Sorry to flood this board with naive questions. Windows is said to internally keep a "paint informatioin structure" for each window. the structure contains the smallest rectangle encompasses the invalid region. But the WM_PAINT message itself doesn't carry this structure, then in general how does the handler know about the invalid region?
I know that the BeginPaint function will properly set the clipping region for the device context (and even "materialize" the "paint informatioin structure" into a "real" PAINTSTRUCT). But some other functions, say GetDC, will not do this. Does that mean BeginPaint is more efficient as it won't paint to unnecessary region?
Thanks for help.