Hi buddy,
I have some questions about the conceptions of dirty-rects and partial update.
The context is that, we use ogl 9.30 package , thus gui is rendered by gpu.
We did some hack job in ew gfx code to make gui rendered to our off-screen frame buffer object (openGL conception), which means we do not want to send gui frame to display directly.
After glFinish, we can get the rendered gui from the fbo color buffer. This works for a long period of time. And now we find our gpu in a poor performance, even takes 30ms+ to draw one
gui frame, I guess it is because we use the CoreGroup_InvalidateArea for the whole screen. thus, for each update, we redraw the whole screen content.
Now the question is, if we do not invoke CoreGroup_InvalidateArea in Update(), how can we get the rendering pass correctly working. As far as I know, we can get the dirty region in root object's
_this->regions, how can I use them to reduce the draw consumptions.
BR
Arnold