I had exactly that problem with a (sounds like) similar application. My behvaiour was waking up after an elapsed time and it would occasionally stop calling the processStimulus method. I tried running it through profilers several times and the only thing I could resolve is that the Behaviour thread was being locked.
I assumed it was a deadlock on my part of the code, yet all my code seemed to exit cleanly. Also, if I replaced the behaviour thread with an asynchronous wakeup (using a Thread that just sleeps then invokes my behaviour) I didn't have the same problem. The downside was that my behaviour was no longer synched with the Java3D render cycle. I tried to make my own wakeup criteria but you can't because of protected methods, etc. In the end I settled for a horribly messy implementation that has never failed. Ready? I put a TransformGroup with no geometry in my scene. I then have an asynchronous thread change its transform periodically (every 33 ms). Finally, I have my old behaviour wake up on transform change. Sounds silly, I know, but as you say the lock-up was very frustrating. Since others don't seem to have that problem, my guess is that we both have thread issues. I also update a Swing GUI, but that should all happen with InvokeLater. I have to update textual displays, vehicle positions and camera orientations from the one stream of data. Good luck! josh ---------------------------------------------------- Joshua L. Richmond, BASc, MSc. Systems Design Engineering / Orbital Express Program MD Robotics 9445 Airport Rd, Brampton, ON, Canada (905) 790-2800 x4332 http://www.mdrobotics.ca ----------------------------------------------------- >>> [EMAIL PROTECTED] 01/21/04 11:37AM >>> John Wright: Talking with someone here at work, they mentioned any textfield updates or swing updates may be a problem because they are not thread safe, so it would be wise to use Runnables with invokeLater. I have Textfield updates and JPanel updates so I'll have to put them in a Runnable to see if that helps. Priya: Right now I'm using wakeup on time...mine does a lot too. I have it update VRML vehicle positions and attitudes, calculate camera positioning while attached to a moving vehicle and calculated burn plume directions...it seems like a lot but I've had other programs do more. Although you make a good point that the elapsed time and overload of work may be interacting. This coupling may cause the failure if the calculations don't finish before the time is up...I'll switch to elapsed frames and see if that helps....the problem is that I can't reproduce the bug. I've put in a try/catch but since I can't reproduce the bug I can't see if that helps yet. Thanks, Mario -----Original Message----- From: Piriyakala Suresh [mailto:[EMAIL PROTECTED] Sent: Wednesday, January 21, 2004 10:19 AM To: ZACZEK, MARIUSZ P. (MARIO) (JSC-DM) (NASA) Subject: [JAVA3D] ? Behaviors: processStimulus issue ? Hi, I have encountered the same problem, and was looking to the mailing archive for any ideas! It may be a good idea to compare notes. I have many other behaviors that have never given me this problem. This behavior is different in 2 ways: it wakes up on elapsed time as opposed to frames by all my other time dependent behaviors, and it has to do a lot during a single processStimulus. Can you tell me anything similar in your situation? Regards, Priya -----Original Message----- From: Discussion list for Java 3D API [mailto:[EMAIL PROTECTED] Behalf Of John Wright Sent: Wednesday, January 21, 2004 10:24 AM To: [EMAIL PROTECTED] Subject: Re: [JAVA3D] ? Behaviors: processStimulus issue ? Mario, I'm afraid you'll probably have to carefully go through your code to spot the error. I doubt it's a common bug in Java 3D as I run Pernica twenty four hours a day seven days a week and I've never observed a behavior spontaneously stopping. Are there any situations where your processStimulas code might fail and not reset the trigger condition? - John Wright Starfire Research "ZACZEK, MARIUSZ P. (MARIO) (JSC-DM) (NASA)" wrote: > > I have my bounds set pretty high (30,000) and I've had this occur even > though I've not moved/rotated that far away ... and this not something I can > even duplicate...it just occurs sometimes (around 5% of the time...though > still annoying as hell). > > thanks, > > Mario > > Mariusz Zaczek > NASA - Johnson Space Center > Automated Vehicles and Orbit Analysis / DM35 > Flight Design and Dynamics Division > Mission Operations Directorate > Bldg: 30A Room: 3040A > > Disclaimer: "The opinions, observations and comments expressed in my email > are strictly my own and do not necessarily reflect those of > NASA." > > "Failure is never quite so frightening as regret." - Cliff's wife (The Dish) > > -----Original Message----- > From: Discussion list for Java 3D API > [mailto:[EMAIL PROTECTED] Behalf Of John Wright > Sent: Wednesday, January 21, 2004 10:09 AM > To: [EMAIL PROTECTED] > Subject: Re: [JAVA3D] ? Behaviors: processStimulus issue ? > > Mario, > > There are several things that might cause this. The most obvious > perhaps is that perhaps you move your view out of the bounds of the > behavior (hence it stops processing). > > - John Wright > Starfire Research > > "ZACZEK, MARIUSZ P. (MARIO) (JSC-DM) (NASA)" wrote: > > > > Has anyone ever experienced your simulation/animation fail or stop? and > you > > couldn't > > restart it? My Java3D code periodically just stops animating and I can't > > seem to figure > > out the problem...it's as if the 'processStimulus' function just stops. Is > > there anything > > that could cause this and is there any remedy for this? > > > > Everything else seems to work..ie buttons, rotating 3d scene, etc...but > > updating it > > via processStimulus just does nothing...I have to restart the application > to > > get it > > to work. > > > > Mario > > > > Mariusz Zaczek > > NASA - Johnson Space Center > > Automated Vehicles and Orbit Analysis / DM35 > > Flight Design and Dynamics Division > > Mission Operations Directorate > > Bldg: 30A Room: 3040A > > > > Disclaimer: "The opinions, observations and comments expressed in my email > > are strictly my own and do not necessarily reflect those of > > NASA." > > > > "Failure is never quite so frightening as regret." - Cliff's wife (The > Dish) > > > > > =========================================================================== > > To unsubscribe, send email to [EMAIL PROTECTED] and include in the > body > > of the message "signoff JAVA3D-INTEREST". For general help, send email to > > [EMAIL PROTECTED] and include in the body of the message "help". > > =========================================================================== > To unsubscribe, send email to [EMAIL PROTECTED] and include in the body > of the message "signoff JAVA3D-INTEREST". For general help, send email to > [EMAIL PROTECTED] and include in the body of the message "help". > > =========================================================================== > To unsubscribe, send email to [EMAIL PROTECTED] and include in the body > of the message "signoff JAVA3D-INTEREST". For general help, send email to > [EMAIL PROTECTED] and include in the body of the message "help". =========================================================================== To unsubscribe, send email to [EMAIL PROTECTED] and include in the body of the message "signoff JAVA3D-INTEREST". For general help, send email to [EMAIL PROTECTED] and include in the body of the message "help". =========================================================================== To unsubscribe, send email to [EMAIL PROTECTED] and include in the body of the message "signoff JAVA3D-INTEREST". For general help, send email to [EMAIL PROTECTED] and include in the body of the message "help". =========================================================================== To unsubscribe, send email to [EMAIL PROTECTED] and include in the body of the message "signoff JAVA3D-INTEREST". For general help, send email to [EMAIL PROTECTED] and include in the body of the message "help".
BEGIN:VCARD VERSION:2.1 X-GWTYPE:USER FN:Josh Richmond TEL;WORK:4332 ORG:MDRobotics Ltd.;Systems Engineering #211 EMAIL;WORK;PREF;NGW:[EMAIL PROTECTED] N:Richmond;Josh TITLE:BASc., MSc. ADR;INTL;WORK;PARCEL;POSTAL:;Mailstop 23;9445 Airport Rd;Brampton;ON;L6S 4J3;Canada LABEL;INTL;WORK;PARCEL;POSTAL;ENCODING=QUOTED-PRINTABLE:Josh Richmond=0A= Mailstop 23=0A= 9445 Airport Rd=0A= Brampton, ON L6S 4J3=0A= Canada LABEL;DOM;WORK;PARCEL;POSTAL;ENCODING=QUOTED-PRINTABLE:Josh Richmond=0A= Mailstop 23=0A= 9445 Airport Rd=0A= Brampton, ON L6S 4J3 END:VCARD =========================================================================== To unsubscribe, send email to [EMAIL PROTECTED] and include in the body of the message "signoff JAVA3D-INTEREST". For general help, send email to [EMAIL PROTECTED] and include in the body of the message "help".
