On Sat, 20 Jan 2007 12:37:25 +0100 Bertrik > I've had the same problems too. > > Isn't it somehow fundamentally wrong to require kernel headers for sane? > Sane is a user-space package not kernel-space. > > I remember seeing the compilation problem only with the epson backend, > but surely other backends must handle byte-order issues. Can't we just > fix the epson backend instead of requiring kernel headers for everyone?
That header happens to have the same name of a kernel header, but it is not a kernel header at all. It's purpose is to provide portable definitions for endiannes conversion. theorically, we should fix all the backends to use that header. -- Best regards, Alessandro Zummo, Tower Technologies - Torino, Italy http://www.towertech.it
