Berkley-isms (was: Why no job control in 386/ix)

[Kemp at DOCKMASTER.NCSC.MIL]

 > You think that ISC's csh is bad, just try the one that ATT ships
 > on their version of the 386 software!  Eech!.
 >
 > My guess at the reason.  Csh is a Berkley [sic] development and
 > SystemV people don't care for Berkley-isms.  I once complained
 > about the absence of '-r' on the System V cp command and was told,
 > in effect, that in that context, '-r' was an offence against
 > Decency and the Natural Order of Things.

As implemented by Berkeley (or at least in SunOS), cp -r *is* an offense
against Decency and the Natural Order of Things.  Of course, so is 'cd
fromdir; tar cf - .  | (cd todir; tar xvf -)'.  So is 'find blah blah
blah | cpio -p'.

I don't think that having a simple command to 'make an exact copy of
this directory hierarchy' is inimical to the Unix Philosophy of 'do one
job and do it well'.  If cp -r wasn't such a botch, it would be worth
having.

Just for the record, is there *any* way to do a recursive copy
correctly?  I.e.  one that doesn't:

 * turn symbolic links into actual files
 * turn link loops into a series of infinitely nested copies
 * alter the modify and change times
 * choke on block and character special files
 * turn holes in sparse files into real disk blocks

   Dave Kemp <Kemp at dockmaster.ncsc.mil>