glTexcoordPointer

i want to achieve a faster way of getting the correct color. . The colors range from [0..4096] and are packed in a (signed) short, but i can use bit operators to get the correct uv coordinates ( 8 bits for u and v tex coordinates ). But the opengl specifications said i can only use GL_INT, GL_SHORT.. etc but GL_BYTE isn't specified to use with glTexCoord. So right now i'm converting this 0-4096 range to u/v coordinates in the range [0..1] in order to use them with glTexPointer and GL_FLOAT. But i'm pretty convinced that i can use the bytes directly without any conversion pericles. Paul