I used them but without luck so far. I'm still trying to capture key press events for a chat box.
Today I changed my custom input handler so that it works with the new methods and added a check in update() which should either test for actions or just capture the chars.
The capturing works but the action is called as well even in capturing mode.
I found it. In the game class I call isValidCommand() and I thought it just processes the events in InputHandler. But these are two different things as KeyBindingManager is checking the keyboard state itself.
So the capturing worked in the custom InputHandler and the action came from the main game class.
I started working on the input system and input handlers. The following changes have already been made:
- InputSystem.createInputSystem is deprecated - not needed any more
- InputSystem.getMouseInput and getKeyInput is deprecated - use MouseInput.get() and KeyInput.get()
- Joystick support is in - please test it, starting point is com.jme.input.joystick.JoystickInput
- one can listen to mouse/joystick/keyboard events
Please post any problems with the new input stuff.
I'll post further advancement here...
I don't want to sound ungrateful, as I know the amount of work that goes into this, but could I repectfully request that next time you make major refactorings you make some posts in advance rather than just go ahead and update the cvs.
Apart from keeping the cvs in synch with other dependant libs (I think jmebui is still broken now, as are some others no doubt)
I know cvs changes, and doing a regular update on cvs one usually expects a few changes, but when a bunch of code changes need to be done and dependant libs are also broken (hence totally breaking your app) it's a little frustrating...even more so when you have to trawl through code to find the changes you need to retrofit.
As there are a number of non-core jme libs out there I also think it is important to co-ordinate refactorings a little more closely so when implemented they are all refactored at once. Better for everyone.
my 0.2 cents eclipse package explorer red x rant:-/
ps: if you did post some warning notice in advance, my apologies as I missed it
INFO: Child (cloth) attached to this node (rootNode)
Unexpected value for DX8 HAT: 17999Unexpected value for DX8 HAT: 17999Unexpected value for DX8 HAT: 17999Unexpected value for DX8 HAT: 17999Unexpected value for DX8 HAT: 17999Unexpected value for DX8 HAT: 17999Unexpected value for DX8 HAT: 17999Unexpected value for DX8 HAT: 17999Unexpected value for DX8 HAT: 17999Unexpected value for DX8 HAT: 17999Unexpected value for DX8 HAT: 17999Unexpected value for DX8 HAT: 17999Unexpected value for DX8 HAT: 17999Unexpected value for DX8 HAT: 17999Unexpected value for DX8 HAT: 17999Unexpected value for DX8 HAT: 17999Unexpected value for DX8 HAT: 17999Unexpected value for DX8 HAT: 17999Unexpected value for DX8 HAT: 17999Unexpected value for DX8 HAT: 17999Unexpected value for DX8 HAT: 17999Unexpected value for DX8 HAT: 17999Unexpected value for DX8 HAT: 17999Unexpected value for DX8 HAT: 17999Unexpected value for DX8 HAT: 17999Unexpected value for DX8 HAT: 17999Unexpected value for DX8 HAT: 17999Unexpected value for DX8 HAT: 17999Unexpected value for DX8 HAT: 17999Unexpected value for DX8 HAT: 17999Unexpected value for DX8 HAT: 17999Unexpected value for DX8 HAT: 17999Unexpected value for DX8 HAT: 17999Unexpected value for DX8 HAT: 17999Unexpected value for DX8 HAT: 17999Unexpected value for DX8 HAT: 17999Unexpected value for DX8 HAT: 17999Unexpected value for DX8 HAT: 17999Unexpected value for DX8 HAT: 17999Unexpected value for DX8 HAT: 17999Unexpected value for DX8 HAT: 17999Unexpected value for DX8 HAT: 17999Unexpected value for DX8 HAT: 17999Unexpected value for DX8 HAT: 17999Unexpected value for DX8 HAT: 17999Unexpected value for DX8 HAT: 17999Unexpected value for DX8 HAT: 17999Unexpected value for DX8 HAT:
Update: problem message was coming from PPJoy - "Creating PPJoy Virtual joystick 1 polling = true", which I uninstalled and message gone. Worked ok before under native jinput so not sure what has changed with the jme/lwjgl implementation.
It seems I really underestimated the number of uses of these methods :-o - sorry for that. I will be more careful next time and will be setting even small methods to deprecated instead of removing them instantly; and consider a warning on major changes, sure.
Concerning other dependent libs: what you say is true - but at least I have sent a patch for bui to Sam and a new version of nui to Guurk before checking in - I thought it would get in quickly…
Concerning the joystick issues - seems like a problem in jinput. So if you want to have joystick support in your game, please report the issue over there.
If you don't want joystick support, I have added a method to disable it (so messages will be gone): simply make the following call before creating display (start the game)
Though I understand your plight, this is the inherent risk you take by running directly from CVS. If you are developing a game and need the assurance of stability during development you should probably be using the release instead. You have no guarantees that the CVS won't drastically change at any point in time because it is, and must remain the "development" code. I think for exactly your situation is why there is a release build.
Though I understand your plight, this is the inherent risk you take by running directly from CVS. If you are developing a game and need the assurance of stability during development you should probably be using the release instead. You have no guarantees that the CVS won't drastically change at any point in time because it is, and must remain the "development" code. I think for exactly your situation is why there is a release build.
darkfrog
Sure, that's a risk..but I'm more prototyping than developing a game so I'm willing to trade off. My point being a suggestion for more coordination and warning, rather than just dump a whole load of new stuff in. There's levels of development. I'm not sure at what point you regard something as being developed enough that you can check/merge it into the main branch, or keep it seperate until tested and complete (at least beyond basic development). Just a thought.
Concerning the joystick issues - seems like a problem in jinput. So if you want to have joystick support in your game, please report the issue over there.
If you don't want joystick support, I have added a method to disable it (so messages will be gone): simply make the following call before creating display (start the game)
I'm using a usb biofeedback interface and a usb Logitech wingman runmlepad (+ 2 P5 usb gloves).
The rumblepad I've been accessing by Jinput directly (ie: non lwjgl and non jme) for some time, along with all the other usb devices and they have been working great. After your update I can't get access to the USB biofeedback device. I had to JoystickInput.setProvider( InputSystem.INPUT_SYSTEM_DUMMY ); and now I can use it fine, as well as access the rumblepad ok by direct (ie non jME) Jinput access.
I suspect that somewhere there is a problem with the jinput implementation, but buggered if I know where it is:-)
Could be the lwjgl implmentation?
I investigated a bit on the message and it looks like this: jInput prints the error message ("Unexpected value for DX8 HAT") and it seems to be caused by lwjgl, which tries to use everything as an 'axis' if it does not know the type…
And, yes, if you initialize the lwjgl controllers and update the display before handling the jinput events, lwjgl Controllers.update consumes all the events
