Hexxen
Italy
Asked
— Edited
I was so amazed of the new function and I'm testing it. I think there's a problem with your algorithm when you save the array matrix into ram. The bigger you set the finger pad the most ram it uses. I'm using the windows developer preview but i'm almost sure this will happen on all O.S.
I made some tests and i figured out that you keep track of too many coords during the fingerpad press function. I think you made this to simulate the movement of the finger.
I thought about a solution too: use a timer to read the coords and send the servos to the new coords or if you want to simulate the finger movement increase the timer loop so it won't save too many coords
Can't see a memory leak...
At beakerhead launch, we had it running on the 2gb tablet for 6 hours with no issues. What kind of memory usage are you seeing that is a concern?
I'll write this in a hurry but just try to set the finger pad something like 800*600 and move the finger/mouse without taking him off. on my computer will run out all my 6gb ram and ARC will crash
Hmmm, it would seem disposing the graphics object is not being garbage collected. What kind of CPU do you have? I might have to force garbage collection if your cpu threads are all eaten up by other tasks.
i7 930. If i take off the finger faster enough, there ram resets to normal <60 mb but the more i move without taking it off the more the ram will increase exponentially
Tried on my laptop (I3 350m 4 gb ram) and won't crash but the ram used during the tracking will run over 1.2 gb
Try the latest version. I now keep the graphics object global. Resize it on the resize event. Using a panel instead of a picturebox. And dispose only occures when the control is removed. redraws on auto invalidate. I should have done it this way to begin wtih, stupid me
Thanx for finding that :)
no problem. you're the best when you fix them