Timestamps are in GMT/BST.
| 02:36:00 | Zoomber | `Reading symbols for shared libraries . done |
| 02:36:00 | Zoomber | Reading symbols for shared libraries . done |
| 02:36:00 | Zoomber | Program received signal SIGTERM, Terminated. |
| 02:36:00 | Zoomber | 0x00000001000de0e9 in Zap::DatabaseObject::setExtent () |
| 02:38:00 | sam686 | it doesn't tell much, try compiling and running in debug mode? |
| 02:38:00 | Zoomber | ok |
| 02:39:00 | Zoomber | wait |
| 02:39:00 | Zoomber | i want to try /maxfps 1001 on my windows emulator |
| 02:41:00 | sam686 | i can go over 1000 fps (maxfps 1001) up to 2000 fps, only when i run bitfighter in a tiny window.. |
| 02:41:00 | Zoomber | i Vsync when i go to a tiny window or fullscreen |
| 02:41:00 | Zoomber | but, in virtual machine, i can only get 69 fps |
| 02:42:00 | sam686 | ok, now it finally freezes on me with 100% CPU usage |
| 02:42:00 | Zoomber | mine usually stay at 2000 or 20000 fps, cant tell which |
| 02:42:00 | Zoomber | when in small window |
| 02:42:00 | Zoomber | gbut cant read it |
| 02:42:00 | sam686 | wait, now it is leaking memory now and eating up all my RAM |
| 02:43:00 | Zoomber | i have over 2 GB free ram |
| 02:43:00 | Zoomber | oh |
| 02:43:00 | Zoomber | now all of a sudden |
| 02:43:00 | Zoomber | i |
| 02:43:00 | Zoomber | have 5 mb memory free |
| 02:43:00 | Zoomber | also |
| 02:43:00 | Zoomber | cpu fine |
| 02:43:00 | Zoomber | but no more ram left |
| 02:44:00 | Zoomber | dam, it used 2.6 GB memory |
| 02:46:00 | sam686 | it was badly slowing my computer down, then i get a "Runtime Error! Program: (empty) This application has requested the Runtime to terminate in an unusual way." |
| 02:48:00 | Zoomber | should i upgrade my RAM to 8 GB? only 30 dollars... |
| 02:49:00 | sam686 | it is a bitfighter problem eating too much RAM when it crashes with /maxfps 1001 |
| 02:50:00 | sam686 | for my computer, it used 2 GB memory, then crash.. My computer only have 1.5 MB ram, so it used a hard disk for extra memory, slowing down my computer. |
| 02:50:00 | sam686 | (it crash bitfighter) |
| 02:51:00 | Zoomber | i wonder why it eats so much ram there |
| 02:51:00 | sam686 | i wonder why it only crash with /maxfps 1001 |
| 02:52:00 | sam686 | i will test again, but with debugger, hit "break all" when it freeze, to see where it get stuck in. |
| 02:53:00 | Zoomber | ok |
| 03:01:00 | sam686 | i found the problem, fixing it... |
| 03:06:00 | Zoomber | what was problem? |
| 03:06:00 | sam686 | there, now I fixed it on my 015a clone.. |
| 03:06:00 | Zoomber | problem? |
| 03:06:00 | sam686 | it was a problem with this: if(U32(maxx - minx) > GridDatabase::BucketRowCount) maxx = minx + GridDatabase::BucketRowCount; |
| 03:07:00 | Zoomber | ok |
| 03:07:00 | Zoomber | pulling now |
| 03:08:00 | sam686 | without U32 part, it may crash when maxx - minx is negative. With U32, it forces it be positive and check. |
| 03:09:00 | sam686 | basically, w = U32(z), if z < 0, then it adds 4294967295 to make it positive (overflow effect) |
| 03:09:00 | sam686 | oh, i meant 4294967296 (not 5) |
| 03:11:00 | Zoomber | trying it now |
| 03:11:00 | sam686 | Here is a range of numbers that 32 bit number is allowed to have: U32 (0 to 4294967295 ) S32 ( -2147483648 to 2147483647 ) |
| 03:11:00 | Zoomber | ok |
| 03:12:00 | sam686 | might be easier to understand with 8 bit numbers: U8 (0 - 255) S8 (-128 to 127) |
| 03:12:00 | Zoomber | ok, i am not crashing now |
| 03:13:00 | Zoomber | mostly happening on map "CPU" when going through a bunch of speedzones |
| 03:13:00 | Zoomber | b |
| 03:13:00 | Zoomber | ut no |
| 03:13:00 | sam686 | conversion from S8 to U8 will add or subtract 256 if it overflows or underflows. |
| 03:13:00 | sam686 | does it work now? |
| 03:29:00 | Zoomber | yes, works now |
| 03:30:00 | | Zoomber Quit (Quit: Zoomber) |
| 07:59:00 | | kodax has joined |
| 08:22:00 | | kodax Quit (Ping timeout: 258 seconds) |
| 08:30:00 | | LordDVG has joined |
| 08:35:00 | | kodax has joined |
| 08:38:00 | | kodax Quit (Read error: Connection reset by peer) |
| 08:39:00 | | kodax has joined |
| 09:50:00 | | kodax Quit (Ping timeout: 245 seconds) |
| 10:10:00 | | kodax has joined |
| 13:03:00 | | LordDVG Quit (Remote host closed the connection) |
| 16:17:00 | | Flynnn has joined |
| 16:20:00 | | kodax Quit (Read error: Operation timed out) |
| 16:31:00 | | Flynnn Quit (Quit: This computer has gone to sleep) |
| 17:01:00 | | Flynnn has joined |
| 17:10:00 | | Flynnn Quit (Quit: Leaving) |
| 17:33:00 | | LordDVG has joined |
| 17:33:00 | | LordDVG Quit (Changing host) |
| 17:33:00 | | LordDVG has joined |
| 17:36:00 | | LordDVG Quit (Remote host closed the connection) |
| 17:36:00 | | LordDVG has joined |
| 17:36:00 | | LordDVG Quit (Changing host) |
| 17:36:00 | | LordDVG has joined |
| 18:26:00 | | Zoomber has joined |
| 18:26:00 | | ChanServ sets mode +v Zoomber |
| 20:51:00 | | raptor has joined |
| 20:51:00 | | ChanServ sets mode +o raptor |
| 20:57:00 | raptor | hello |
| 20:58:00 | | LordDVG Quit (Remote host closed the connection) |
| 21:06:00 | raptor | so sam686, Zoomber, found a memory leak? |
| 21:09:00 | sam686 | it is the problem in gridDB.cpp line 544: if(maxx - minx >= gridDB->BucketRowCount) |
| 21:09:00 | sam686 | using if(U32(maxx - minx) >= gridDB->BucketRowCount) will fix it.. |
| 21:09:00 | raptor | interesting, how does that become a memory leak? |
| 21:10:00 | raptor | so probably an overflow problem? |
| 21:10:00 | sam686 | when maxx - minx overflows to negative number, then it will skip that check if not checking as unsigned |
| 21:16:00 | raptor | ahh.. i get it |
| 21:16:00 | raptor | so not a traditional leak, but very fast build up |
| 21:17:00 | sam686 | yes, as as a loop lie this: for(S32 y = miny; maxy - y >= 0; y++) {GridDatabase::BucketEntry *be = gridDB->mChunker->alloc(); ...} will eat too much memory if the loop is run millions of times |
| 21:21:00 | raptor | so is a U32 cast better than an abs()? |
| 21:21:00 | raptor | seems like a lot of casting is happening in that function...\ |
| 21:22:00 | sam686 | abs may not work, as integer abs won't go to positive when the S32 value is -2147483648 (0x80000000) |
| 21:23:00 | | BFLogBot - Commit b0e95bb03d52 | Author: sam8641 | Log: Fix problem with GridDatabase freeze problem. |
| 21:25:00 | sam686 | when converting F32 to S32, when float to int overflows, the value is 0x80000000. Might be different overflow behavior in linux or other compilers. |
| 21:35:00 | sam686 | I came up with some list of problems that exists with 016. http://208.107.52.15/bitfighter/110904bugs.txt |
| 21:35:00 | raptor | oh goodie |
| 21:36:00 | raptor | looks like the editor is winning! |
| 21:36:00 | raptor | CircleSpawn doesn't spawn anymore? i thought it took about 20 seconds |
| 21:37:00 | sam686 | 10. was suppose to be: 10. Editor: Can't edit heal rate for Turrets and ForceField |
| 21:38:00 | raptor | i looked at why raw buttons don't work - i didn't see anything obvious |
| 21:39:00 | raptor | i can't duplicate #4 on my linux box |
| 21:40:00 | sam686 | number 19 may also be a problem in linux.. |
| 21:41:00 | raptor | does 19 occur when you run the dedicated part? |
| 21:41:00 | raptor | or client |
| 21:41:00 | sam686 | i guess number 4 is only a problem with ubuntu.. |
| 21:41:00 | sam686 | i think it is client only, expecially when client is hosting. |
| 21:41:00 | raptor | ok, let me try... hosting |
| 21:49:00 | raptor | ok, i'm hosting again |
| 21:49:00 | raptor | i turned off the sdl hardware surface flag |
| 21:59:00 | raptor | yeah, i'd mark #19 as invalid |
| 21:59:00 | raptor | also #2 doesn't occur on my linux, is it that big a concern? |
| 22:02:00 | sam686 | in #2, hiding mouse pointer then showing mouse pointer moves the pointer back to the middle, such as pressing up arrow may move the highlighted menu back to the middle... |
| 22:03:00 | sam686 | #1 may be more annoying then #2.. |
| 22:03:00 | raptor | hmm, looks like it can one of two things: 1. SDL bug with windows (or necessary workaround) 2. bitfighter code always resetting the mouse pointer at center |
| 22:04:00 | sam686 | well then, why does neither #1 or #2 not happen in linux? |
| 22:05:00 | sam686 | i think this might be a bug in windows version of SDL, that does not happen in linux. |
| 22:05:00 | raptor | yeah, i think so too |
| 22:05:00 | raptor | at least for #2 |
| 22:09:00 | raptor | can you explain bug #1 a bit better? |
| 22:10:00 | sam686 | easy, go full screen, go in-game as if you joined a server or hosting, then press ALT + TAB. mouse will then get stuck at top left corner of screen. |
| 22:10:00 | sam686 | note that linux seems to not let you alt+tab in fullscreen, but wildows does. |
| 22:11:00 | raptor | ok, i can alt+tab just fine in linux between all three window modes |
| 22:12:00 | raptor | also, now cursor stuck |
| 22:12:00 | raptor | but switching to either full-screen mode makes the cursor recenter |
| 22:12:00 | raptor | but not going from fullscreen to windowed |
| 22:13:00 | sam686 | my alt+tab don't work in ubuntu... |
| 22:14:00 | sam686 | in fullscreen bitfighter.. |
| 22:14:00 | raptor | well that stinks |
| 22:15:00 | raptor | i found somthing odd: http://208.107.52.15/upload/screenshot_6.png |
| 22:15:00 | raptor | at certain angles the shots don't go to the middle of the cross-hairs |
| 22:16:00 | raptor | some sort of rounding, i imagine.. |
| 22:18:00 | sam686 | this is 2 places it freezes. http://208.107.52.15/bitfighter/linux_ubuntu_freezes.txt |
| 22:18:00 | sam686 | the SDL_GL_SwapBuffers freezes while in console mode (ctrl + alt + F1) |
| 22:19:00 | sam686 | the Zap::Event::onResize(int, int) seems to be endless loop when maximized, but not fullscreen. |
| 22:22:00 | raptor | that's odd |
| 22:22:00 | raptor | that would be an SDL problem.. |
| 22:23:00 | raptor | if you look at the code, that method is only called if SDL sets the appropriate flag |
| 22:24:00 | Zoomber | hi |
| 22:25:00 | raptor | hi |
| 22:25:00 | raptor | sam686: can you put a logprintf statement at Event.cpp:443 |
| 22:25:00 | raptor | and see when it is actually triggered? and do things like maximize window/ full screen/ etc... |
| 22:26:00 | Zoomber | raptor: |
| 22:26:00 | Zoomber | let me show you |
| 22:26:00 | Zoomber | oh, nevermind |
| 22:26:00 | Zoomber | i thought i could reproduce your shooting bug |
| 22:27:00 | Zoomber | it looks like you have maxfps set in that picture though |
| 22:27:00 | Zoomber | the phasers are all a certain distance from each other, where on my screen, they are much closer together |
| 22:27:00 | raptor | i am running 016 |
| 22:27:00 | sam686 | ok, but what logpringf? in Exent.cpp line 443? |
| 22:28:00 | raptor | like logprintf("in onResize()!"); |
| 22:28:00 | raptor | something simple, just to see when it is actually triggered |
| 22:30:00 | Zoomber | Ahh raptor, i found my mouse problem |
| 22:30:00 | sam686 | yes, it is calling onResize about 200 times a second when maximized... |
| 22:31:00 | raptor | oh yikes |
| 22:31:00 | raptor | that's a serious bug |
| 22:31:00 | Zoomber | found out, i was only having slow mouse when in "scale mode" (lets me scale my OS window to whatever i want) |
| 22:32:00 | raptor | yeah sam686, in linux, i only ever get onResize if it is in windowed mode, like it should be... |
| 22:32:00 | raptor | i might have to recompile sdl libraries just for windows... sigh |
| 22:33:00 | sam686 | could be something to do with bitfighter fighting to set the window to 4:3 ratio, but os won't set it to that in windowed maximize mode. (my laptop at 1024 x 576) |
| 22:34:00 | raptor | my laptop is 1600x900 |
| 22:34:00 | sam686 | my desktop lcd is at 1280 x 1024. |
| 22:34:00 | raptor | same ratio as yours |
| 22:35:00 | sam686 | my crt is 17 inch, but the resolution depends (up to 2048 x 1536 at 55 Hertz, or 1280 x 1024 at 86 hertz) |
| 22:35:00 | | ZoombreGiantImac has joined |
| 22:36:00 | ZoombreGiantImac | hi |
| 22:36:00 | raptor | so where does the bug occur? on your CRT? |
| 22:36:00 | sam686 | what bug? my desktop windows computer? |
| 22:36:00 | ZoombreGiantImac | the fps crash bug? |
| 22:36:00 | raptor | the onResize being called a hundred times a second in full screen |
| 22:37:00 | sam686 | that bug is only on my laptop's ubuntu. |
| 22:37:00 | raptor | ohhh... |
| 22:37:00 | raptor | can you tell my the SDL version on your laptop? |
| 22:38:00 | sam686 | what command do i use to find out? |
| 22:38:00 | ZoombreGiantImac | YES |
| 22:38:00 | ZoombreGiantImac | can still access my remote VNC desktop |
| 22:38:00 | raptor | ummm... |
| 22:38:00 | raptor | check your package management systeM? |
| 22:38:00 | raptor | not sure on a non-RPM based system... |
| 22:39:00 | sam686 | there is something called "Ununtu Software Center" that i can try to check... |
| 22:39:00 | ZoombreGiantImac | yes |
| 22:39:00 | sam686 | sdl |
| 22:39:00 | ZoombreGiantImac | thats what you would want to check |
| 22:39:00 | raptor | do you have a graphical front-end for apt? |
| 22:39:00 | ZoombreGiantImac | i think it may be similar, if not is |
| 22:40:00 | ZoombreGiantImac | i tried to install bitfighter there, and couldn't even do that :[ |
| 22:40:00 | sam686 | not sure.. |
| 22:40:00 | ZoombreGiantImac | but with apt-get command, easy to install stuff. leme go to my imac and try that |
| 22:41:00 | sam686 | when i do SDL for installed software, it says i have libsdl-ttf2.0-0 libsdl1.2-dev |
| 22:41:00 | raptor | those would be package names |
| 22:41:00 | raptor | what about the version? |
| 22:41:00 | sam686 | and libsdl1.2debian-pulseaudio libsdl1.2debian |
| 22:41:00 | raptor | should be something like 1.2.12 |
| 22:41:00 | raptor | or 1.2.14 |
| 22:42:00 | raptor | so for instance, on rpm system, the package name is: libSDL-1_2-0 |
| 22:42:00 | sam686 | libsdh-ttf2.0-0 have version 2.0.9-build1 |
| 22:42:00 | raptor | ok, how about the basic libsdl12 |
| 22:42:00 | Zoomber | sudo apt-get install libsdl1.2-dev libsdl1.2debian |
| 22:42:00 | Zoomber | and y to updates |
| 22:43:00 | raptor | on rpm systems, the version is: 1.2.14-40.1 |
| 22:43:00 | raptor | something like that |
| 22:43:00 | raptor | 1.2.14 <- version; 40.1 <- build revision |
| 22:44:00 | raptor | so full package name is something horrible like this: libSDL-1_2-0-1.2.14-40.1 |
| 22:44:00 | sam686 | my libsdl1.2-dev have version 1.2.14-4ubuntu1.1 |
| 22:44:00 | raptor | there we go! |
| 22:44:00 | raptor | 1.2.14 |
| 22:44:00 | raptor | ok |
| 22:45:00 | raptor | mine is the same, but i don't have that resize problem... |
| 22:46:00 | raptor | i wonder if it is your video drivers |
| 22:47:00 | Zoomber | sam686 |
| 22:47:00 | Zoomber | can find that by going to System > administration |
| 22:47:00 | Zoomber | > system monitor |
| 22:49:00 | sam686 | what does the system monitor shows? it just looks at CPU, network, and memory usage. |
| 22:49:00 | Zoomber | you have to click on the "system" tab |
| 22:50:00 | Zoomber | for me it shows my Ubuntu version and kernal/gnome, my hardware "memory and processor"< and available disk space |
| 22:50:00 | raptor | ubuntu can vary wildly between versions FYI |
| 22:51:00 | raptor | any linux distro can.. |
| 22:51:00 | Zoomber | true, and am not sure what version sam has, maybe 10.something |
| 22:52:00 | sam686 | http://208.107.52.15/upload/systemmonitor.png - doesn't seem to show anything about the graphics. |
| 22:53:00 | Zoomber | i see |
| 22:54:00 | Zoomber | an intel atom processor with 1 GB ram will work very very well with windows 7.. |
| 22:54:00 | raptor | i'll be back.. |
| 22:54:00 | Zoomber | raptor: do you use windows 7 at all? curious |
| 22:54:00 | sam686 | though, i don't think it is graphics hardware related, but mostly the SDL version, different dirtrobution (ubuntu) having different version of several things.. |
| 22:55:00 | sam686 | though, i also have windows XP in my laptop as dual boot (with ubuntu) |
| 22:57:00 | Zoomber | might be a good idea to upgrade that windows XP to 7, as windows xp take lots of resources on a small laptop like that, and 7 will be much faster with it..unlike vista |
| 22:58:00 | sam686 | upgrade to windows 7 == pay $$$$$$ |
| 22:58:00 | sam686 | installing and using ubuntu = free |
| 22:59:00 | sam686 | better off getting a new, fast computer.. |
| 22:59:00 | Zoomber | too bad microsoft doesnt allow me to lend you my windows 7 to upgrade.. |
