PDA

View Full Version : .trk file playback bug, possible affect on flightmodel. Please test and post results.



EightyOne_Dano
05-05-2005, 06:46 AM
I have a track file here:

http://homepage.ntlworld.com/daniel.clarke17/OUCH.zip

Play the track file with default settings and post back with your CPU type and wether my plane hits the tower on takeoff or gets airbourne.

So far all INTEL CPU's have shown the tower crash and all AMD CPU's have shown a takeoff.

Thanks,

Dano.

EightyOne_Dano
05-05-2005, 06:46 AM
I have a track file here:

http://homepage.ntlworld.com/daniel.clarke17/OUCH.zip

Play the track file with default settings and post back with your CPU type and wether my plane hits the tower on takeoff or gets airbourne.

So far all INTEL CPU's have shown the tower crash and all AMD CPU's have shown a takeoff.

Thanks,

Dano.

Heinz_Schuss
05-09-2005, 09:34 PM
I have an Intel P3-600 MHz and I see all 4 planes takeoff fine. Straight down the runway, no problems.

I have had some track weirdness which could be aircraft out of sync with terrain position, or position in time.

In a dogfight I flew, I do a victory roll just before the bandit hits the ground (so that I'm already inverted when he goes in), and then I land perfectly.

In the track, I start the victory roll only after the bandit hits, and then I land off to the side of the runway and bounce along in the snow.

It was quite a nice landing too http://forums.ubi.com/images/smilies/16x16_smiley-sad.gif

Schuss

Blackjack174
05-11-2005, 01:47 AM
A while back I myself have transmitted a .trk file to oleg , and the funny thing was that he could not play it back correctly http://forums.ubi.com/images/smilies/59.gif , he had (at that time) a Pentium 4 and I have a AMD Athlon, in the end I had to send him an .ntrk of the .trk
Maybe its only CPU extensions that mess up a track (no SSE2 etc.) because essentially the .trk "reflies" the mission with your control input and recalculates EVERY move from your plane and the AI again ,even any deviation from steering input handling will mess things up, and best guess is the flightmodel doesnt change , but the input http://forums.ubi.com/images/smilies/1072.gif
But that could only the programmers know for sure....

TAGERT.
05-12-2005, 08:29 AM
<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">Originally posted by EightyOne_Dano:
I have a track file here:

http://homepage.ntlworld.com/daniel.clarke17/OUCH.zip

Play the track file with default settings and post back with your CPU type and wether my plane hits the tower on takeoff or gets airbourne.

So far all INTEL CPU's have shown the tower crash and all AMD CPU's have shown a takeoff.

Thanks,

Dano. </div></BLOCKQUOTE>The *.trk files are very dependend on the system they were recorded on.. and the state of that system at the time of recording. If you want an idependend file that you can share with your friends that does not care about what kind of PC it was recorded on, then do as Oleg does and record *.ntrk files.

In a nut shell, the *.ntrk files contain the RESULTS of the game/sim, the *.trk files contain the INPUTS to the game/sim that than in turn get re-calculated by what ever PC it is running on.

EightyOne_Dano
05-12-2005, 10:10 AM
and that Tagert is my whole point. What concerns me is the fact that out of all machines used to playback the track so far all Intel chips(bar a single P3) have reported the crash and all AMD have not. This leads me to believe that the flight model data is being corrupted/miscalculated by one or other type of chip.
If it was a ransom spread I would not be concerned but as it is such a definite split it does have me concerned.

Dano.