Creating a Panic UI image (either the default or loadable) The key steps are: create an indexed image using the MacOS X system 8 clut, saved in QuickTime uncompressed 256 color format. Run it through the genimage tool to create a C structure or a kernel loadable file. ===== Create the image Using an application like Photoshop, create an image to be used as the image displayed at panic time. Your selection of colors is limited to those found in the MacOS X system 8 clut; in the application you're using, make sure you are in "indexed mode" and that the supplied CLUT (appleClut8.act) has been selected. * The appleClut8.act is the default Mac OS X CLUT. Keep in mind the following. * There must be at least 20 lines at the bottom of the image reserved. This is used by the system for displaying extra panic information. There can be more than 20 lines, but you'll have to provide this information when generating the image for the kernel. * You must determine the colors used by the font for displaying the panic information. There are forground and background colors. The default foreground is "100% White". It is represented by a 24-bit value of 0xFFFFFF. The default background is "13% White, or Dark Gray". It is represented by a 24-bit value of 0x222222. To change the defaults, you'll have to provide this information when generating the image for the kernel. Save the completed image as a TIFF (still indexed off the CLUT). ===== Convert the TIFF indexed image to QuickTime RAW Using Preview, open the TIFF image. Use File:Export to save the TIFF image in QuickTime image format with options of "None" for compression and "256 Colors" for the depth. Quality should be "Best". The saved results should be a .qtif formatted RAW image. ===== Generate an image for the kernel. To generate the default kernel panic image file "panic_image.c", in your working directory, execute: genimage -i -n -fg <24-bit color> -bg <24-bit color> ** options other than -i are optional. To genertate a kernel loadable panic image file, execute: qtif2kraw -i -o -n -fg <24-bit color> -bg <24-bit color> ** options other than -i and -o are optional. ===== Other Info The reason an 8-bit image was choosen, was because it is easy to convert to 24 or 16 bit colors. The system does not typically run in 8-bit mode. If the system is in 8-bit mode. Then we have to check to see if the active CLUT is the same as the one that the image was created with. If the CLUTs are different. The image is converted to grayscale and the nearest matching gray in the active CLUT is used.