Timestamps are in GMT/BST.
| 02:03:58 | | IAmBeard Quit (Quit: Leaving) |
| 02:47:04 | | zoomber_mbp has joined |
| 02:47:09 | zoomber_mbp | hi raptor |
| 02:48:23 | | zoomber_mbp Quit (Client Quit) |
| 03:44:33 | raptor | hi |
| 03:44:35 | raptor | bye |
| 03:44:46 | sam686 | hi |
| 03:45:21 | raptor | hi |
| 03:46:00 | sam686 | i seem to now find that rev b909d416e42a is not crashy anymore.. |
| 03:46:21 | sam686 | but, a51117a398a have a problem with NULL mGeometry crash... |
| 03:46:40 | raptor | yes, your fix seemed to fix everything |
| 03:46:44 | raptor | how did you find that? |
| 03:53:13 | sam686 | mostly, getFiringSolution gets 2 arguments, and robot_helper_functions.lua seem to convert one arg to 2 args, the first arg is item:getClassID(), the second is the item itself.. |
| 03:53:37 | sam686 | i find the first arg was completely wrong (zero most of the time) |
| 03:53:54 | sam686 | and that was asteroid |
| 03:53:59 | raptor | how did you dump the argument from th *L? |
| 03:54:10 | raptor | i'm not very good with the lua stuff... |
| 03:54:48 | sam686 | I simply put a breakpoint somewhere in LuaRobot::getFiringSolution, to look at type number |
| 03:54:58 | | FlynnnNT has joined |
| 03:59:31 | raptor | watusimoto thought it was weird that your fix would fix the stack overflow |
| 03:59:43 | raptor | he seemed to think that maybe LuaExceptions might leak |
| 04:26:25 | raptor | i've been thinking about reimplementing a replay Journal |
| 04:26:40 | raptor | looks like a lot of work |
| 04:27:00 | raptor | this is an excellent article: http://www.gamasutra.com/view/feature/2029/developing_your_own_replay_system.php |
| 04:30:52 | sam686 | looks like thinks break starting at rev 6ff27680abc5, added "GeomObject.cpp" file, null mGeometry |
| 04:37:08 | sam686 | as for replay system, computers with more then 1 GB of free disk space can easily handle "Saving The State" by writing each frame to disk.. |
| 04:39:51 | sam686 | i think the old bitfighter 014 or older sort of use the "Deterministic " replay system, but it has to also record all incoming network communications, not just the controller input.. |
| 04:51:20 | raptor | i know that IAmBeard said saving physics states was better - but I think it would use LOTS of memory |
| 04:51:27 | raptor | which may not be better |
| 04:52:19 | raptor | besides - that article talks about a hybrid technique - deterministic with 'keyframes' where you'd save the entire physics state every minute for faster jumping.. |
| 04:53:06 | sam686 | i am sure it won't use lots of RAM, when those frames can be saved in one file while saving the next and the next... |
| 04:53:42 | sam686 | its kind of like saving screenshots in file, you won't screenshot and hold all the screenshot in RAM, would you? |
| 04:55:35 | sam686 | but either way, "deterministic" may save lots of disk space compared to saving each frame state.. |
| 04:57:30 | sam686 | what the article is talking about is on a limited memory, no hard disk playstation 2 system.. |
| 04:58:56 | raptor | so do you think deterministic would be better or state-saving would be better? |
| 04:59:58 | sam686 | probably both, state-save every 20 seconds or so to allow fast seeking.. |
| 05:00:30 | raptor | i agree |
| 05:00:33 | sam686 | hundreds of asteroids can make state-save only grow very big on file size |
| 05:00:44 | raptor | yes, i bet it would be HUGE |
| 05:02:39 | sam686 | in a network game like bitfighter, the server can do the replay saving, as it can save everything in the game. |
| 05:02:52 | raptor | yes |
| 05:02:57 | sam686 | but, maybe a client can do replay saving too, mostly a replay of making a level in editor? |
| 05:03:17 | raptor | that's an intereting idea |
| 05:03:26 | raptor | and you could distribute it like a tutorial |
| 05:05:44 | sam686 | the keyframe stuff sort of reminds me of video compressing having keyframe about every 2 seconds, as in mpeg 1-4 video compression.. |
| 05:06:56 | sam686 | lagarith RGB lossless compression fast enough for my computer, though every frame is keyframe for compression isn't that much... (same with huffyuv) |
| 05:10:45 | raptor | yes, me too |
| 05:13:18 | raptor | should we think about separating the client GUI rendering into a different thread |
| 05:13:19 | raptor | ? |
| 05:14:20 | sam686 | not sure if that will improve performance, unless you are hosting... |
| 05:15:20 | sam686 | bitfighter can run on old computer, most of old computer don't have more then one core CPU.. |
| 05:15:37 | sam686 | most newer computer is so fast, it can run bitfighter without multithreading.. |
| 05:15:48 | raptor | yeah, ok - i just keep thinking that there should be some way to take advantage of multithreading for servers... |
| 05:16:16 | sam686 | and many VPS servers have only one CPU allowed to use.. |
| 05:16:31 | raptor | yes - i think we should probably just keep bitfighter very low CPU |
| 05:24:43 | sam686 | i somewhat find the code a bit more messy with all the Geometry and EditorObject stuff, in bitfighter 013, it looks cleaner with mostly the lack of any editor specific functions in game objects, the stuff dedicated server don't use. |
| 05:25:07 | raptor | yes, that was when they were all WorldItems |
| 05:25:26 | raptor | and the Editor was almost completely self-contained |
| 05:25:49 | sam686 | although, that tends to make UIeditor.cpp a bit messy, but good thing that was only one file only.. |
| 05:26:04 | sam686 | ... until the mess spreads to so meny files.. |
| 05:26:24 | raptor | yeah - at least we don't have the diamond class problem anymore.. |
| 05:27:08 | raptor | i only find one relationship that is messy now: http://bitfighter.org/~raptor/doxygen/post017b/class_zap_1_1_burst_projectile.html |
| 05:28:05 | sam686 | all those "isSelected" functions in Geometry can easily be tracked on UIEditor only, by using Vector<Object*> ListOfSelectedObjects; |
| 05:30:46 | sam686 | i see a problem, Geometry::flip calls setVert, will not call MoveObject::getVert ... because Geometry is not part of MoveObject |
| 05:30:58 | sam686 | maybe thats why flip don't work.. |
| 05:31:26 | raptor | i haven't done any testing - i figure watusimoto isn't done moving things around |
| 05:33:29 | sam686 | i am guessing the game code size and binary size could easily be much smaller in code size if the LUA and most of Editor stuff is taken away... |
| 05:33:38 | raptor | i'm falling asleep |
| 05:33:44 | raptor | yep |
| 05:33:52 | sam686 | ok, night.. |
| 05:33:54 | raptor | good night |
| 05:34:00 | | raptor Quit () |
| 05:39:19 | | sam686 Quit (Ping timeout: 245 seconds) |
| 05:40:01 | | sam686 has joined |
| 05:40:01 | | ChanServ sets mode +v sam686 |
| 06:18:54 | | sam686 Quit (Ping timeout: 245 seconds) |
| 06:19:31 | | sam686 has joined |
| 06:19:31 | | ChanServ sets mode +v sam686 |
| 07:18:54 | | sam686 Quit (Ping timeout: 245 seconds) |
| 07:19:34 | | sam686 has joined |
| 07:19:34 | | ChanServ sets mode +v sam686 |
| 07:41:22 | | kodaws has joined |
| 07:50:21 | | FlynnnNT Quit (Quit: This computer has gone to sleep) |
| 07:51:22 | | FlynnnNT has joined |
| 07:52:56 | | watusimoto has joined |
| 07:52:56 | | ChanServ sets mode +o watusimoto |
| 08:46:24 | | sam686 Quit (Ping timeout: 245 seconds) |
| 09:10:15 | | FlynnnNT Quit (Quit: Leaving) |
| 09:10:36 | | FlynnnNT has joined |
| 10:08:43 | | FlynnnNT Quit (Quit: Leaving) |
| 13:25:12 | | IAmBeard has joined |
| 15:24:42 | IAmBeard | 'morning |
| 15:53:31 | | LordDVG has joined |
| 16:28:35 | | raptor has joined |
| 16:28:35 | | ChanServ sets mode +o raptor |
| 16:28:55 | raptor | good mornign |
| 16:50:02 | | LordDVG Quit (Remote host closed the connection) |
| 16:50:20 | | watusimoto Quit (Read error: Operation timed out) |
| 16:57:20 | | kodaws Quit (Read error: Connection reset by peer) |
| 17:26:15 | | Heyub has joined |
| 17:32:25 | | koda has joined |
| 17:33:06 | | Heyub|2 has joined |
| 17:33:18 | Heyub|2 | I think my airport card got burned up... |
| 17:33:29 | raptor | oh yikes |
| 17:33:36 | raptor | too much web-browsing? :) |
| 17:33:38 | Heyub|2 | I can't keep connection to anything |
| 17:34:12 | | Heyub Quit (Ping timeout: 246 seconds) |
| 17:34:25 | Heyub|2 | It runs very hot... |
| 17:34:35 | Heyub|2 | I literally mean it might have fried |
| 17:34:54 | raptor | ice cube trays |
| 17:35:06 | Heyub|2 | Six inch desk fan. |
| 17:35:11 | raptor | stick the laptop on ice cube trays |
| 17:35:14 | raptor | or that |
| 17:36:00 | Heyub|2 | I bought a six inch USB fan pad for it but windows wont run it any faster than slow... |
| 17:39:02 | | Heyub has joined |
| 17:40:51 | | Heyub|2 Quit (Ping timeout: 246 seconds) |
| 17:43:39 | | Heyub Quit (Ping timeout: 246 seconds) |
| 19:00:15 | | Watusimoto has joined |
| 19:31:26 | Watusimoto | hi |
| 19:31:30 | raptor | hello |
| 19:36:34 | raptor | sam686 said something about mGeometry being null somewhere and crashing |
| 19:36:39 | raptor | since your latest merge |
| 19:36:44 | Watusimoto | great |
| 19:37:06 | raptor | i haven't gotten a crash yet, but i've specifically stayed away from the editor... |
| 19:37:12 | Watusimoto | I've got a lot of things ripped apart right now |
| 19:37:20 | Watusimoto | trying to kill off luaItem |
| 19:37:28 | Watusimoto | and that whole branch of the tree |
| 19:37:28 | raptor | ooooo |
| 19:37:33 | raptor | cool |
| 19:37:36 | Watusimoto | well, I;ve killed it, but nothing works |
| 19:37:49 | Watusimoto | now bfObject inherits right from luaObject |
| 19:38:02 | Watusimoto | and everything flows more normally |
| 19:39:44 | Watusimoto | I've also decided I should probably get rid of EditorObject again |
| 19:39:55 | Watusimoto | I really like the idea of it, but I just can;t make it work |
| 19:40:06 | Watusimoto | it depends on too many other things |
| 19:40:17 | Watusimoto | and it is easy to get sucked back into the dread diamond |
| 19:40:28 | Watusimoto | which we DO NOT WANT |
| 19:40:36 | raptor | do not want |
| 20:02:13 | raptor | well, just tell me if you want doxygen re-run |
| 20:24:41 | | sam686 has joined |
| 20:24:41 | | ChanServ sets mode +v sam686 |
| 20:37:18 | Watusimoto | luarobot is gone |
| 20:37:27 | raptor | RIP |
| 20:51:28 | | Heyub has joined |
| 20:53:16 | Watusimoto | luaship is gone |
| 20:59:13 | raptor | oh b oy |
| 21:02:56 | Watusimoto | of course, nothing compiles |
| 21:16:48 | raptor | hooray! |
| 21:28:55 | Watusimoto | link errors out the wazoo! |
| 21:31:12 | raptor | new file missing? or instantiated something in a header? |
| 21:43:05 | Watusimoto | just tons of broken stuff |
| 21:44:23 | Watusimoto | built! |
| 21:51:06 | sam686 | starting with "6ff27680abc5", the adding of GeomObject.cpp, is when the problem of NULL mGeometry started.. |
| 22:00:36 | raptor | ok heading out, be back later |
| 22:00:38 | | raptor Quit () |
| 22:24:04 | Watusimoto | what's the NULL problem? |
| 22:27:01 | sam686 | trying to add spawn points in editor |
| 22:27:22 | sam686 | goes for some other objects as well.. |
| 22:31:03 | Watusimoto | odd |
| 22:31:12 | Watusimoto | I'm having problems with spawns in the editor as well |
| 22:31:16 | Watusimoto | but different |
| 22:31:22 | Watusimoto | but messed up |
| 22:31:44 | Watusimoto | toString is being called instead of processArguments |
| 22:32:11 | sam686 | i don't have any of your latest changes yet (the last one i see was 25 hours ago) |
| 22:32:25 | Watusimoto | you don't want them just now |
| 22:32:45 | sam686 | maybe your changes breaks even more? |
| 22:32:56 | Watusimoto | quite likely! |
| 22:33:27 | Watusimoto | never seen this before... I step into one method, but arrive at another |
| 22:33:34 | Watusimoto | ok, time for a complete rebuild |
| 22:33:56 | Watusimoto | if I can get into the editor, I'll take a look at the error you're having |
| 22:34:38 | sam686 | make sure you are in debug build, as usual, it helps with debugging.. |
| 22:35:48 | Watusimoto | yes |
| 22:35:52 | Watusimoto | always |
| 22:36:06 | sam686 | those printf in LuaWrapper.h is making my bitfighter_server.log files grows up to 0.5 GB in size... |
| 22:37:18 | Watusimoto | full rebuild fixed my problem |
| 22:37:21 | sam686 | its those logprintf, besides increasing log file sizes, it slows down the hosting with robots.. |
| 22:37:23 | Watusimoto | wow |
| 22:37:38 | Watusimoto | well, all that garbage will be stripped out when things are stable |
| 22:37:55 | Watusimoto | mmm |
| 22:38:02 | Watusimoto | null geometry in a resource item |
| 22:44:12 | Watusimoto | there's garbage in the database |
| 22:45:23 | sam686 | the null mGeometry problem started in rev (6ff27680abc5) Break out GeomObject into a separate file, make GeomObject a parent class of Database... |
| 22:46:25 | sam686 | maybe a problem with copy constructors? |
| 22:46:36 | Watusimoto | my guess is it's the EditorObjectDatabase |
| 22:46:47 | Watusimoto | that somehow things are being miscasted |
| 22:47:20 | sam686 | this null mGeometry also happens when flags is trying to return in CTF or Zone control.. |
| 23:16:38 | | raptor has joined |
| 23:16:38 | | ChanServ sets mode +o raptor |
| 23:16:43 | raptor | hi again |
| 23:19:55 | | koda Quit (Quit: koda) |
| 23:27:35 | sam686 | i will go, may be back in about an hour.. |
| 23:27:40 | raptor | byebye |
| 23:47:01 | raptor | uh oh - if i don't commit something soon, i'm in danger of disappearing from the first google code changes page! |
| 23:48:17 | Heyub | what? |