryes, i have a config file and everything set up for it
21:00:42Watusimotooh, great
21:00:48Watusimotois that in the repo?
21:00:52raptorso don't do too much extra work!
21:00:55raptoryes
21:01:00Watusimotojust playing with it so far
21:01:19raptorthere's a subdir with notes in the config file
21:01:39destroyerimo_ Quit (Ping timeout: 260 seconds)
21:01:43Watusimotosee it
21:03:36WatusimotoI came up with a few different options than what you have in your .sh script
21:04:16raptordifferent than ones also specified in the .spec file?
21:04:35raptorI'm using mingw, by the way, so i don't rage against the machine as much...
21:04:48raptorerr MSYS, I mean
21:05:09WatusimotoI need to review the .spec
21:05:40Watusimotowe need the --noconsole option for windows
21:05:49raptorthat's in the spec
21:05:55Watusimotooh, good
21:06:11Watusimotook, you're ahead of me on this
21:06:26raptoryeah, I can compile it and give you the EXE
21:06:38Watusimotono need -- I have the tools here to build it myself
21:10:20raptorok, let merge those changes...
21:10:36sam686do you want the console to be hidden in -dedicated ? (windows)
21:11:48sam686main.cpp have USE_HIDING_CONSOLE part for windows-only, it may show console at first, but once SDL windows shows the console is hidden (release build)
21:12:09sam686and the console won't hide if ran from command window
21:13:53Watusimotoah, I think we want to show the console, so people can kill their server ???
21:13:58phillreederI'm having a hard time thinking of what the seeker targeting should be looking for
21:14:01Invisible has joined
21:14:29Watusimotophillreeder: one idea is to have it look for ships within LOS
21:14:29raptorother ships!
21:14:31phillreederit seems like no matter what the situation, the bot should target the bot with the least advantage and the work around it
21:14:50sam686showing console, its an easy way to prevent stupid people from double-clicking the -dedicated shortcut and accidenly run hundreds of dedicated server
21:14:58kaenspicey here
21:15:09Watusimotohi spicey
21:15:19kaenHey, how is it going?
21:15:39Watusimotosam686: ideally we would show a tray icon in windows for the master, but I know we're a long way from that ideal
21:15:49Watusimotoit's going well here!
21:15:58Watusimotowhat have you done with kaen?
21:16:38Watusimotophillreeder: the seekers kind of putt along until the find something
21:16:40raptorphillreeder: you probably don't have to choose a target in such a complicated way - maybe just the closes enemy ship with line-of-sight
21:16:53raptor*closest
21:16:54Watusimotoso to use them most effectively, you want them to lock on quickly so they accelerate
21:17:10phillreederthe algo looks for targets within range, then prioritizes based on energy, distance, line of sight.
21:17:23Watusimotothey can see around corners a little, but maybe you shoud just focus on los
21:17:36raptorimmutable changeset, what the heck
21:18:07WatusimotoI think you should worry less about energy, and just pick the most optimal target from a target POv
21:18:38phillreederok, so from I'm getting here, there's almost no point to having it in the robot.cpp
21:18:39Watusimotoand if the target is at the extreme end of range, maybe approach instead of shoot?
21:18:57Watusimotothe target selection?
21:19:07phillreederyea, there's very little outside range and LOS
21:19:18Watusimototarget selection could probably be in the bot, yes
21:19:46Watusimotoas for where to aim, you may need to experiment a bit -- leading, trailing, or aiming directly at the target
21:20:09phillreederI know we want to adopt it to the turrets and whatever, but it seems really specific to bots at this point
21:20:12Watusimotowith phasers, we lead the target by just teh right amount...
21:20:17kaentwo dimensional differential equation!
21:20:27Watusimotosounds like spicey is gone
21:20:55kaenhaha yeah she hijacks my laptop when I'm cooking for her
21:21:04Watusimotophillreeder: if you find someting simple, that works as makes for a fun bot, then that's great
21:21:09Watusimotono need to complicate things
21:21:21raptorspikey has made a recent contribution to bitfighter in fact - kaen may now become our android maintainer :)
21:21:25phillreederok one last thing before I set off: this is practically a whole different playstyle for the bot, how is s_bot going to choose between the playstyles
21:21:45kaenhahaha that's true!
21:22:24WatusimotoI think there are a couple of options: either the bot can make a global decision (always be the seeker bot), or it can make a tactical decision (too far for phasers, use seekers)
21:22:51Watusimotothe assignment was ambiguous on that point; what do you think would be most fun to play against?
21:22:59Watusimotoprobably a bot that mixed it up a bit
21:23:11phillreederI'll play with the second a little and see if it works well, hopefully I can get it to fel right
21:23:13Watusimotofired a seeker then swooped in with phasers or soemthing
21:23:15phillreederthat was my initial impression
21:23:26Watusimotoexperimentation is the best way
21:24:26Watusimotoraptor: do you think we should use upx on windows for the notifier?
21:25:14WatusimotoI'm thinking that it woudl complicate the build process, and that we'll get compression in transit from the installer; the only difference would be installed size
21:25:28raptorWatusimoto: already do
21:25:41WatusimotoI don;t see a windows build script... is there one?
21:25:47WatusimotoI see the .sh one
21:25:51raptorthat's it
21:25:54raptorfor use with msys
21:25:59Watusimotoah, I see
21:26:18Watusimotothat will require a very specialized build environment then
21:26:27raptorWatusimoto: I think I should be building on windowx xp 32bit
21:26:52raptorI don't trust your vs 2013 to build something fully backwards compatible
21:27:26Watusimotowindows is always backwards compatible; I do trust it, but am not opposed to you doing the final build
21:27:37sam686I have visual 2008 that runs on my windows XP and 7
21:27:51sam686I kindof have visual 2010 on windows 7 though
21:27:58raptorthe reason I say this is that the backwards compatibility does not hold when compiling in newer IDEs...
21:28:02Watusimotobut I do want to make sure that the build environment is as easy to get up and running as possible
21:28:17raptorthey link in different APIs, even if it is the 'same' library
21:28:25raptorit's why bitfighter doesn't run on windows 2000
21:28:54sam686unless compiled from visual 2008, those are said to suppose to run windows 2000 i think
21:29:09sam686in fact I use 2008 right now
21:29:27raptorexactly! binaries I compile on vs 2010 do not work on windows 2000, and I have not been able to get them to work..
21:29:28Watusimotopersonally, I only feel compelled to suport >= XP
21:29:36raptoryeah, me too...
21:29:37sam686but again anyone still use 2000?
21:30:01Watusimotoprobably not
21:30:45WatusimotoI don't like having to have msys installed to build the notifier
21:30:59Invisible Quit (Quit: Invisible)
21:31:10raptorthen create a batch rile
21:31:12raptorbatch file
21:31:25raptorthe sh is super simple and minor modifications should work in a .bat
21:31:46Watusimotothe batch file is no problem, excpet for the path to upx
21:31:54Watusimotowhich is where we started this conversation
21:31:54sam686interesting,i have some msys-like stuff in mingw. C:\MinGW\msys\1.0\bin
21:32:06raptoruhh
21:32:13Watusimotoif we didn;t use upx, it would be super simple
21:32:13raptorupx was just a zip file with an exe in it
21:32:26raptori put it somewhere, then pointed to the folder
21:32:55Watusimotoright, but the .sh file has a hardcoded path in it, which seems... hardcoded :-)
21:32:57raptorthat seems simple to me... unless you mean you don't want to have to download it
21:33:26raptorbuilding this thing will require hard-coded stuff
21:33:33Watusimotono
21:33:36raptorif used in dos or msys
21:33:40Watusimotowithout upx it wouldn't
21:33:43raptoryes... look at the spec file
21:34:21Watusimotoit has one in there:
21:34:32Watusimotoroot_path = 'c:/hg/bitfighter.tools/bitfighter-notifier'
21:34:43raptoryep
21:34:56raptorthat was the minumum I could get away with...
21:35:22Watusimotowhat is that? I can;t find that in the pyinstaller docs
21:35:30raptorbut if you want to take out upx, then that's fine - i'm sure the installer compression would make up for it
21:35:43raptorit's not in the docs
21:35:49Watusimotowhat is it?
21:36:02raptora variable to make running the spec easier
21:36:07raptorthe spec is just python
21:36:36Watusimotoah, so it is
21:36:36raptorit uses that path to know where to analyze the python script and to pull some other resources, like the ico file
21:36:47Watusimotocouldn't those all be relative paths?
21:37:55raptori remember trying that but failed - I don't remember why, but I still think it *should* work
21:38:02sam686notifier? it won't run for me, and this is all I get: http://sam6.25u.com/upload/text1311/131129_21-37-33.txt
21:38:18raptorI think it had to do with the spec running from the pyinstaller path
21:38:34raptorand i couldn't guarantee a relative path from that on everyone's system
21:38:45sam686where should i "cd" to then?
21:38:51raptorsam686: it'll work after its compiled
21:39:01raptorsee the dependencies in the .py script itself
21:39:11Watusimototo run from the .py you need the libs
21:39:26Watusimotothe exe is an all-in-one; no python required
21:39:37sam686perhaps its the systram thats missing
21:39:41sam686systray
21:40:42Watusimotobecause when I build the exe manually, I did so from the bf-tools folder
21:41:18Watusimotoand since everything is there, I didn;t require any paths at all
21:41:30raptoroh really - but did you use the spec?
21:41:34Watusimotono
21:41:36WatusimotoI did this
21:41:41Watusimotopyinstaller --onefile --noconsole bitfighter_notifier.py
21:41:48Watusimotoit's not complete, but it worked
21:41:58Watusimotoeven sucked in the icons when run from a folder with icons in it
21:42:13Watusimotoso maybe it is complete
21:42:31WatusimotoI wasn't even planning on using a .spec file
21:42:39Watusimoto(unless we needed it for other platforms)
21:42:40raptorhow big was the output?
21:43:35Watusimotoas I recall about 4.2MB
21:43:49raptorok good, then it did pull in the runtime
21:43:52Watusimoto(I think I deleted it on a subsequent aborted run)
21:44:06Watusimotohow big is your build?
21:44:07raptormy output (with upx) was 3.1 MB or so
21:44:10Watusimotook
21:44:20Watusimotoso I think upx gives us nothing of real value
21:44:44Watusimotothe installer will compress it just as well, and 1MB on disk is nothing
21:45:21sam686bitfighter.exe release build for me is 5099 KB
21:45:40Watusimotoso my plan was to add the pyinstaller command to the nsis builder, and have it all sucked in with one command
21:45:43sam686soo.. it looks like python is size-inefficient somehow..
21:46:01Watusimotowell, you get all of python with that exe
21:46:53Watusimotofor my plan to work, however, the notifier should be in the same repo as the game
21:47:02Watusimotoso we know where everything is
21:48:28sam686maybe if we cook up some windows C++ tray icon notifyer, we can then have it display while running -dedicated and notify on player join..
21:49:11Watusimotowould that offer a big advantage over running the exisitng notifier? (which will work in that context as well as others)
21:49:46BFLogBot Commit: 965b490df992 | Author: buckyballreaction | Message: Add in GEOFBOT's notifier changes from GCI
21:50:15raptorI don't mind bringing the notifier into the main repo, now that it will be a part of this complete breakfast
21:50:26sam686Could also be tiny bandwidth, using TNL to communicate (as a seperate MASTER_NOTIFIER NetConnection) and only the master instantly tell the notifier
21:50:28raptorbut maybe put it under its own directory
21:51:37raptorthat would be ideal, but it does not offer rapid development on 3 platforms
21:52:45sam686what else should we do? Convert entire TNL to python?
21:53:25raptoror just leave the simple http JSON fetch
21:54:15raptordoing a c++ one would offer speed/size advantages, but the disadvantage of having to do a lot more platform-specific development
21:54:26kaenalso a JSON parser
21:54:39raptorkaen: he's thinking of just using TNL to ping the master
21:54:46kaenohh I see
21:55:03raptoralso, development on OSX, to me, means death
21:55:08sam686more like connect to master.. and just wait for the master to signal notifier
21:55:23sam686thats the most instant way to get notified
21:55:29kaenok, but there's the UI integration
21:55:35kaenwhich is 100% the hardest part in my opinion
21:55:44Watusimotoone possibility would be to use a standard libary like Qt to do notifier stuff
21:56:00Watusimoto(if it is offered)
21:56:05kaenimo Qt is a nightmare to link against
21:56:08Watusimotothat might remove some of the ui pain
21:56:09raptorQT!
21:56:13raptor12 MB later...
21:56:25Watusimotoit's an option; there are others
21:56:43raptornot on release day!
21:57:06Watusimotoso I'm goint to move forward with moving notifier into the main repo and hooking the build into nsis
21:57:14kaenjust thinking about all the for a c++ notifier like the python is making me nervous
21:57:19Watusimotoand documenting the new requirement (pyinstaller)
21:57:28kaenall the deps*
21:57:41sam686the notifier stuff can always wait for later, like 019a..
21:57:50raptorWatusimoto: don't hook up the building with nsis, but hook up the resultant exe
21:57:56raptoralso please pull latest from repo
21:58:07raptori just commited geofbot's changes
21:58:17Watusimotook, then build from a bat file and package with nsis?
21:58:24raptorWatusimoto: yes
21:58:26WatusimotoI could live with that
21:58:36raptorthat way we don't break nsis if we don't have all the deps
21:58:41Watusimotoyes
21:58:42Watusimotook
22:01:49sam686CS_PROTOCOL_VERSION 37 might want to change to 38 on release... many people have old 019 that can cause some problem if attempted to connect to released 019..
22:02:19raptoryes, i'm waiting for everyone else's commits to change the version numbers
22:02:59raptori'm doing my commit review now...
22:03:29raptorkaen, do you need to re-run the luadocs?
22:04:30Watusimotoannoyingly, hg is giving me a 403 Forbidden error when I try to commit to the tools repo
22:04:32raptorit's been almost 2000 revisions...
22:04:57raptorWatusimoto: set up your google code credentials?
22:05:12sam686CS_PROTOCOL_VERSION 38 on release.. the comments in version.h, just have it say pre_019 = 37 // 019 = 38
22:05:15Watusimotonot with this repo yet
22:05:23Watusimotobut I don't get a chance to enter my pw
22:05:27Watusimotolet me try from cmd line
22:06:08Watusimotook, it's not a tortoise thing, it's an hg thing
22:06:55raptorI.. I don't think I can enumerate the bug fixes...
22:07:48sam686hg push https_path?
22:08:01raptorah yes, https is needed too
22:08:44Watusimotoc:\Users\Chris\Documents\bf-tools>hg push
22:08:45Watusimotopushing to https://watusimoto%40bitfighter.org@code.google.com/p/bitfighter.tools/
22:08:45Watusimotosearching for changes
22:08:45Watusimotoabort: HTTP Error 403: Forbidden
22:08:47sam686or in turtoiseHG, press the 2 green arrows and then set the path, click the lock icon to change password, and click save (floppy disk icon)
22:09:03WatusimotoI think that hg must have an incorrectly cached password somewhere
22:09:31raptorthe password is the auto-generated google code one
22:09:34sam686I wonder why people still use the floppy disk icon for save button?
22:10:04Watusimotoah, good idea
22:10:10WatusimotoI got a pw prompt
22:10:14Watusimotohooray!
22:12:03Watusimotopushed!
22:12:04BFLogBot Commit: d7fe2d1c7ce7 | Author: watusimoto | Message: Comments
22:12:05BFLogBot Commit: 12484e1bac08 | Author: watusimoto | Message: Merge
22:12:07BFLogBot Commit: d24cf4e17310 | Author: watusimoto | Message: Merge
22:12:08BFLogBot Commit: 05f3b2973831 | Author: watusimoto | Message: Merge
22:12:39raptor1 commit and 3 merges?
22:13:36raptoramazing... fun graph
22:22:37WatusimotoI've been dragging those comments around for a while
22:23:44sam686I don't get it, did the latest comment of 3 merge disappeared from google code?
22:23:44Watusimotoso... top level folder called "notifier"?
22:24:05sam686http://code.google.com/p/bitfighter/source/list
22:24:07Watusimotoit should not have
22:24:36Watusimotowrong repo?
22:24:43sam686I polled your changes once it worked.
22:24:57Watusimotoit's the bf-tools rep
22:25:01sam686I tried polling again into my different clone, I don't see merge 3 time changes
22:25:19sam686oh a different repo
22:25:21sam686I get it.
22:25:37sam686Confused over which Repo I am looking at
22:30:39sam686I did notice something, I see raptor's gmail address and watusimoto email address in commit authors (in hg web) but not mine. http://sam6.25u.com/hg/bf019/ and logbot doesn't show gmail part in raptor's commit
22:33:33kaeniirc you have to configure mercurial for that
22:33:56kaen[ui]
22:33:56kaenusername = kaen
22:33:56kaenemail = bkconrad@gmail.com
22:34:03sam686ok, though not too bothered about that..
22:35:35sam686the commit username can be different from push username/password
22:35:54sam686and it appears it just uses email for the commit username if the commit username left blank.
22:36:21sam686no email/no commit username, it tells you need a username.
22:37:15phillreederalright Watusimoto I think I got somewhere
22:37:42phillreederthe bots feel significantly more....dynamic
22:37:52Watusimotogood, I think
22:38:18kaencan I see?
22:38:24phillreederit feels a lot less clustered
22:38:57phillreederhere's the code for the s_bot http://pastie.org/8517799
22:40:20sam686Authors:? now that you made changes to it...
22:41:16phillreederI don't feel like I did enough to write my name on it yet
22:41:28sam686its ok
22:49:26phillreederI decided against putting new logic in the robot.cpp, but getIdealTarget() could be migrated and improved upon
22:50:09kaenI like the idea of prototyping in lua, and moving to c++ if there's a significant performance gain to be had.
22:53:18raptorI agree
22:53:56raptorWatusimoto: whatever you don't get done tonight, I'll continue
22:54:44Watusimotook, thanks
22:54:53raptorI haven't been doing too much - we've had several crises today
22:55:04Watusimotothe danger of putting too much bot logic in c++ is bots become homogeneous
22:55:17raptorthe last was my second-born wedged himself between boards of the fence, with his arms forward
22:55:30Watusimotogood grief
22:55:42raptorand when I pulled hiim out, he lost a healthy chunk of skin
22:57:04Watusimotoouch
22:59:42kaenWatusimoto, agreed
22:59:54kaenplus, why maintain c++ code that you don't have to?
23:00:04Watusimotoc++ is good
23:00:08Watusimotovery tasty
23:00:20Watusimotoso you don't haev to maintain lua code!
23:00:24kaenhehe
23:01:05kaenI think I've just been brain damaged from writing so much dynamic interpreted code
23:01:18kaenI feel out of place when I don't have anonymous function literals.
23:02:17kaenthe only thing I can say that I like about c++ is the static typing
23:03:14phillreederagreed. Watusimoto I want you to check over what I have to see if you like where I'm headed, should I put it as finished on the task page?
23:03:44Watusimotowhat did you do? c++ and lua work?
23:04:18phillreederI was playing with c++ but decided it would be better to prototype in lua and maybe later move some logic to c++
23:04:28phillreederso for not it's all in the .bot
23:05:07Watusimotothe just do a pastie for the moment, or the gci thing... whatever works
23:05:44phillreederhttp://pastie.org/8517830
23:06:14phillreederthat's that...no where near done but it'll give you a sense of what I want to do
23:06:31Watusimotoso just use it like any other bot?
23:06:52phillreederyep, it's a rewrite of s_bot. I got to go for the night, see y'all tomorrow
23:07:27Watusimotolater!
23:07:36kaenlater!
23:08:10kaenso there's nothing I can do to help?
23:08:50Watusimotocheck the bot!
23:08:59Watusimotofor me he seems to put on armor then do nothing
23:09:20Watusimotooh, maybe I need to rebuild
23:12:44WatusimotoI think phillreeder is going to need to walk me through this tomorrow
23:13:45Watusimotowell, I cannot get pyinstaller to build in the new location
23:13:56Watusimotoan error message pops up then disappears too quickly to read
23:15:02kaenhmm I couldn't actually get the bot to fire
23:17:11Watusimotome neither
23:18:30Watusimotomaybe that's what he meant when he said nowhere near done
23:25:57raptorok, are there any bugs we want to finish before release, like that editor one? or allowing neutral Cores?
23:28:10kaenneither of those interest me
23:28:31raptoryeah, me neither, unless someone can solve that bug with a magic want
23:28:33raptor*wand
23:28:44kaenit's already bested me twice
23:37:27raptorme too!
23:39:39amgine123 has joined
23:39:39Watusimotowhat's the editor error?
23:39:44amgine123hey raqptor
23:39:51amgine123hey wattisimo
23:40:06amgine123is raptor afk i need to ask him osmthing
23:41:02kaenmost likely. just ask, amgine123
23:41:24kaenWatusimoto, the editor error is a weird snapping thing with forcefield and turrets
23:41:33amgine123raptor can you type in the channel /msg servbot #bitfighter fantasy on ?
23:41:34WatusimotoI can fix that :-)
23:41:41amgine123i found that btw
23:42:32amgine123it gives you more channel abilitsy if you type that
23:42:39amgine123but only @ can do it
23:45:18sam686there is CoreItem::canBeHostile() CoreItem::canBeNeutral(), they are both false, allowing neutral core on editor is a simple as making those return true.
23:46:56raptoryes, but we chose to disallow them
23:47:09raptorit was for more than just a numbering problem - but i don't remember why
23:47:41sam686a lot of neutral core problem was back in 017
23:47:46sam686but those problem are gone on 018
23:47:52amgine123ratpro look at above chat
23:48:39sam686as of now, CoreGame only count how many core you have
23:49:11sam686on 017, it counts -1 for every core that isn't yours and a whole lot of confusion related to that method..
23:49:25BFLogBot Commit: e4e34b4a1971 | Author: watusimoto | Message: Import notifier code from bf-tools repo
23:49:27BFLogBot Commit: d8ba89132cf4 | Author: watusimoto | Message: Remove paths from spec file
23:49:28BFLogBot Commit: 6479a1d76325 | Author: watusimoto | Message: Merge
23:49:30BFLogBot Commit: d7322066d1a1 | Author: watusimoto | Message: Create a bat file to help build the notifier on windows
23:50:02Watusimotokaen: are you looking for something to do?
23:50:12kaenyes
23:50:46amgine123!help
23:50:46BFLogBotType !commands to see a list of commands. Please don't spam the main channel
23:51:35amgine123ratypro try the command trust me its worth it
23:52:01BFLogBot Commit: cfd5755e2385 | Author: watusimoto | Message: Remove notifier -- code is now in main repo
23:52:18Watusimotowell, I have about an hour before I completely pass out and need to decide between two tasks
23:52:26Watusimoto1) try to fix that editor error
23:52:41Watusimoto2) get the installer to include the notifer on windows
23:52:58raptorI can do 2, if needed
23:53:00WatusimotoI can definitely do 2), possibly 1), definitely not both
23:53:21Watusimotoit should be failry easy
23:53:55Watusimotook, I'll take a stab at the editor task
23:54:31amgine123but i had a possibly stupid idea about the join polywalls creating a hole problem
23:56:43kaenI'll work on 1 as well
23:56:46sam686Why do we allow hostile spawn point in editor if they do totally nothing? (i think, or is it something with levelgen?)
23:56:54Watusimotook, I updated the checklist to mention building notifier on windows
23:57:33amgine123is it possible with the join walls issue to make it so if there is a hole to automaticly fill the entire hole ( 1 object) then deleted the filled area ?
23:57:49amgine123fill join then deleted ?
23:57:53raptorkaen: did you say luadocs was up to date?
23:58:05kaenyes
23:58:13Watusimotofor nsis, we'll need to 1) copy bitfighter_notifier.exe from pyinstaller/dist and 2) add a shortcut to it on the start menu
23:58:32amgine123ratpro you should acess fantasy commands is would be very helpfull for you
23:59:09Watusimotoman, that checklist is getting big
23:59:58Watusimotoadded Desura

Index Search ←Prev date Next date→

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