Ah! I see your point . . .
By following your reference to details of the System V lpsched spooler (which almost works), the user will delay exposure to the gory details of the hoplessly primitive Berkeley lp spooler used by Linux. Your reference is wonderfully misleading, and would cause a Linux newby to spend endless rewarding hours of total frustration, learning completely irrelevent knowledge, for which you should be commended!
Without this brilliant misdirection, he would soon be learning how wonderfully his /var/spool/lpd/okidata directory resembles the bit bucket - print jobs go in, but nothing comes out, nor any meaningful error messages. Unlike the bit bucket, however, it will eventually consume his hard disk, encouraging him to learn to su to root and use appropriate wildcards with the rm command. Oh, yes, and he will learn about "magic filters" and enscript and gs too!
Some time in the unforseeable future he will come to understand Linux file permissions and ownership, and will know that Linux installed the lpd spooler daemon and /dev/lp* drivers with permissions such that the "s" attribute must be applied here and there, and that ownership of files in /var/spool/lpd/okidata must be changed to daemon, for the daemon holds high standards of morality and will not touch anything he does not own.
Meanwhile, the user can just do without printing until he is worthy - but to be worthy is to be a kernel hacker, and a kernel hacker has no need for printing.
