View Issue Details
| ID | Project | Category | View Status | Date Submitted | Last Update |
|---|---|---|---|---|---|
| 0000332 | FSSCP | user interface | public | 2005-02-27 10:15 | 2005-05-23 05:00 |
| Reporter | LOA--Paul | Assigned To | taylor | ||
| Priority | normal | Severity | crash | Reproducibility | always |
| Status | resolved | Resolution | fixed | ||
| Summary | 0000332: Command briefing avi scrambles, game crashes two seconds later | ||||
| Description | After the mission where you are helping to test the Pegasus stealth fighters and you have to defend the Aquataine from ambush and you go through debriefing/load the next level, the command briefing starts. Right when it starts, the avi begins to play about 5 frames in, then scrambles, and then the game crashes. | ||||
| Additional Information | Always reproduces. Cannot advance any further into game. Tried running with only the core VP file, crash still occurs. | ||||
| Tags | No tags attached. | ||||
|
|
What command lines are you using? What version of windows? What soundcard? What graphics card? What drivers? D3D or OGL? What version of DirectX? Can you try this on other missions? What build(s)? Any errors? |
|
|
No command lines, vanilla FS Open Windows XP Home, Service Pack 2 Nvidia 6600GT (6.6.3.1) Sound Blaster Live! (D800) (3/18/02) D3D DirectX 9 Other missions seem to work. 3.6.5 FSOpen Just the usual "Program had an error and crashed, send report to Microsoft?" -It's a CTD besides that single message PC: P4 3.0GHZ 1GB Corsair RAM Nvidia 6600GT |
|
|
Try OGL and see if it does anything different. |
|
|
Just tried OGL, still CTD's at the exact same spot (about a second into the command briefing. I get to see it start, the words still being put on the screen, and then it crashes) |
|
|
Just tried it on a different pilot and replayed the entire campaign with a fresh install of the VPs. Still occured at the exact same spot with same symptoms. Always reproduced. |
|
|
Can you find out what exact ani is causing it and upload the version of the file you've got? Are you using the HOTU version by any chance? |
|
|
Reminder sent to LOA--Paul Could you please look at this bug again, and see if it is still occuring? It'd be great if you could upload the problematic ani to Mantis. |
|
|
Er, I'd have to play through the campaign again. My single player pilots erase whenever I use a multiplayer one, so I guess that's another bug but I posted about it on the HLP SCP forum. I did manage to fix the problem and finish the campaign by temporarily deleting my tango files, so it was exactly a problem with the avi. I'm not sure what avi, but it was the one at that mission described in the first report. |
|
|
Putting this into feedback, not seeign any other reports of this. |
|
|
Unless you're going to fix it, this should be reassigned. ;) Not sure whose department this should be; assigning to kasperl for the time being. |
|
|
Asigning to Taylor, he seems to be the one with the most knowledge about ani's and stuff. LOA-Paul: Could you please upload the ani in question? |
|
|
Is this the bug he just committed fixes for? |
|
|
@ADM: Haven't figured out the avi problem yet. The recent pilot fixes were only for retail release to current upgrade and not really for his problem. I've got no clue why they are getting deleted, it should always prompt before doing something like that (either through the upgrade message or out-of-date message). @LOA--Paul: I need more info on the pilot file problem before I can figure out what's going on there. For the avi issue, I can't find anything wrong with the anim itself. I think the mission you are referring too is "The King's Gambit" and the first anim in the cmdbriefing is cb_sm2-04a.ani. Start the game with the -allslev cmdline option and try just that mission in the techroom. If that is the one that crashes then let me know and I'll attach a test anim with the same name to test with. That is the only anim that has ever crashed for you though right? |
|
|
LOA-Paul: Are you using a HOTU build? If you are, that anim might just be corrupted or something. |
|
|
Reminder sent to LOA--Paul Could you please look at this bug again and upload the ani that's cauing the crash? |
|
|
Long standing feedback, will close in a week, unless someone objects. |
|
|
I've had this happen to me in FS1... when I replaced the file in question, it started working again. So I'm guessing this is just a corrupt cbanim. It would be nice to fail gracefully instead of crashing, though. Taylor (heh), any idea how to recognize a corrupt cbanim and simply stop displaying it, rather than continue to display it and crash? If you need something to test with, I've got a cbanim that reliably reproduces the crash in all versions of Freespace (FS1 through present). |
|
|
I don't know if it's possible to fail gracefully or not, just depends on exactly how it's crashing. I need a corrupt cbanim first to experiment with but it should be possible to make the decoder handle errors a bit better. |
|
|
Okay, for some odd reason Mantis won't let me upload files today. O.o I tried it with both Firefox and IE. I've uploaded them here instead: http://fs2source.warpcore.org/temp/cbanims.zip These crashed FS1 and FS2 for me in the past, but curiously they don't do it any more... I'm not sure why. They *do*, however, scramble and fail to display properly after a certain point. To see the scramble effect, put these cbanims in Freespace2/data/cbanims/ and run the FSPort. Choose mission 1-08 for cb_hol and mission 3-01 for cb_starmap10. |
|
|
Yep these are crashing pretty good. I added a quick check to bail on a playback stream if it detects corruption but I want to add a little more feedback to it. As it is you will only know it's corrupted if running a debug build but I would like to have at least a single frame notification that it's ended prematurely. I'll have it throw up some TV style test bars or something as the last couple of frames and then commit the change. I would greatly prefer if it could continue a stream if a frame is invalid. The RLE compression makes this nearly impossible though since the frame size varies and there is no sort of end or beginning frame marker. This is pretty easy to do if the exact problem is known before hand but that's rather useless in a real like situation. For now the bail-on-first-bad-frame option is the best I can come up with. Assuming that I ever have time to finish that other ANI replacement this won't be a problem for it since it uses beginning frame markers as an extra position check during streaming. That feature was originally just for better reverse and ping-pong play support but would work here too. |
|
|
I'm going to go ahead and resolve this, but it's not the complete fix that I want. The corrupted anis are handled but it would still die if only a partial file exists. Note also that a corrupted ani that triggers certain RLE decoding errors can still cause a crash. A partial fix for that is coming when I get the chance to test it better. A simple check for partial files is also in testing. The ANI format just doesn't allow for really good error checking I'm afraid. Fixered. |
| Date Modified | Username | Field | Change |
|---|---|---|---|
| 2005-02-27 10:15 | LOA--Paul | New Issue | |
| 2005-02-27 10:33 | kasperl | Note Added: 0001692 | |
| 2005-02-27 18:18 | LOA--Paul | Note Added: 0001696 | |
| 2005-02-27 18:26 | taylor | Note Added: 0001697 | |
| 2005-02-27 18:40 | LOA--Paul | Note Added: 0001698 | |
| 2005-02-28 04:32 | LOA--Paul | Note Added: 0001701 | |
| 2005-02-28 13:56 | kasperl | Note Added: 0001702 | |
| 2005-03-12 19:41 | kasperl | Note Added: 0001742 | |
| 2005-03-13 06:50 | LOA--Paul | Note Added: 0001806 | |
| 2005-03-13 19:27 | administrator | Note Added: 0001826 | |
| 2005-03-13 19:27 | administrator | Assigned To | => administrator |
| 2005-03-13 19:27 | administrator | Status | new => feedback |
| 2005-03-15 01:19 | Goober5000 | Note Added: 0001943 | |
| 2005-03-15 01:19 | Goober5000 | Assigned To | administrator => kasperl |
| 2005-03-15 16:52 | kasperl | Note Added: 0001963 | |
| 2005-03-15 16:52 | kasperl | Assigned To | kasperl => taylor |
| 2005-03-15 16:52 | kasperl | Status | feedback => assigned |
| 2005-03-16 03:32 | administrator | Note Added: 0001970 | |
| 2005-03-16 03:56 | taylor | Note Added: 0001971 | |
| 2005-03-16 12:56 | kasperl | Note Added: 0001972 | |
| 2005-03-31 08:55 | taylor | Status | assigned => feedback |
| 2005-03-31 12:07 | kasperl | Note Added: 0002072 | |
| 2005-04-04 16:55 | kasperl | Note Added: 0002085 | |
| 2005-04-04 16:55 | kasperl | Assigned To | taylor => kasperl |
| 2005-05-12 03:10 | Goober5000 | Note Added: 0002357 | |
| 2005-05-12 03:10 | Goober5000 | Assigned To | kasperl => taylor |
| 2005-05-12 08:35 | taylor | Note Added: 0002362 | |
| 2005-05-22 03:11 | Goober5000 | Note Added: 0002448 | |
| 2005-05-22 18:35 | taylor | Note Added: 0002451 | |
| 2005-05-22 18:35 | taylor | Status | feedback => assigned |
| 2005-05-23 05:00 | taylor | Status | assigned => resolved |
| 2005-05-23 05:00 | taylor | Resolution | open => fixed |
| 2005-05-23 05:00 | taylor | Note Added: 0002461 |
