| *** tpb has joined #freeorion | 00:00 | |
| GeofftheMedio | how does it know what line the crash is at then, if the code changed since the symbols were generated? | 00:00 |
|---|---|---|
| kr0ddnnoMobil | the crash is in GiGi | 00:00 |
| kr0ddnnoMobil | the debug symbols of gigi didnt change | 00:00 |
| GeofftheMedio | missing the point, but nevermind | 00:01 |
| kr0ddnnoMobil | yes, i know what you mean | 00:01 |
| kr0ddnnoMobil | we dont need updated symbols for that | 00:02 |
| kr0ddnnoMobil | cubathy: please download again | 00:02 |
| kr0ddnnoMobil | now it shouldnt crash on line 288, maybe we now get a crash on another line | 00:03 |
| GeofftheMedio | one can only hope... | 00:03 |
| kr0ddnnoMobil | if not, then FO is known to not run on ATI open source and intel drivers :-) | 00:03 |
| kr0ddnnoMobil | for the moment | 00:03 |
| cubathy | we have menu | 00:04 |
| kr0ddnnoMobil | thats the point where my notebook freezes | 00:04 |
| cubathy | i'm into a game - no idea what I'm doing but randomly clicking stuff seems ok :) | 00:05 |
| kr0ddnnoMobil | thats imposible! grrrrr | 00:05 |
| kr0ddnnoMobil | and why does my notebook freeze? | 00:05 |
| cubathy | blame acer - usually a safe bet | 00:05 |
| kr0ddnnoMobil | GeofftheMedio: i disabled the two lines | 00:06 |
| kr0ddnnoMobil | // if (glUseProgramARB) | 00:06 |
| kr0ddnnoMobil | // glUseProgramARB(0); | 00:06 |
| GeofftheMedio | I was just about to ask that | 00:06 |
| GeofftheMedio | hmm. | 00:06 |
| kr0ddnnoMobil | thats what i did some days before when we where diskussing here | 00:06 |
| GeofftheMedio | I know | 00:07 |
| kr0ddnnoMobil | I think on my notebook its a sound card problem | 00:07 |
| GeofftheMedio | I've asked tzlaine about them | 00:07 |
| kr0ddnnoMobil | But i cannot find out | 00:07 |
| GeofftheMedio | he says they're important | 00:07 |
| GeofftheMedio | I think it may only matter when switching between 2d and 3d mode | 00:07 |
| GeofftheMedio | which doesn't happen if just running without combat enabled | 00:07 |
| GeofftheMedio | kroddn: try disabling sound and music in the options? | 00:07 |
| kr0ddnnoMobil | aha | 00:07 |
| kr0ddnnoMobil | i did | 00:07 |
| GeofftheMedio | hmm | 00:07 |
| kr0ddnnoMobil | but it seems to be using sound nevertheless | 00:07 |
| GeofftheMedio | it will still start up and initialize the sound system even if they're turned off | 00:08 |
| GeofftheMedio | but just won't try to output any sounds | 00:08 |
| kr0ddnnoMobil | yes | 00:08 |
| GeofftheMedio | probably there needs to be a --no-sound command line flag | 00:08 |
| kr0ddnnoMobil | As I said, i am not sure about what leads to the freeze | 00:08 |
| kr0ddnnoMobil | The systems lags for some seconds and then freezes | 00:09 |
| *** kr0ddnnoMobil has quit IRC | 00:10 | |
| *** kroddnoMobil has joined #freeorion | 00:10 | |
| *** ChanServ sets mode: +o kroddnoMobil | 00:10 | |
| kroddnoMobil | I'll do experiments tomorrow, no its time to sleep | 00:11 |
| GeofftheMedio | night | 00:11 |
| cubathy | thanks a lot | 00:12 |
| GeofftheMedio | cubathy: could you try running "freeorion --test-3d-combat" and then starting a game with 10 stars and 3 AIs, and then hitting turn a few times until a combat happens? | 00:12 |
| kroddnoMobil | cubathy: its would be interesting to know if you can play a game without crash/freeze | 00:12 |
| GeofftheMedio | kroddn: I'm assuming that cubathy can hit turn a few times without a crash, since there was no mention of doing so causing a crash | 00:13 |
| cubathy | yep, i can, running 3d now... | 00:13 |
| cubathy | ok - it came up resolving combat or something, and now is showing a half empty screen, just see buttons "Turn" and "End Combat" | 00:15 |
| cubathy | doesn't seem to let me out when I click end combat | 00:16 |
| GeofftheMedio | you can zoom with the scrollwheel, double click to recenter, and rotate the view with middle-button and mouse move | 00:16 |
| cubathy | no trace | 00:16 |
| GeofftheMedio | it's not supposed to let you out of combat... work in progress | 00:16 |
| cubathy | it doesn't show anything to rotate/zoom | 00:16 |
| GeofftheMedio | no background, no star, no wormhole swirly things? | 00:17 |
| cubathy | noope | 00:17 |
| GeofftheMedio | hmm | 00:18 |
| GeofftheMedio | odd | 00:18 |
| GeofftheMedio | haven't tried it in a few days though, so it might be normal | 00:18 |
| GeofftheMedio | not a major issue at the moment either way, as it's still being set up | 00:18 |
| cubathy | ok, I'll try and debug it a bit tomorrow if I can get my build working | 00:21 |
| cubathy | is it maybe because of the shader language API change? | 00:26 |
| GeofftheMedio | there's nothing fancy happening on the line in question... it's just telling the driver to disable all shaders | 00:26 |
| GeofftheMedio | glUseProgram(0) should always work on GL 2.0 or greater, as far as I know | 00:28 |
| cubathy | I'd presume glXGetProcAddress((const GLubyte* )"glUseProgramARB") is NULL | 00:28 |
| cubathy | thats what I'm saying - should it use glUseProgram instead of glUSeProgramARB | 00:28 |
| GeofftheMedio | there's a if () before the function pointer is called. if it was null, it wouldn't be called | 00:29 |
| cubathy | ...true :) | 00:29 |
| GeofftheMedio | you can try removing the ARB on those lines | 00:29 |
| GeofftheMedio | I think I suggested that to kroddn, and it didn't help | 00:29 |
| cubathy | I'll play around with it tomorrow, my build still isn't working | 00:30 |
| cubathy | anyway. thanks for all the help | 00:31 |
| cubathy | night | 00:31 |
| *** cubathy has quit IRC | 00:36 | |
| *** mithro has joined #freeorion | 00:51 | |
| *** bernardh has quit IRC | 01:18 | |
| *** tomboy65 has quit IRC | 02:32 | |
| *** STalKer-X has quit IRC | 03:20 | |
| *** STalKer-Y has joined #freeorion | 03:20 | |
| CIA-76 | FreeOrion: tzlaine * r3565 /trunk/FreeOrion/ (UI/CombatWnd.cpp combat/OpenSteer/CombatShip.cpp): | 03:27 |
| CIA-76 | FreeOrion: Three fixes to bugs uncovered by Valgrind: | 03:27 |
| CIA-76 | FreeOrion: - Initialized CombatShip::m_enter_starlane_start_turn. | 03:27 |
| CIA-76 | FreeOrion: - Initialized CombatWnd::m_star_brightness_factor. | 03:27 |
| CIA-76 | FreeOrion: - In CombatWnd::CombatTurnUpdate(), moved the call to | 03:27 |
| CIA-76 | FreeOrion: PathingEngine::TurnStarted() outside the loop it was mistakenly placed in. | 03:27 |
| *** kr0ddnoMobil has joined #freeorion | 03:44 | |
| *** ChanServ sets mode: +o kr0ddnoMobil | 03:44 | |
| *** kroddnoMobil has quit IRC | 03:48 | |
| CIA-76 | FreeOrion: tzlaine * r3566 /trunk/FreeOrion/combat/OpenSteer/PathingEngine.cpp: Fixed the PathingEngine dtor, which was intermittently crashing the AI clients. | 04:02 |
| CIA-76 | FreeOrion: tzlaine * r3567 /trunk/FreeOrion/UI/CombatWnd.cpp: Removed certain bits of testing cruft from CombatWnd. | 04:03 |
| CIA-76 | FreeOrion: tzlaine * r3568 /trunk/FreeOrion/server/ServerApp.cpp: | 04:19 |
| CIA-76 | FreeOrion: Increased the minumum number of server-side PathingEngine update iterations, to | 04:19 |
| CIA-76 | FreeOrion: increase agreement between the client and server object positions. | 04:19 |
| *** aleksil has joined #freeorion | 05:56 | |
| kr0ddnoMobil | GeofftheMedio: could it be possible that calling glUseProgram(0) is possible only after a prior call to it with a valid program? I don't know exactly why tzlaine calls it, as in the whole source there is never an own program added by glUseProgram (or an according ARB funciton) | 07:19 |
| *** enigmatic has joined #freeorion | 07:23 | |
| GeofftheMedio | I don't see any mention of that anywhere | 07:28 |
| kr0ddnoMobil | Maybe in ogre itself? Would you ask tzlaine again if we could skip this code for the current 0.3.14? | 07:34 |
| kr0ddnoMobil | it would make have sense to call it if a subprogram would have called it with an own programm, but that would be dirty too | 07:35 |
| kr0ddnoMobil | as far as I understand what I read you would use glUseProgramARB(0) only after you have added your own Programm before | 07:36 |
| kr0ddnoMobil | and that has to be done by the code that adds it, not by this function Enter2DMode() | 07:36 |
| GeofftheMedio | it's much simpler to have all rendering code not need to worry about cleaning up its shader programs, and instead have every bit of code set up the rendering state for itself | 07:40 |
| GeofftheMedio | enter2dmode doesn't assume anything about the previous rendering state, so calls useprogram(0) to set things up for 2D rendering | 07:41 |
| GeofftheMedio | this isn't an ogre issue, unless something very weird is going on | 07:41 |
| GeofftheMedio | what happens if you remove all this code: | 07:44 |
| GeofftheMedio | typedef void (*BindBufferARBFn)(GLenum, GLuint); | 07:44 |
| GeofftheMedio | #if OGRE_PLATFORM == OGRE_PLATFORM_WIN32 | 07:44 |
| GeofftheMedio | BindBufferARBFn glBindBufferARB = (BindBufferARBFn)wglGetProcAddress("glBindBufferARB"); | 07:44 |
| GeofftheMedio | #elif OGRE_PLATFORM == OGRE_PLATFORM_APPLE | 07:44 |
| GeofftheMedio | BindBufferARBFn glBindBufferARB = (BindBufferARBFn)aglGetProcAddress("_glBindBufferARB"); | 07:44 |
| GeofftheMedio | #else | 07:44 |
| GeofftheMedio | BindBufferARBFn glBindBufferARB = (BindBufferARBFn)glXGetProcAddress((const GLubyte* )"glBindBufferARB"); | 07:44 |
| GeofftheMedio | #endif | 07:44 |
| GeofftheMedio | if (glBindBufferARB) { | 07:44 |
| GeofftheMedio | glBindBufferARB(GL_ARRAY_BUFFER_ARB, 0); | 07:44 |
| GeofftheMedio | glBindBufferARB(GL_ELEMENT_ARRAY_BUFFER_ARB, 0); | 07:45 |
| GeofftheMedio | glBindBufferARB(GL_PIXEL_PACK_BUFFER_ARB, 0); | 07:45 |
| GeofftheMedio | glBindBufferARB(GL_PIXEL_UNPACK_BUFFER_ARB, 0); | 07:45 |
| GeofftheMedio | } | 07:45 |
| GeofftheMedio | typedef void (*UseProgramARBFn)(GLuint); | 07:45 |
| GeofftheMedio | #if OGRE_PLATFORM == OGRE_PLATFORM_WIN32 | 07:45 |
| GeofftheMedio | UseProgramARBFn glUseProgramARB = (UseProgramARBFn)wglGetProcAddress("glUseProgramARB"); | 07:45 |
| GeofftheMedio | #elif OGRE_PLATFORM == OGRE_PLATFORM_APPLE | 07:45 |
| GeofftheMedio | UseProgramARBFn glUseProgramARB = (UseProgramARBFn)aglGetProcAddress("_glUseProgramARB"); | 07:45 |
| GeofftheMedio | #else | 07:45 |
| GeofftheMedio | UseProgramARBFn glUseProgramARB = (UseProgramARBFn)glXGetProcAddress((const GLubyte* )"glUseProgramARB"); | 07:45 |
| GeofftheMedio | #endif | 07:45 |
| GeofftheMedio | if (glUseProgramARB) | 07:45 |
| GeofftheMedio | glUseProgramARB(0); | 07:45 |
| GeofftheMedio | and replace with | 07:45 |
| GeofftheMedio | glBindBuffer(GL_ARRAY_BUFFER_ARB, 0); | 07:45 |
| GeofftheMedio | glBindBuffer(GL_ELEMENT_ARRAY_BUFFER_ARB, 0); | 07:45 |
| GeofftheMedio | glBindBuffer(GL_PIXEL_PACK_BUFFER_ARB, 0); | 07:45 |
| GeofftheMedio | glBindBuffer(GL_PIXEL_UNPACK_BUFFER_ARB, 0); | 07:45 |
| GeofftheMedio | glUseProgram(0); | 07:45 |
| GeofftheMedio | (maybe remove the _ARB in those GL constants too) | 07:45 |
| kr0ddnoMobil | those calls without _ARB are not existing | 07:54 |
| GeofftheMedio | ok, try adding | 07:54 |
| GeofftheMedio | # include <GL/gl.h> | 07:54 |
| GeofftheMedio | # include <GL/glu.h> | 07:54 |
| kr0ddnoMobil | damn, i should have a look at those things before i post ^^ | 07:55 |
| kr0ddnoMobil | but that doesnt change anything | 07:56 |
| kr0ddnoMobil | glBindBuffer is available only if the GL version is 1.5 or greater. | 07:56 |
| GeofftheMedio | and glUseProgram is only available in GL 2.0 or greater | 07:56 |
| GeofftheMedio | what gl version are you running? | 07:58 |
| kr0ddnoMobil | i am RUNNING gl 1.5 | 07:58 |
| kr0ddnoMobil | but on the system I compile I use the dummy framework | 07:58 |
| kr0ddnoMobil | hm, maybe i sould install a winxp on my notebook to see if win32 version of FO is running properly | 08:00 |
| GeofftheMedio | if you want... | 08:00 |
| kr0ddnoMobil | to much work ^^ | 08:10 |
| GeofftheMedio | could you have a look what ogre does to get gluseprogram so that it can work even if the gl version is less than 2.0 ? | 08:10 |
| *** tomboy64 is now known as MacBash | 08:13 | |
| kr0ddnoMobil | GeofftheMedio: seems that they do it the same way as tzlaine did it | 08:13 |
| kr0ddnoMobil | (glUseProgram = (PFNGLUSEPROGRAMPROC)glewGetProcAddress((const GLubyte*)"glUseProgram")) | 08:14 |
| kr0ddnoMobil | hm, they use glewXXX | 08:14 |
| *** tomboy65 has joined #freeorion | 08:15 | |
| GeofftheMedio | I think glew is win32 only? | 08:15 |
| kr0ddnoMobil | glew is a extendet framework, isnt it� | 08:16 |
| GeofftheMedio | I seem to recall "GL Extension Wrangler" | 08:16 |
| *** MacBash has quit IRC | 08:19 | |
| kr0ddnoMobil | GeofftheMedio: i changed the call to glXGetProcAddress((const GLubyte*)"glUseProgram") | 08:20 |
| kr0ddnoMobil | no crash | 08:20 |
| *** kroddnoMobil has joined #freeorion | 08:22 | |
| *** ChanServ sets mode: +o kroddnoMobil | 08:22 | |
| kroddnoMobil | GeofftheMedio: guess what... | 08:24 |
| GeofftheMedio | useful for does glXGetProcAddress((const GLubyte*)"glUseProgram") return a NULL pointer, though? | 08:25 |
| GeofftheMedio | (ignore the "useful for") | 08:25 |
| kroddnoMobil | I think glXGetProcAddress((const GLubyte*)"glUseProgram") returns NULL, so no crash happens. Bzt then I encounter that problem that my notebook encounters | 08:25 |
| *** kr0ddnoMobil has quit IRC | 08:25 | |
| kroddnoMobil | Hm, it returns a valid object | 08:27 |
| kroddnoMobil | $1 = (GG::OgreGUI::UseProgramARBFn) 0x99a4a0 <glUseProgramObjectARB> | 08:27 |
| GeofftheMedio | isn't the other problem audio related? | 08:27 |
| kroddnoMobil | i do not know | 08:28 |
| kroddnoMobil | The main menu comes up, and then it freezes after some seconds | 08:28 |
| kroddnoMobil | no errors, no output, no syslog entries | 08:28 |
| GeofftheMedio | try commenting out all the contents of all the functions in Sound.cpp | 08:29 |
| kroddnoMobil | Apr 30 10:20:58 natter kernel: [519632.517831] evdev.c(EVIOCGBIT): Suspicious buffer size 511, limiting output to 64 bytes. See http: | 08:29 |
| kroddnoMobil | hm | 08:29 |
| kroddnoMobil | didn't realize that syslog error before | 08:30 |
| kroddnoMobil | Apr 30 10:30:28 natter pulseaudio[1622]: ratelimit.c: 136 events suppressed | 08:30 |
| kroddnoMobil | so my intention seems to be right | 08:31 |
| GeofftheMedio | ? | 08:31 |
| kroddnoMobil | pulseaudio is audio ^^ | 08:31 |
| GeofftheMedio | ah. | 08:32 |
| GeofftheMedio | so yeah: comment out code in Sound.cpp | 08:32 |
| GeofftheMedio | if that works, we can add a --no-sound command line flag as well | 08:32 |
| kroddnoMobil | maybe you can look at http://userweb.kernel.org/~dtor/eviocgbit-bug.html | 08:37 |
| tpb | <http://ln-s.net/6-UF> (at userweb.kernel.org) | 08:37 |
| kroddnoMobil | this exolains something about that bug | 08:37 |
| kroddnoMobil | and maybe we can find the problem | 08:37 |
| kroddnoMobil | if it is not a bug in openAL | 08:38 |
| GeofftheMedio | I don't find ioctl in sound.cpp | 08:39 |
| GeofftheMedio | is that call within openal? | 08:41 |
| *** kr0ddnoMobil has joined #freeorion | 08:43 | |
| *** ChanServ sets mode: +o kr0ddnoMobil | 08:43 | |
| kr0ddnoMobil | guess what | 08:44 |
| GeofftheMedio | ...? | 08:44 |
| kr0ddnoMobil | I could start a game | 08:45 |
| GeofftheMedio | after doing what? | 08:45 |
| kr0ddnoMobil | But as soon as the main screen appeard, it froze again | 08:45 |
| *** kroddnoMobil has quit IRC | 08:46 | |
| kr0ddnoMobil | still entries in my syslog | 08:46 |
| GeofftheMedio | after what changes? | 08:46 |
| kr0ddnoMobil | after disabling the functions in sound.cpp | 08:46 |
| kroddn | I'll try to find out what this syslog messages come from | 08:49 |
| kroddn | maybe its not only sound | 08:49 |
| GeofftheMedio | can you pastebin your sound.cpp | 08:49 |
| GeofftheMedio | ? | 08:49 |
| kroddn | is there other places where sound is initialized? | 08:49 |
| GeofftheMedio | I don't think so | 08:50 |
| planetmaker | just a comment from the off: pulseaudio is known to cause freezes / hangs on some systems | 08:51 |
| planetmaker | especially ubuntu | 08:51 |
| *** kr0ddnoMobil has quit IRC | 08:51 | |
| GeofftheMedio | are you sure the evdev.c thing is sound related? | 08:52 |
| kroddn | Okay, this was my fault... i commented out the wrong way | 08:52 |
| GeofftheMedio | I'm finding some references to it that appears to be OIS related | 08:52 |
| kroddn | GeofftheMedio: there are other log lines like that: Apr 30 10:30:28 natter pulseaudio[1622]: ratelimit.c: 136 events suppressed | 08:53 |
| GeofftheMedio | hmm | 08:53 |
| kroddn | thats why I thought of sound | 08:53 |
| kroddn | maybe its ois | 08:53 |
| GeofftheMedio | again: could you pastebin your sound.cpp ? | 08:53 |
| planetmaker | http://paste.openttd.org/225721 <-- explanation | 08:54 |
| tpb | Title: OpenTTD PasteBin (at paste.openttd.org) | 08:54 |
| kroddn | http://freeorion.psitronic.de/src/Sound.cpp | 08:55 |
| GeofftheMedio | you still need to return 0 from RefillBuffer | 08:57 |
| *** bernardh has joined #freeorion | 09:07 | |
| *** kroddnoMobil has joined #freeorion | 09:35 | |
| *** ChanServ sets mode: +o kroddnoMobil | 09:35 | |
| *** kroddnoMobil has quit IRC | 10:15 | |
| *** kroddnoMobil has joined #freeorion | 10:28 | |
| *** ChanServ sets mode: +o kroddnoMobil | 10:28 | |
| kroddnoMobil | GeofftheMedio: this all didn't make a difference | 10:30 |
| kroddnoMobil | Where did you find the infos according OIS ? | 10:30 |
| GeofftheMedio | http://www.orxonox.net/changeset/5695/code/trunk/src | 10:31 |
| GeofftheMedio | search page for ioctl | 10:31 |
| tpb | Title: Changeset 5695 for code/trunk/src – Orxonox (at www.orxonox.net) | 10:31 |
| kroddnoMobil | there are a lots of them :-) | 10:35 |
| GeofftheMedio | first | 10:37 |
| kroddnoMobil | lenny has ois 0.99 | 10:39 |
| kroddnoMobil | i will try upgrading to 1.2 | 10:39 |
| kroddnoMobil | its easy to upgrade that, so I'll give it a try | 10:40 |
| kroddnoMobil | thats not fixed in 1.2.0 :-( | 10:56 |
| *** kroddnoMobil has quit IRC | 11:12 | |
| *** kroddnoMobil has joined #freeorion | 11:13 | |
| *** ChanServ sets mode: +o kroddnoMobil | 11:13 | |
| kroddnoMobil | GeofftheMedio: I now fixed the errors that syslog reported by fixing the code in ois itself | 11:16 |
| GeofftheMedio | I'm surprised that worked | 11:17 |
| kroddnoMobil | But it still freezes, although I am now able to move the mouse | 11:17 |
| kroddnoMobil | Means: its not freezing completely, the mouse will move every 3-5 seconds | 11:17 |
| GeofftheMedio | it would probably be good to post a bit about the changes you're making | 11:17 |
| GeofftheMedio | so google can index them, and anybody else trying to compile freeorion can duplicate them | 11:18 |
| planetmaker | kroddnoMobil, is that SDL + pulseaudio you're using? | 11:18 |
| kroddnoMobil | no, we are using ogre instead of SDL | 11:19 |
| planetmaker | http://paste.openttd.org/225721 <-- did you check wether it still might be related to the pulseaudio behaviour described there? Wether that exactly might be the difference between "now" and "earlier"? | 11:21 |
| tpb | Title: OpenTTD PasteBin (at paste.openttd.org) | 11:21 |
| planetmaker | it's mainly a pulseaudio thing - so it might show with other front-ends, too | 11:21 |
| planetmaker | basically some change in handling the scheduling of audio changed... | 11:22 |
| planetmaker | s/changed/was introduced/ | 11:23 |
| kroddnoMobil | i'll try tha | 11:24 |
| kroddnoMobil | Saying goodby, until later ;-) | 11:24 |
| planetmaker | enjoy | 11:25 |
| planetmaker | it's just an idea and might be completely off. But I see a bug report about that every week. | 11:25 |
| *** kroddnoMobil has quit IRC | 11:25 | |
| *** kroddnoMobil has joined #freeorion | 11:33 | |
| *** ChanServ sets mode: +o kroddnoMobil | 11:33 | |
| kroddnoMobil | did not work | 11:34 |
| kroddnoMobil | One thing is possitive - we now know that FO can run on ATI or Intel systems with open source drivers | 11:36 |
| kroddnoMobil | planetmaker: which OS do you have? | 11:36 |
| kroddnoMobil | Hm, i could try running FO on a Knoppix | 11:37 |
| planetmaker | kroddnoMobil, the only one I can run FO on is OSX. But I know that bug from yeah.. OpenTTD forums and bug tracker where it shows with ubuntu and pulseaudio | 11:39 |
| planetmaker | my linux has too old OpenGL unfortunately. And... I shouldn't mess too much with this system ;-) | 11:39 |
| *** mithro has quit IRC | 11:40 | |
| kroddnoMobil | Do you have a linux? | 11:41 |
| kroddnoMobil | The version I provide does not need any messing up :-) | 11:41 |
| kroddnoMobil | just extract in /tmp and run ./freeorion | 11:41 |
| planetmaker | I do have linux here at university | 11:41 |
| planetmaker | Yes... but opengl? | 11:41 |
| kroddnoMobil | but maybe without opengl :-) | 11:42 |
| kroddnoMobil | give it a try | 11:42 |
| kroddnoMobil | maybe a software renderer is available | 11:42 |
| planetmaker | I tried to compile the last year's stable - which failed due to that reason | 11:42 |
| kroddnoMobil | It won't be playable, but maybe a positive feedback | 11:42 |
| planetmaker | it will be on suse 11.0 | 11:43 |
| planetmaker | can you give me the svn URL? | 11:45 |
| planetmaker | nvm ... https://freeorion.svn.sourceforge.net/svnroot/freeorion/trunk | 11:47 |
| tpb | <http://ln-s.net/1MR4> (at freeorion.svn.sourceforge.net) | 11:47 |
| kroddnoMobil | http://freeorion.psitronic.de/download/nightly/ | 11:49 |
| tpb | Title: Index of /download/nightly (at freeorion.psitronic.de) | 11:49 |
| kroddnoMobil | but wait some minutes, i'll release a new one | 11:50 |
| planetmaker | just highlight me | 11:50 |
| kroddnoMobil | And I'll try knoppix | 11:53 |
| kroddnoMobil | planetmaker | 12:00 |
| kroddnoMobil | http://freeorion.psitronic.de/download/nightly/ | 12:00 |
| tpb | Title: Index of /download/nightly (at freeorion.psitronic.de) | 12:00 |
| kroddnoMobil | download the freeorion_rev3568_i386_static.tar.gz | 12:00 |
| kroddnoMobil | and I'll boot knoppix | 12:00 |
| *** kroddnoMobil has quit IRC | 12:00 | |
| planetmaker | seems like an insuccessful re-boot? | 12:16 |
| planetmaker | anyway... the provided version seems to run fine. | 12:17 |
| planetmaker | Mostly that is as I can't figure out how to send ships around and the sensor area around the HW doesn't show either. | 12:17 |
| planetmaker | But it's not more sluggish than I'd expect on this machine | 12:17 |
| planetmaker | (and most things are sluggish here :-P ) | 12:17 |
| *** kroddnoMobil has joined #freeorion | 12:22 | |
| *** ChanServ sets mode: +o kroddnoMobil | 12:22 | |
| planetmaker | welcome back :-) | 12:25 |
| planetmaker | <planetmaker> anyway... the provided version seems to run fine. | 12:25 |
| planetmaker | <planetmaker> Mostly that is as I can't figure out how to send ships around and the sensor area around the HW doesn't show either. | 12:25 |
| planetmaker | <planetmaker> But it's not more sluggish than I'd expect on this machine | 12:25 |
| planetmaker | <planetmaker> (and most things are sluggish here :-P ) | 12:25 |
| planetmaker | in turn two the ships seem accessible, too. But maybe I was just stupid | 12:25 |
| kroddnoMobil | There are two fleets | 12:26 |
| planetmaker | well. yes | 12:26 |
| kroddnoMobil | one with scouts and colony ship | 12:26 |
| planetmaker | both are accessible | 12:26 |
| kroddnoMobil | you can drag&drop the scouts into the new fleet field and then send each scout to a system | 12:26 |
| kroddnoMobil | to a unexplored system i mean | 12:26 |
| kroddnoMobil | If i remember right you have to left-click on the fleet and then left click on the target syste, | 12:27 |
| kroddnoMobil | Knoppix doesnt run properly, i couldnt test :( | 12:27 |
| planetmaker | yes.. actually there's one problem: I don't see the starlanes | 12:28 |
| kroddnoMobil | that can be enabled in the options | 12:28 |
| GeofftheMedio | uh... no it can't | 12:29 |
| GeofftheMedio | more so, it can't be disabled | 12:29 |
| kroddnoMobil | ah, i understand^^ | 12:29 |
| GeofftheMedio | you should always be able to see lanes | 12:29 |
| GeofftheMedio | maybe this is a side effect of not calling glUseProgram(0) ? | 12:29 |
| kroddnoMobil | planetmaker: can you paste the output from glxinfo ? | 12:29 |
| GeofftheMedio | planetmaker: if you open the options screen and to to the galaxy map tab... | 12:29 |
| GeofftheMedio | ... is fog of war scanlines enabled? | 12:30 |
| GeofftheMedio | also, you CAN enable the visibility circles on that page | 12:30 |
| planetmaker | http://paste.openttd.org/225722 | 12:30 |
| tpb | Title: OpenTTD PasteBin (at paste.openttd.org) | 12:30 |
| planetmaker | GeofftheMedio, how can I do that? Options? | 12:30 |
| GeofftheMedio | menu, options | 12:30 |
| kroddnoMobil | open GL 1.4 | 12:30 |
| planetmaker | yes. That's what I said: old crappy opengl :-) | 12:31 |
| GeofftheMedio | if GL is that low, various rendering options will have been disabled | 12:31 |
| GeofftheMedio | such as the scanlines | 12:31 |
| planetmaker | scan radius found. | 12:34 |
| kroddnoMobil | If i remember right i didnt have starlanes before my crashes too | 12:35 |
| *** bernardh has quit IRC | 12:35 | |
| planetmaker | hm... it was no good idea to try to enable galaxy gas: | 12:35 |
| planetmaker | ./freeorion: line 143: 14693 Segmentation fault ${FO_TRACE} ${FO_GDB_ARGS} ./freeorion --resource-dir ./default $@ | 12:35 |
| kroddnoMobil | as expected :-) | 12:35 |
| kroddnoMobil | that needs opengl 1.5 | 12:35 |
| planetmaker | uhm... expected crash? | 12:35 |
| kroddnoMobil | yes ^^ | 12:36 |
| planetmaker | I'd expect usually an ifdef ;-) | 12:36 |
| kroddnoMobil | not in an unstable alpha version ^^ | 12:36 |
| planetmaker | :-P ok | 12:36 |
| GeofftheMedio | it's difficult to use a compile time switch to disable functionality that can only be tested for by getting a string on the machine running the code | 12:37 |
| planetmaker | well, yes, that can't, I agree | 12:37 |
| kroddnoMobil | planetmaker: http://freeorion.psitronic.de/download/nightly/freeorion/application/freeorion <- Exe without my "Fix" | 12:38 |
| tpb | <http://ln-s.net/6:CA> (at freeorion.psitronic.de) | 12:38 |
| planetmaker | but something like "fool! get a better machine and library support! :P " ;-) | 12:38 |
| GeofftheMedio | it disabled the gas rendering based on your gl version | 12:38 |
| kroddnoMobil | thats work in progress i think - in earlier versions fo didn even start when havin open gl less than 1.5 | 12:38 |
| GeofftheMedio | the only foolish part was turning it back on | 12:38 |
| planetmaker | hehe | 12:39 |
| planetmaker | user usually are way more stupid than one can expect :-P | 12:39 |
| kroddnoMobil | are you downloading? | 12:40 |
| planetmaker | already done | 12:40 |
| planetmaker | gigabit backbone is fast :-P | 12:40 |
| kroddnoMobil | thats the advantage of compiling on a root server ^^ | 12:41 |
| kroddnoMobil | no slow upload needed | 12:41 |
| planetmaker | yeah. And being attached to the backbone helps with download :-P | 12:41 |
| kroddnoMobil | grml | 12:41 |
| kroddnoMobil | In �freeorion_rev3553_i386_static.tar.gz� speichern. | 12:41 |
| kroddnoMobil | 68% [===================================================================> ] 115.407.030 243K/s ETA 3m 28s | 12:41 |
| kroddnoMobil | 243 kb... | 12:41 |
| planetmaker | hm... that binary is missing libCg.so | 12:43 |
| kroddnoMobil | ? | 12:43 |
| kroddnoMobil | no ^^ | 12:43 |
| kroddnoMobil | you replaced the wrong file i would suppose | 12:43 |
| kroddnoMobil | replace application/freeorion and run ./freeorion | 12:43 |
| kroddnoMobil | libGc is inside application/lib | 12:44 |
| planetmaker | yes. I didn't know I needed to replace that one. | 12:44 |
| planetmaker | Well... : | 12:44 |
| planetmaker | ./freeorion: line 143: 14920 Segmentation fault ${FO_TRACE} ${FO_GDB_ARGS} ./freeorion --resource-dir ./default $@ | 12:44 |
| planetmaker | we've seen that ;-) | 12:45 |
| kroddnoMobil | can you try ./freeorion --gdb | 12:45 |
| kroddnoMobil | and then type start | 12:45 |
| kroddnoMobil | it should crash on something line 288 :-) | 12:45 |
| planetmaker | no idea | 12:46 |
| planetmaker | ---Type <return> to continue, or q <return> to quit--- | 12:47 |
| planetmaker | (no debugging symbols found) | 12:47 |
| planetmaker | Program received signal SIGSEGV, Segmentation fault. | 12:47 |
| planetmaker | [Switching to Thread 0xb743d6d0 (LWP 14956)] | 12:47 |
| planetmaker | 0x000010c9 in ?? () | 12:47 |
| planetmaker | no debugging symbols found btw | 12:47 |
| kroddnoMobil | ah okay, cannot work without debugging symbols of course | 12:47 |
| kroddnoMobil | but surely its the crash we were working on | 12:47 |
| planetmaker | it's exactly the same as I got when enabling the gas in the galaxy... | 12:48 |
| GeofftheMedio | you might want to add some std::cout << "About to execute line 228" << std::endl; in there | 12:48 |
| GeofftheMedio | you might want to delete your config.xml then | 12:48 |
| kroddnoMobil | remove the whole $HOME/.freeorion | 12:49 |
| planetmaker | whom? Me? | 12:49 |
| kroddnoMobil | yes, you | 12:49 |
| GeofftheMedio | "do just what he told you to" | 12:49 |
| planetmaker | ok... and then? | 12:50 |
| kroddnoMobil | the run again | 12:52 |
| planetmaker | gdb or not? | 12:52 |
| kroddnoMobil | gdb | 12:52 |
| kroddnoMobil | did you download the debug symbols? | 12:52 |
| planetmaker | wait... new binary? | 12:53 |
| kroddnoMobil | http://freeorion.psitronic.de/download/nightly/freeorion_rev3553_i386_static.tar.gz | 12:53 |
| tpb | <http://ln-s.net/6:M5> (at freeorion.psitronic.de) | 12:53 |
| kroddnoMobil | �h stop | 12:53 |
| kroddnoMobil | http://freeorion.psitronic.de/download/nightly/freeorion_rev3553_debugsymbols.tar.gz | 12:53 |
| tpb | <http://ln-s.net/6:MC> (at freeorion.psitronic.de) | 12:53 |
| kroddnoMobil | be sure to be inside the right dir when extracting | 12:55 |
| planetmaker | hm, yeah. Are you aware that those debug symbols are for another version? | 12:55 |
| kroddnoMobil | I think its freeorion main installation dir where you have to extract it to | 12:55 |
| planetmaker | r3553 vs r3568 ? | 12:55 |
| kroddnoMobil | the exe you downloaded is from3553 | 12:55 |
| kroddnoMobil | its an older version without my recent patches | 12:56 |
| planetmaker | ah, damn... need to re-download that :-) | 12:56 |
| kroddnoMobil | why? | 12:56 |
| planetmaker | I thought it was a new whole install the debug thingy and deleted the old ;-) | 12:56 |
| kroddnoMobil | aha | 12:56 |
| kroddnoMobil | then download both http://freeorion.psitronic.de/download/nightly/freeorion_rev3553_i386_static.tar.gz and http://freeorion.psitronic.de/download/nightly/freeorion_rev3553_debugsymbols.tar.gz | 12:56 |
| tpb | <http://ln-s.net/6:M5> (at freeorion.psitronic.de) | 12:56 |
| planetmaker | Breakpoint 1 at 0x853d871: file client/human/chmain.cpp, line 61. | 13:02 |
| planetmaker | Starting program: /home/ingo/freeorion/freeorion/application/freeorion --resource-dir ./default | 13:02 |
| planetmaker | [Thread debugging using libthread_db enabled] | 13:02 |
| planetmaker | [New Thread 0xb74116d0 (LWP 15206)] | 13:02 |
| planetmaker | [Switching to Thread 0xb74116d0 (LWP 15206)] | 13:02 |
| planetmaker | main (argc=3, argv=0xbfbfbf54) at client/human/chmain.cpp:61 | 13:02 |
| planetmaker | 61 client/human/chmain.cpp: No such file or directory. | 13:02 |
| planetmaker | in client/human/chmain.cpp | 13:02 |
| planetmaker | should I go on from there? call a backtrace? ...? | 13:11 |
| kroddnoMobil | yes | 13:18 |
| kroddnoMobil | bt | 13:18 |
| planetmaker | #0 main (argc=3, argv=0xbfbfbf54) at client/human/chmain.cpp:61 | 13:21 |
| planetmaker | that's all | 13:21 |
| planetmaker | continue? | 13:22 |
| kroddn | thats a strange position for the crash | 13:39 |
| kroddn | did you remove $HOME/.freeorion/ completely? | 13:39 |
| planetmaker | yes | 13:39 |
| planetmaker | rm -rf ~/.freeorion should do the job | 13:40 |
| planetmaker | actually... the dir wasn't re-created by that point | 13:41 |
| planetmaker | it doesn't exist as of now | 13:41 |
| planetmaker | and... when I continue, I get that: | 13:46 |
| planetmaker | http://paste.openttd.org/ | 13:46 |
| planetmaker | which might rather be the crash | 13:46 |
| tpb | Title: OpenTTD PasteBin (at paste.openttd.org) | 13:46 |
| planetmaker | so it doesn't find a file - but that's non-fatal | 13:46 |
| planetmaker | but that's some kind of breakpoint | 13:47 |
| GeofftheMedio | missing link number | 13:48 |
| planetmaker | I don't understand | 13:50 |
| GeofftheMedio | normally there's a number after the / in a paste ...? | 13:50 |
| planetmaker | which slash? | 13:51 |
| planetmaker | oh. :-) | 13:52 |
| planetmaker | http://paste.openttd.org/225725 | 13:52 |
| tpb | Title: OpenTTD PasteBin (at paste.openttd.org) | 13:52 |
| planetmaker | The link I posted :-) | 13:52 |
| planetmaker | sorry | 13:52 |
| kroddn | The point where your crash is is strange | 14:10 |
| kroddn | never had that before | 14:10 |
| kroddn | Is that the end of the backtrace? | 14:11 |
| kroddn | please post the highest #XX position, that may be inside the file GG/Ogre/OgreGUI.cpp | 14:12 |
| kroddn | #1 0x085235d9 in HumanClientApp::Enter2DMode (this=0xbfe8f390) at client/human/HumanClientApp.cpp:471 | 14:12 |
| kroddn | GeofftheMedio: we should add to the code that it shows which GG revision was used | 14:13 |
| kroddn | I now don't know what i used with 3553 | 14:14 |
| kroddn | 469 void HumanClientApp::Enter2DMode() | 14:16 |
| kroddn | 470 { | 14:16 |
| kroddn | 471 OgreGUI::Enter2DMode(); | 14:16 |
| GeofftheMedio | gigi hasn't changed in a week | 14:16 |
| kroddn | GeofftheMedio: right, but it would be a good improvement | 14:16 |
| GeofftheMedio | hmm | 14:17 |
| kroddn | For some reason the debug symbols do not contain gigi debug | 14:17 |
| GeofftheMedio | might be difficult to do in a portable way... | 14:17 |
| GeofftheMedio | that doesn't depend on gigi being located in a particular place when building freeorion | 14:17 |
| kroddn | ? | 14:17 |
| GeofftheMedio | unless you convinced tzlaine to add a Version() function to GG | 14:17 |
| kroddn | i mean the SVN revision | 14:18 |
| GeofftheMedio | I know | 14:18 |
| kroddn | like it is shown for freeorion | 14:18 |
| GeofftheMedio | I know | 14:18 |
| GeofftheMedio | but how do you get the SVN version? | 14:18 |
| kroddn | how does FO do it? | 14:18 |
| GeofftheMedio | SVN command and rewriting version.cpp | 14:18 |
| GeofftheMedio | only actually works for the scons build though | 14:18 |
| planetmaker | usually you'd do it in a configure script ^ | 14:18 |
| kroddn | so why don't do that for gigi | 14:18 |
| planetmaker | kroddn, that's the full backtrace I get, yes | 14:19 |
| kroddn | planetmaker: the part for gigi is missing | 14:19 |
| GeofftheMedio | as I suggested, you could convince tzlaine to add a Version function to GG | 14:19 |
| kroddn | but likely it is the line that caused the crash before | 14:19 |
| planetmaker | kroddn, how would I obtain that? That's a verbatim copy&paste from my xterm | 14:19 |
| GeofftheMedio | but since it currently only works for the scons build, and not the cmake, we'd need to figure out a way to make it work for cmake as well | 14:20 |
| planetmaker | hm, why support two build systems actually? | 14:20 |
| GeofftheMedio | because kroddn doesn't like cmake | 14:20 |
| planetmaker | :-P | 14:21 |
| GeofftheMedio | but really, I'm not actually suggesting that we need to | 14:21 |
| GeofftheMedio | I'm saying we need to make it work for cmake | 14:21 |
| kroddn | no, thats NOT the fact | 14:21 |
| kroddn | I am not able to build FO correctly with cmake | 14:21 |
| kroddn | I cannot do some static linking tricks with it | 14:22 |
| planetmaker | uh... I didn't want to kick an ant hive :S | 14:22 |
| kroddn | no problem, there is no ant hive :-) | 14:22 |
| GeofftheMedio | consider that one of the reasons why I say you dislike it | 14:22 |
| kroddn | <planetmaker> kroddn, how would I obtain that? That's a verbatim copy&paste from my xterm <--- you wouldnt be able to obtain it, because there is no debug info inside the debug-files | 14:23 |
| kroddn | I had problems getting gigi compiled in a Debug version using cmake ;-) | 14:23 |
| kroddn | I dont know for sure if the debug symbols work with the latest release | 14:25 |
| planetmaker | ok | 14:26 |
| kroddn | and the latest version doesn't crash on your system, thats the other problem :-) | 14:29 |
| kroddn | or say the problem is the missing starlanes | 14:29 |
| planetmaker | well, yes. But that's an old OpenGL. Not sure it's worth to worry about this system too much. | 14:33 |
| kroddn | it is goog to know where the crashes come from | 14:45 |
| kroddn | thanks for the help ! | 14:48 |
| planetmaker | :-) | 14:48 |
| planetmaker | you're welcome, my pleasure | 14:49 |
| planetmaker | weekdays in European evening hours I could do some testing sometimes. Tell me, if you need it | 14:49 |
| kroddn | yes, of course we need that | 14:50 |
| planetmaker | anyway... now I'm off for 1st May weekend :-) | 15:01 |
| CIA-76 | FreeOrion: id-ego * r3569 /trunk/FreeOrion/ (2 files in 2 dirs): | 15:12 |
| CIA-76 | FreeOrion: - Noticed that Xcode project wasn't compiling Vec3Utilities.cpp source file | 15:12 |
| CIA-76 | FreeOrion: - ServerApp.cpp: std::max(unsigned int, std::size_t) can result in inconsistent types in function template, changed local variables to std::size_t | 15:12 |
| CIA-76 | FreeOrion: id-ego * r3570 /trunk/FreeOrion/Xcode/Info.plist: Mac app bundle version bump. | 15:17 |
| kroddn | planetmaker: wenn GeofftheMedio mal nicht hier ist k�mma auch Deutsch quatschen :-) | 15:18 |
| kroddn | GeofftheMedio: how can I improve the increase of the constuction meter ? | 15:58 |
| kroddn | is there any research for that? | 15:58 |
| kroddn | GeofftheMedio: 3D comabt starts successfully on my nvidia system - but i cannot end/exit the combat window | 16:08 |
| kroddn | GeofftheMedio: I've got a positive feedback from an opengl 1.4 Netbook user, he is running FO | 16:13 |
| kroddn | strange thing those missing starlines on planetmaker's system | 16:13 |
| kroddn | http://img7.abload.de/img/freeorion9uh7.png | 16:17 |
| *** enigmatic has quit IRC | 18:43 | |
| CIA-76 | FreeOrion: tzlaine * r3571 /trunk/FreeOrion/UI/CombatWnd.cpp: | 20:00 |
| CIA-76 | FreeOrion: Fixed the bug that was causing all (or at least nearly all) of the | 20:00 |
| CIA-76 | FreeOrion: disagreement between the server and client object positions. The CombatObject | 20:00 |
| CIA-76 | FreeOrion: added as user data to each Ogre::SceneNode was not being set to the new | 20:00 |
| CIA-76 | FreeOrion: incoming CombatObject in CombatWnd::CombatTurnUpdate(), causing old object | 20:00 |
| CIA-76 | FreeOrion: state to be used. This is now fixed. | 20:00 |
| *** kr0ddnoMobil has joined #freeorion | 20:03 | |
| *** ChanServ sets mode: +o kr0ddnoMobil | 20:03 | |
| *** kroddnoMobil has quit IRC | 20:04 | |
| kr0ddnoMobil | GeofftheMedio: i have the same problem running ufoai on my notebook, so the crash is definitively no FO specific problem | 20:23 |
| kr0ddnoMobil | I'll upgrade now to Ubuntu 10.04 | 20:24 |
| STalKer-Y | just ignore GeofftheMedio and talk in deutsch :> | 21:06 |
| kr0ddnoMobil | GeofftheMedio needs input input input! | 21:11 |
| *** kr0ddnoMobil has quit IRC | 21:23 | |
| GeofftheMedio | google translate says: if time is not here GeofftheMedio kömma also German chat :-) | 22:34 |
| GeofftheMedio | so... that's nice. | 22:34 |
| GeofftheMedio | kroddn: since there's no transition to and from 3D mode in the current build, you can probably leave those gluseprogram lines commented out for the v0.3.14 release | 22:35 |
| GeofftheMedio | it's going to matter again once 3D combat is working, though... | 22:35 |
| *** welterde has quit IRC | 23:05 | |
| *** kroddnoMobil has joined #freeorion | 23:53 | |
| *** ChanServ sets mode: +o kroddnoMobil | 23:53 | |
Generated by irclog2html.py 2.5 by Marius Gedminas - find it at mg.pov.lt!
