I'm trying to implement texture-mapped-font drawer in OpenGL.
It currently works like so:
1. Load a .fnt file (containing character description, position etc).
2. Load a texture.
3. Splice the texture into smaller sub-textures according to the fnt file.
4. Assign each sub-texture to its character.
Is it really necessary though? Can I not somehow only draw the parts I need from the bigger texture map? Or do I actually have to generate the sub-textures.