#bitfighter IRC Log

Index Search ←Prev date Next date→

IRC Log for 2013-10-14

Timestamps are in GMT/BST.

00:00:29Flynnn has joined
00:40:43Nothing_Much Quit (Remote host closed the connection)
01:03:31Nothing_Much has joined
01:17:31koda has joined
01:37:02Flynnn Quit (Quit: This computer has gone to sleep)
01:45:36watusimoto has joined
01:45:36ChanServ sets mode +o
02:10:26Invisible1 has joined
02:17:27tjcarter Quit (Read error: Connection reset by peer)
02:17:45tjcarter has joined
03:56:37Invisible1 Quit (Ping timeout: 265 seconds)
04:34:20Invisible1 has joined
05:08:43LordDVG has joined
05:45:19YoshiSmb has joined
05:52:20Invisible1 Quit (Ping timeout: 272 seconds)
05:53:20YoshiSmb Quit (Ping timeout: 250 seconds)
06:29:11LordDVG Quit (Remote host closed the connection)
06:30:30HylianSavior has joined
06:47:08Invisible1 has joined
07:07:18kodab has joined
07:11:04koda Quit (Ping timeout: 246 seconds)
08:03:27Invisible1 Quit (Ping timeout: 240 seconds)
08:09:27tjcarter Quit (Read error: Connection reset by peer)
08:09:29Invisible1 has joined
08:09:41tjcarter has joined
08:16:17Invisible1 Quit (Ping timeout: 252 seconds)
08:26:44kodab Quit (Ping timeout: 252 seconds)
08:35:39bobdaduck has joined
08:35:49bobdaduckwatusimoto:
08:35:53bobdaduckIs bitfighter capitalized?
08:36:19bobdaduckI always deliberately lower case my name as style and I'm not sure...
08:39:13Invisible1 has joined
08:59:57HylianSavior Quit (Quit: Leaving)
09:23:49watusimotobobdaduck: generally, yes
09:23:54bobdaduckokay.
09:24:01bobdaduckAnd can you tell me how to read a stack trace?
09:24:29watusimotoBitfighter
09:24:43watusimotoah... start at the top (or bottom, depending on nature of stacktrace)
09:25:36bobdaduckhttp://pastie.org/8401566
09:28:11HylianSavior has joined
09:28:26bobdaduckbroke timer.lua
09:28:31bobdaduckdeltaTime is now nil.
09:28:32bobdaduckwhat?
09:28:36kaenthe top one is the inner most function call
09:28:52HylianSavior Quit (Read error: Connection reset by peer)
09:29:00kaen"local variables" are variables declared with local inside of the corresponding function
09:29:01bobdaduckHUSH YOU
09:29:05bobdaduckI'M STILL PISSED ABOUT CIRCLES
09:29:56kaenI'm not really sure how you set deltaT, but nils are contagious
09:30:43kaenand those (*temporary) entries are temporary results of the arithmetic on the line in question
09:31:01kaenor return values of nested function calls
09:31:43kaenoh whoa
09:31:49kaenthat is indeed inside of timer.lua
09:32:23kaenbobdaduck, are you using . instead of : ?
09:32:26bobdaduckI've never used scheduleRepeatWhileTrue. I'm almost inclined to assume that that's what's broken, not my code
09:32:48bobdaduckAnd, no.
09:33:01kaenare you supplying a delay value?
09:33:18kaenyou should paste the code making that call
09:33:36bobdaduckuh
09:33:39bobdaduckthe code is 2000 lines long
09:33:53bobdaduckMaybe I can pull out the relevant parts
09:34:08kaenyou should just paste the whole thing because it could all be relevant
09:34:21kaenthat's what line numbers are for :)
09:37:21bobdaduckhttp://pastie.org/8401592
09:37:34bobdaduckThat's not the whole thing
09:38:02bobdaduckBut any problems are happening in that somewhere.
09:38:42kaenvar2 is nil
09:39:54kaenit will be nil whenever arg3 is a string that can't be interpretted as a number
09:39:59kaensee, that wasn't so hard
09:40:26kaentook you longer to argue with me than for me to solve your problem :)
09:41:06bobdaduckuh?
09:41:22bobdaduckline 25 of the pastie arbitrarily sets var2
09:41:57kaenit's hitting line 23
09:42:06kaenhttp://pgl.yoyo.org/luai/i/tonumber
09:42:18kaenbookmark that page and use those docs enthusiastically
09:43:44bobdaduckIf that's true then I should have been dealing with this way before now
09:43:59bobdaduckany time someone types a phrase longer than one word.
09:44:13bobdaduckbecause I use the exact same method for var
09:45:10kaenyou do nilchecks on var before you use it
09:47:01bobdaduckIts pretending to be fixed
09:50:22bobdaduckafter wrapping the entire function in a nil check
09:51:07Invisible1 Quit (Ping timeout: 240 seconds)
10:06:07Nothing_Much Quit (Read error: Connection reset by peer)
10:13:37Nothing_Much has joined
10:13:51bobdaduck stomps out, still pissed about circles
10:13:55bobdaduck has left #bitfighter
10:16:12HylianSavior has joined
10:46:19HylianSavior Quit (Quit: Leaving)
11:01:17watusimoto Quit (Ping timeout: 252 seconds)
11:05:35Flynnn has joined
11:46:06LordDVG has joined
11:47:38Flynnn Quit (Ping timeout: 272 seconds)
11:53:09Watusimoto has joined
11:58:07HylianSavior has joined
11:59:05LordDVG Quit (Remote host closed the connection)
12:15:32Nothing_Much Quit (Remote host closed the connection)
12:16:54Nothing_Much has joined
12:21:35Nothing_Much Quit (Remote host closed the connection)
12:51:41Watusimoto Quit (Ping timeout: 265 seconds)
13:16:23Flynnn has joined
13:18:55Flynnn Quit (Client Quit)
13:29:24Watusimoto has joined
13:37:58Flynnn has joined
13:51:17koda has joined
14:10:38Nothing_Much has joined
14:17:52Nothing_Much Quit (Read error: Connection reset by peer)
14:18:42Nothing_Much has joined
14:20:49thread_ has joined
14:21:57thread_Bob doesn't seem to be on the IRC channel as much recently
14:31:58Flynnn Quit (Quit: This computer has gone to sleep)
14:36:12Nothing_Much Quit (Read error: Connection reset by peer)
14:40:42kaenindeed
14:49:43Flynnn has joined
14:57:36Nothing_Much has joined
14:59:31Flynnn Quit (Quit: This computer has gone to sleep)
15:01:14Flynnn has joined
15:04:49Flynnn Quit (Client Quit)
15:23:44Watusimotoman, I'm getting sick of tests... trying to reproduce a engieneer bug by simulating key strokes to the client
15:24:04Watusimotobut I need a level and need to fly around the level to get a resource and start engineering
15:24:23Watusimotohopefully this work will push forward our test writing techniques
15:25:10WatusimotoI'm handcrafting SDL events at the moment
15:27:46raptor has joined
15:27:57ChanServ sets mode +o
15:29:02raptorhello
15:31:19kodaHiii
15:31:45kodaWatusimoto: You so lucky to have tests around
15:32:00thread_hey
15:32:06Watusimotohey koda
15:32:17kodaBtw I know it's early but will I see any of you at next fosdem?
15:32:21Watusimotoour testing is very very spotty
15:32:27Watusimotowhen is it? next Feb?
15:32:31kodaY
15:32:35Watusimotoin Brussels?
15:32:52koda31/01-1/02 same place of last year
15:33:00WatusimotoI sure hope so
15:33:04raptorif it ever happens within 100 miles or so, I might go...
15:33:15Watusimotoit was definitely worth the minimal effort it required last year :-)
15:33:21koda100000miles :p
15:33:36kodaraptor: You should have FOSM nearby
15:33:59Watusimotoyeah -- it's not like there's no cool open source events in the us
15:34:19Watusimotomaybe not so much in SLC, but perhaps vegas?
15:34:46kodaWatusimoto: Did I tell you that I jumped in a guy who had the same hat you said you were going to wear but he wasn't you? :)
15:35:00WatusimotoI think you did mention it
15:35:10WatusimotoI'll wear it again :-)
15:35:18Watusimotoor something
15:35:49Watusimotoor maybe I'll wear this http://img1.etsystatic.com/007/0/6852185/il_570xN.404583777_23av.jpg
15:35:58Watusimoto(mine has no pom-pom, sadly)
15:36:18koda Quit (Remote host closed the connection)
15:38:10raptorWatusimoto: I was thinking of adding ability to repair to a teleporter, but reducing it's health
15:38:32Watusimotosorry, i did respond to that last night, briefly before falling alseep again
15:38:40WatusimotoI said sure
15:38:48Watusimotorepair teleporter makes sense
15:38:55raptorkaen brought up the issue that the main difficulty with them is that you have to kill them at the start point
15:39:05Watusimotoyes, true
15:39:19raptorso a reduction in health might balance it a bit better
15:39:21WatusimotoI see... so you put the intake in a secure location, and it becomes permanent
15:39:49Watusimotoactually, you could probably put the intake somewhere inaccessible and make it invulnerable
15:40:10raptorhuu?
15:40:13raptorhuh?
15:40:32Watusimotosome levels have a room you spawn in with teleporters going out
15:40:45Watusimotoactually, that wouldn't work... there would be no resource item in there
15:41:14Watusimotobut yes, I think your proposal makes sense
15:41:25raptori'm just thining of balancing the item despite map maker abuse...
15:41:30raptorok
15:41:32Watusimotoanother idea would be to make it possible to destroy the outlet
15:41:47Watusimotoperhaps by making it 1) invisible (as it currently is) and 2) really small
15:42:05Watusimotothough I guess you could blanket the area with bursts and be sure to kill it
15:42:09Watusimotoso scratch that
15:42:10raptorwell, it does have an outline on it
15:42:30WatusimotoI like your proposal better
15:42:33raptoryeah - killing the outlet really kills the whole object
15:42:40Watusimotoyes
15:45:53raptorok, what do you think 1/2 health as it has now? 3/4?
15:46:04Watusimotohow many shots should it take to kill?
15:46:18WatusimotoI feel like it should be easier to kill with phaser than burst
15:46:23Watusimotoas phaser requires more skill
15:46:24raptorright now it takes 20 or so, I think (equal to turrets/FFs)
15:46:28Watusimoto20?
15:46:31Watusimotoway high
15:46:35raptorI will test...
15:46:43Watusimotowhat about 1/2 then?
15:46:48raptorok
15:47:10Watusimoto(maybe define its health in terms of phaser damage, to be clear :-)
15:48:37raptorwell, technically engineered items have damage reduction at 1/4
15:49:05raptorso phaser used to do .21, it would do 0.0525
15:49:15raptornow it's .19 so .0475
15:49:37raptoroh yeah... 1/2
15:50:54koda has joined
15:51:10kodaback ^^
15:52:20raptorsamurai jack
15:57:15raptorWatusimoto / kaen I'm getting an assert everytime I engineer a teleporter: Assert: packUpdate out of range teleporter number in teleporter.cpp line 415
15:57:18Watusimotokoda: we're using gtest, and it seems really great
15:57:30Watusimotoraptor: that's what I'm writing the test for!!
15:57:38raptoroh
15:57:45raptorwhen did that appear??
15:57:54WatusimotoI noticed it yesterday
15:58:06kodagtest? i wonder how that interfaces with cmake and freepascal ^^
15:58:27WatusimotoI'm guessing it arrived when we (I) was monkeying with the teleporter destinations stuff
15:58:32Watusimotofixing a different bug
15:58:46Watusimotowhich would have been a week or two ago
15:58:53Watusimotoit's not like we test this very much
15:59:10raptorwant me to do a quick hg bisect?
15:59:15Watusimotosure
15:59:18kodahttp://code.google.com/p/googletest/?
15:59:27Watusimotobetween bisect and the tests, we'll nail it for sure!
15:59:30Watusimotokoda: yes
15:59:43Watusimotoit probably will not help you with a pascal project
16:01:32kodaoh you write unit tests in c++
16:01:38kodabut how to interface it hmmm
16:01:48koda is too tired now but will look it up
16:02:08kodaand to think unit testing was one gsoc ideas...
16:02:12raptorseems more like integration tests
16:02:18raptor needs to learn the system
16:04:17WatusimotoI like bigger, grander tests... unit tests can be silly sometimes
16:05:14kodaunit testing: https://gs1.wac.edgecastcdn.net/8019B6/data.tumblr.com/1106e60d96005cd07bb4860bb6b1d82a/tumblr_mgo4ucGWTk1qzlfumo1_500.gif
16:07:54Nothing_Much Quit (Remote host closed the connection)
16:08:08raptorWatusimoto: also notice that now when you start on a level with teleporters, they start as though a ship has gone into them
16:08:53WatusimotoI hadn't noticed... will look
16:12:41Watusimotokoda: indeed!
16:14:30raptorJust thought I should let you know that I've noticed a big improvement on you checking in cross-platform broken code in that past few months - it makes bisect a lot easier
16:15:17raptorWatusimoto: the culprit is: changeset: 8200 917dca20608a
16:15:46Watusimotothat's the one!
16:16:05raptoraww man, looks complicated..
16:16:09Watusimotowell, luckily I did test coverage with that fix, so it will make it easier to fix the new bug without breaking the old
16:16:37Watusimotono not really
16:17:21Watusimotoit just pushes cleans up the pack/unpack for teleporter
16:18:00Watusimotothere's lots of trivial cleanup in there
16:18:13Watusimoto(picking one at random)
16:18:14Watusimoto- for(S32 i = mDestManager.getDestCount() - 1; i >= 0; i--)
16:18:14Watusimoto+ for(S32 i = getDestCount() - 1; i >= 0; i--)
16:18:46Watusimotothat one, for example, replaces a call to dest manager with a built-in method that does the same
16:18:56Watusimotomost of the edits are stuff like that
16:19:11WatusimotoI was just trying to consolodate calls and make the code more uniform
16:20:22Watusimotobetween that and replacing getVert(0) with getOrigin() accounts for probably half the changes
16:25:01Nothing_Much has joined
16:31:37fordcars has joined
16:45:12thread_ Quit (Quit: Page closed)
17:05:37kaenwell, because of how old the codebase is, a lot of the classes are very tightly coupled
17:05:46kaenso what should be unit tests end up as integration tests
17:06:08kaenthat being said, wat's done some really good decoupling work so we're getting closer to true unit tests
17:06:38kaenand as we write tests bit by bit, our coverage well keep getting better
17:06:58kaenthere's really no sense in writing tests for stable, functional production code
17:07:13kaenbut rather we should just do it as we add features or refactor
17:09:47Watusimotoyes
17:11:02WatusimotoI just wrote a method that inquires with the engineering helper menu to find out what key it should press to activate the creation of a teleporter entrance
17:11:08WatusimotoKEY_3 is the answer
17:11:21raptorheh
17:11:24Watusimotowill probably never change... but if it does, we're covered!
17:11:40Watusimotoso now I need to simulate a keypress for KEY_3
17:12:04Watusimotothis is a true integration test!
17:14:44raptorpurple cornbread!
17:15:12kaenI've got a working UIEditor test
17:15:20kaenthe final frontier...
17:15:24raptorwhoa! really?
17:15:30kaenyep!
17:15:36kaenand it doesn't even segfault
17:16:04kaennow we just need to work on implementing sane default constructors and we'll be golden
17:16:49kaenthis GamePair thing is too much of a crutch for my taste
17:19:48Watusimotoit's a total crutch, but saves a lot of repeat code
17:19:53kaenindeed
17:20:00Watusimotoif you can see a good way to refactor it away... I'd be happy!
17:20:33kaenideally things that need a Game in order to function would just create one in their default constructor
17:20:39Watusimotothe problem is the games need a gameType, and that is a game object, and so can;t be made in the constructor
17:20:49koda Quit (Quit: koda)
17:20:58Watusimotowe need a coupled pair of games, which share a gameType
17:21:16Watusimotoso there are several steps that are required, which don;t fit into a constructor
17:22:05kaenwell, those objects should create "zero-value" game, which would create a "zero-value" gametype, etc. on down the line
17:22:16kaenonly in their default constructor, which would be used only in tests
17:24:03kaenthen they would keep the normal "injected" constructors for use in production or tests that need special injected objects
17:24:39WatusimotoI see the design you are describing, and agree it would be nice; I'm just not sure it would work in this case, specifically because of the GameType object
17:25:06Watusimotoif you created your defaults as you described, then read a chunk of level code, you'd have two gametypes
17:25:26Watusimotoor you'd need a way to propagate to the client a message to discard the first gametype
17:25:40Watusimoto_maybe_ you could do it wiht a change level message
17:25:46kaenonly in paired client/server tests, where we could actually use GamePair for what I intended
17:26:11Watusimotoso a default level (i.e. "") gets loaded first, with a default gametype, then you read the level code, and then change to the new level
17:26:39Watusimotothat might work, and that might get you the elegance you crave
17:26:42kaenthat sounds exactly like what I'm envisioning
17:29:39Watusimotobut you still need to create your client/server games at the same time, so they will be linked
17:29:47Watusimotoand for that you need... a gamepair!
17:30:25kaenright, right
17:30:34Watusimotoor if you wanted to test client-client interaction.., a game triplet1
17:30:36kaenthis default construction business is more for unit tests i.e. with a single game
17:31:16kaenso that, for instance, if I want to test an EditorUserInterface, I don't have to first construct a ClientGame in my tests
17:32:31kaenI could just do: EditorUserInterface editor;
17:32:33kaenand get to work
17:32:53kaenand then also not have to juggle pointers and stuff in my test
17:34:14Watusimotoyes, good point
17:34:21WatusimotoI am just amazed that this all works
17:34:41WatusimotoI can pump events and keys into a client and the server responds
17:36:37kaenhehe
17:48:28Watusimotobed time for me
17:48:48BFLogBot Commit: d882b3607997 | Author: watusimoto | Message: Several intertwined rendering fixes
17:48:50BFLogBot Commit: bdb2fe136b3f | Author: watusimoto | Message: Make sure every call to getShip can handle NULL -- this was the cause of the crash fixed in 996cc24470b3394f34925d4e0b9cc51b29432bc1
17:48:51BFLogBot Commit: f3d3090b1576 | Author: watusimoto | Message: Test coverage demonstrating teleporter engineer assertion failure. Does not fix. Yet.
17:48:55WatusimotoI am checking in a test that demonstrates the teleporter bug, but the fix will have to wait until tomorrow.
17:50:48Watusimototomorrow I start french classes!
17:52:32fordcarsHehe you'll learn french?
17:52:52Watusimotooui
17:53:25fordcarsHehe cool
17:54:06fordcarsIf you need any help, French and English are both my primary languages :P
17:57:44raptor Quit ()
17:58:17Watusimotogreat! My wife and kids also speak fluent french, so I should be set!
17:58:25Watusimotobon nuit!
17:59:09fordcarsNight!
18:02:59Watusimoto Quit (Ping timeout: 268 seconds)
18:16:47Flynnn has joined
18:27:34Flynnn Quit (Quit: This computer has gone to sleep)
18:40:19fordcars Quit (Ping timeout: 250 seconds)
18:42:31BFLogBot Commit: c29f12752eb8 | Author: buckyballreaction | Message: Fix compiling with SDL 1.2. SDL_Keycode is SDL2 only. The related gtests are probably broken
18:53:50BFLogBot Commit: 1c6fd0396916 | Author: buckyballreaction | Message: Engineered teleporters now only take half as many hits to kill
18:56:51fordcars has joined
18:58:34Nothing_MuchHey fordcars
18:58:39fordcarshi
18:58:49Nothing_MuchWhat's up?
18:59:27fordcarsNot much
18:59:33fordcarsYou?
18:59:46Nothing_MuchTrying to figure out this convoluted mess called OpenID
19:00:04Nothing_MuchI seriously don't understand it at all
19:00:50fordcarsHah ok
19:01:15fordcarsLooks cool though
19:02:13Nothing_MuchI don't understand how to use it on launchpad
19:02:33Nothing_MuchApparently I have 3, 2 on Gmail and 1 on Yahoo
19:02:35Nothing_MuchBleh
19:02:44Nothing_MuchSuch a mess when I want Launchpad only
19:03:30fordcarshah
19:03:42fordcarsWhat are you looking for?
19:03:52fordcarsSome server-side stuff for logging-in?
19:04:04Nothing_MuchI'm TRYING to use Launchpad
19:06:04Nothing_MuchOh screw it, I-
19:06:12fordcarsOh is launchpad a distribution version control system?
19:06:23fordcarsLike mercurial?
19:06:48Nothing_MuchDistribution control system?
19:06:50Nothing_MuchI think so
19:07:22Nothing_Muchcheck it out though
19:07:25Nothing_Muchlaunchpad.net
19:07:29fordcarsWhat's up with Bazarr?
19:07:45Nothing_MuchI think it's more like Github, but it also hosts .debs for whatever you want.
19:07:59fordcarsThere is an irc channel: #launchpad
19:08:02fordcarsOn freenode
19:08:45Nothing_MuchI guess I'll just stick to the crappy Gmail stuff
19:08:54fordcars:P
19:08:55Nothing_MuchBut if Gmail expects me to use my first and last name
19:09:04Nothing_MuchOho boy am I gonna complain.\
19:09:11fordcarsJust make a new one?
19:09:44Nothing_MuchNo
19:09:45Nothing_MuchI mean
19:09:55Nothing_MuchGmail FORCES you to pick a First and Last name
19:10:03Nothing_MuchIt's BS
19:10:04fordcarsHah maybe
19:10:07fordcars:P
19:10:25Nothing_MuchIt is, especially when I wanna keep my internet and real life separate.
19:10:26fordcarsI hate it when it asks that
19:16:08Nothing_MuchYeah
19:17:35Nothing_MuchAny luck on the RPi?
19:19:01fordcarsNot yet, I had quite allot of work to do these past days :P
19:19:08Nothing_MuchAh, cool
19:20:07Nothing_MuchOMFG
19:20:14Nothing_MuchThat scared the hell outta me..
19:20:39BFLogBot Commit: e1f59639ae23 | Author: buckyballreaction | Message: Fix admin making commands on master: - Fix relaying some admin commands to clients as chats (oops) - Fix allowing admins to send private messages
19:22:08fordcarsWhat??
19:23:49Nothing_MuchI thought I lost the entire source of the trailer... ;_; but it recovered itself properly :)
19:35:16fordcarshaha
19:36:51amgine1234567890 has joined
19:36:55amgine1234567890sup
19:48:05kaenhi o/
19:49:45Nothing_MuchHello
20:17:32Flynnn has joined
20:38:44Flynnn Quit (Quit: This computer has gone to sleep)
20:49:01fordcars Quit (Ping timeout: 250 seconds)
20:49:20Flynnn has joined
21:29:26fordcars has joined
21:29:38Flynnn Quit (Quit: This computer has gone to sleep)
21:30:26Flynnn has joined
21:30:26Flynnn Quit (Client Quit)
22:09:28HylianSavior Quit (Quit: Leaving)
22:13:35HylianSavior has joined
22:13:42HylianSavior Quit (Read error: Connection reset by peer)
22:19:35amgine1234567890 Quit (Ping timeout: 250 seconds)
23:13:12Platskies has joined
23:16:00HylianSavior has joined
23:59:15fordcars Quit (Ping timeout: 250 seconds)

Index Search ←Prev date Next date→

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