|
2006-01-09 17:35
|
screen0031.png (35,876 bytes) |
|
2006-01-09 17:36
|
NAN.png (57,774 bytes) |
|
2006-01-09 17:36
|
|
|
|
Does this happen every time? I just played through the attached mission twice and it nothing strange happened. |
|
|
No.. It only happens a few times... But where it happened very often the last time: FSPort, mission "Out of the Dark, Into the Night" (fs1a1m05.fs2) - when you select to fly with the Valkyrie... Well... At last it shoots ME away... And in this mission almost everything can do this... An missile, an exploding ship when I'm near... Maybe the Valkyrie doesn't have enough mass an gets catapulted out of the area ;) |
|
|
This may be linked to an error I was attempting to track down last week. Does it still happen if you use -loadallweps? I think I asked that in the forums but I don't remember if I got an answer or not. |
|
|
Grab current CVS and give it another try. |
|
|
First of all: Without -loadallweps, this mission crashes... And WITH -loadallweps, it doesn't... But in BOTH cases, this strange bug can happen... And for new CVS-Builds... WMCoolmon is doing a lot of Scripting-Stuff... I'm sure it's hard work, but almost every change breaks compiling under Linux ;) So I wasn't able to compile a new CVS-Build for tree days... |
|
|
For current CVS change the LUA_API call to int and it should build again. It would be nice if all those changes were tested a bit better though ;) The crash at least should be fixed. I can't seem to get the other stuff to happen though. I put in most of the code from my own tree that could be related so perhaps it was fixed by one of those things. If it works for you then we'll know for sure. |
|
|
Well... I think it happens more often with "custom" oder "modified" weapons... But I had this happen with "normal" weapons, too... Mostly with beams... Oh... Now it only happens a few times with "normal" weapons... Very seldom indeed... But it happens... Hmm... I think I'll attach a TBM for FSPort which modifys the Disruptors. Maybe you can find an error there... Maybe I just "overdid" it and some variables there are to high... |
|
2006-01-19 15:18
|
|
|
|
Still worked fine for me. I played it a half dozen times just to be sure. Although I encountered a totally different issue (and so far can't recreate it), I haven't experienced any of the problems you've mentioned. Enough people are starting to see these same things that there is obviously a problem somewhere. What cmdline options are you using? Any other custom tables or anything? Are you using the mediavps (current, older, etc.) or just the plain data? There has got to be something that I'm not doing here, or just don't have which is bringing it on. Try and get a full debug log (with a debug_filter.cfg file in place) too and attach that. |
|
|
Hmm.. Strange... Maybe it's an issue of my system ;) Okay... Infos: Build: Always the newest (self-compiled) MediaVP: 3.6.8 Beta (all VPs) CMDLINE: -ambient_factor 75 -spec_exp 11 -spec_point 0.6 -spec_static 0.8 -spec_tube 0.4 -fov 0.51 -spec -glow -jpgtga -nomotiondebris -2d_poof -dualscanlines -ship_choice_3d -targetinfo -3dwarp -snd_preload -fps -warp_flash -rearm_timer -img2dds -missile_lighting Extra Tables and ships: * In FSPort: Only a TBM to use the new Leviathan-Model and the better Disruptors * In FS2_Open: A few of TrashMans Ships with the necessary weapons (Ships: CarLancer, CIHaven, Columbus, Gallileo, SDF Marcross, Phoenix, Preatorian and the Retribution), a better Akethon SDG, a BFBlue and BFWhite, a HUGE Blue beam, a new AAA-Beam, a new/modified Kayser, a new/modified Helios, a modified Macross, two new/modified Deimos and a new/modified Erinyes. Tha Akethon is the only TBM that overrides the original Weapon... System: * gcc-3.4.5 * glibc-2.3.5-r2 * Kernel 2.6.14-gentoo-r5 * dev-lang/python: 2.4.2 * sys-apps/sandbox: 1.2.12 * sys-devel/autoconf: 2.13, 2.59-r6 * sys-devel/automake: 1.4_p6, 1.5, 1.6.3, 1.7.9-r1, 1.8.5-r3, 1.9.6-r1 * sys-devel/binutils: 2.16.1 * sys-devel/libtool: 1.5.22 * media-libs/openal-20050504-r1 * media-libs/libsdl-1.2.9-r1 I'll upload a debug-log soon... EDIT: Okay... The debug-Build crashes with "ASSERTION FAILED: "is_valid_vec(&pi->rotvel)" at physics/physics.cpp:550", when I usually get catapulted away... I uploaded the log and will run the debug build through gdb... EDIT2: Okay... gdb was useless ;) It only told me, the build got terminated by SIGTRAP. P.S.: Other ships disappear very seldom... Mostly the own ship gets catapulted away... bearbeitet am: 01-19-06 12:58 |
|
2006-01-19 17:51
|
|
|
|
The physics rotvel problem is what I had noticed too. Maybe it wasn't unrelated like I thought it was. :) For some reason it never occurred until I used the tbm, but it seems strange that it would be the cause. I'm trying to trace exactly where the rotvel vector is getting set wrong, so far without luck. I have figured out 2/3rds of the assignments that aren't the problem at least. Hopefully I'll have enough time tonight to track it down and we'll see if that was the real problem or not. |
|
|
I'm giving up for now, I'll try again over the weekend. I did figure out that it's the object (a ship that's blowing up) that has it's position set the NaN values that is the problem. Once it's NaN it filters down and screws everything else up. I haven't figured out where the object pos is getting set wrong though, that's what I have let to do. I suppose it's possible that the ship should be dead and the flag isn't set yet, but I can't say for sure. If you want to take a crack at it yourself just add a "Assert(is_valid_vec(&cur_pos));" to obj_move_all(). When you hit the Assert() then you know you are on the bad ship. Where it messes up before that point is the question. |
|
|
[intense sarcasm] Well, it took 3 days but I finally tracked this down to a massive and terrifying div-by-0 bug. I'm totally thrilled to finally have conquered this problem and am eager to move on to the next! [/intense sarcasm] [truth] Please replace every word in the above statement with the worst curse words in your vocabulary. The numbers should be replaced by at least 4 such words, hopefully the worst of them all. Feel free to complete this assignment in as many languages as you are capable. [/truth] And also try current CVS to see if it's totally fixed... Please excuse me, I'm going to pass out now... |
|
|
I love the humor of you guys :) Taylor: Maybe you should take a rest? Until FS2_Open isn't totally crashing or doing other bad stuff you can treat some bugs (most of mine, at least) as low priority... You guys are working hard... Maybe too hard sometimes ;) EDIT: Seems to be fixed... Thanks :) bearbeitet am: 01-21-06 12:46 |
|
|
Fixered. |
- Anonymous
