I've come across some strange behavior when specifying parts of the image to draw. If cx2 and cy2 are not the right- and bottom-most edges of the image (respectively) then you need to add 1 to the values of each of them to get the part you want.
In other words, cx1 and cy1 are using (0,0) for the upper-left corner of the source image, but cx2 and cy2 are using (1,1) for the upper-left corner of the source image. The really strange thing is that if cx2 is at the bottom-most part of the image it reverts back to the standard limit of (image width)-1, and if cy2 is at the right-most part of the image it reverts back to the standard limit of (image height)-1
In the image below, the blue lines cross at cx1, cy1, and the red lines cross at cx2, cy2, but in order to get the red lines to show up, you've got to use cx2+1 and cy2+1.
I've included a sample project to show the problem.
| Attachments: |
|
File comment: Example. partImageTest.zip [1.03 KiB] Downloaded 1448 times |
)
Actually, I decided to help out by doing a calculator for a demonstration of user input (that came up in some other topic). I happen to be one of those weirdos who does graphics and coding, and I like everything to be exact with my interfaces (which probably comes from doing so many Winamp skins).