Recently I wrote a code that can split animated GIF image to several frame-images.
The output images has the same logical screen descriptor/global color table, and their own graphics control extension/image descriptor/local color table/image data.
The code works great, but when I use imagecreatefromgif/imagecreatefromstring to open each frame-image, they lost the logical screen width/logical screen height, also the position on logical screen, their width and height become the dimension from image descriptor.
I'm using GD 2.0, haven't tested on 2.1 yet.
I'm just wondering whether it is a bug or not, if GD intend to do like this, I can just write very few codes to fix this.