PhysX 3.x Plugin

#241 aah ok yeah that makes sense, i never actually tested the triggerdata this whole time (oops), i only ever tested the physics player against the health items,ammo etc.

#242 Also anyone else reading physX 3.3.1 was released a little while ago (april i think), it's a good idea to update it if you haven't already.

#243 removed

#244 Entity/component system discussion thread.

#245 Thanks for the tip on updated physx core, will grab that today...

Any significant performance improvements or anything with the latest SDK?

Fantastic to see progress so soon on the trigger stuff, keep up the great work!

Cheers
P

#246 Release notes physxinfo.com/news/12283/physx-sdk-3-3-1-is-available

There is a bug in the last repo update i made, if you add a PhysicsShape it will kill the trigger stuff. I'll fix it up during the week.

#247 Do you know if GMK will work out of the box with PhysX 3.3 in Torque3D? It was based of 2.8.4, so will it work now? If not, could we work on "porting" it to PhysX 3.3?

#248 Unfortunately it won't work out of the box with PhysX 3.3

Sorry getting this plugin to work with GMK is not something i will be doing.

#249 Ok. That's a dark area for me, so my GMK port will need to be updated by someone else.

GMK is a really good plugin, dumping it out just isn't right. We should at least keep it alive, as it is currently being help up by me. I can't port it to PhysX 3.3, because I don't know the API or have the knowledge. If anyone can take on porting my port (lol), then please do so. I hope that GMK will stay alive.

#250 Too be completely honest raa i have just never had interest in GMK but maybe someone else will fix it up for you.

@Paul:

This may take a little bit to track down what is going wrong, the whole trigger system seems to be buggy. The trigger won't activate if a PhysicsShape or RigidShape is anywhere in the scene,doesn't even have to be anywhere near the trigger, i suspect it's anything derived from GameBase? It is very, very buggy. It does this with Torque physics module and all three physics modules.

Has anyone else noticed any of these issues?

If anyone wants to reproduce the problem, just use the terrain mission within the full template, create a trigger and walk through it and it should work ok. Try it again but this time place a RigidShape anywhere (the bouncing boulder i used) and now the trigger won't activate. It makes no difference if you use torque physics module or any of the other three physics plugins.

#251 Just an update on this trigger bug, if multiple triggers are in the one scene only one of them will work (i think the first trigger that is loaded?)

#252 Crazy!

That is annoying for sure...

I remember there was some trigger bug in 3.5 that was fixed immediately after, did that maybe not get rolled into your physx3 branch?

Just an idea, I remembered something about the triggers getting screwed up, not sure if it would cause these issues though.

I used to have multiple triggers in a scene no problem in earlier T3D versions, including I think 3.0. The latest version of our Star Trek game was updated to 3.0 and we have lots of triggers that you fly through in the shuttle and they work. So all was well as of 3.0 at least in a non-physics build.

Will grab your latest commit off the repo later and give it a test, see if does the same for me.

#253 Actually i haven't tested that, i'll grab the vanilla T3D and give it a test

#254 The problem still happens on the latest official development branch. I gotta say it would make it a lot easier on the official repo if they used version numbers for the branches. Have master as the latest stable version and than have version number for branches and get rid of the development branch.

#255 Ya agreed, "development" is unreliable at best.

Bummer about this bug... Do you mean that even in a non-physics build of stock T3D 3.5 that only one trigger per scene is working now?

If so, wow they busted it good since 3.0 lol. Never used to be an issue.

#256 Yes on the official repo both master and development have this bug (physics or no physics). I'll try and dig through and see if i can find what has changed. Great example here why version number branches would be handy ;-)

#257 Aaah shit sorry i'm leading you on a wild goose chase. Something was wrong with the mission i created. I re-created and the two or more trigger thing now works just fine across all versions and so does the rigidbody thing.

Totally disregard all my posts above.

#258 Haha!

No sweat. Had not started chasing yet so no harm done :)

#259 Actually paul you where correct before, the problem is in 3.5 and my physx branch. If you save a mission file created with those versions the triggers screw up. Looking through the notes this was corrected in 3.5.1. I will have to get my repo in sync correctly with the official repo.

#260 Ok the mess is sorted out lol

The last repo update merged all the changes from the latest development branch from the official repo into the physx3 branch.

Do note there is still a fix in trigger.cpp that i added the other day that does fix up the triggers properly(with physics enabled) compared to the current GG development version.

@Paul: That request about PhysicsShape and triggers should now be fairly simple (fingers crossed), should hopefully have it done today.