On Thu, 24 Jan 2002, Kurt J. Lidl wrote:

# On Thu, Jan 24, 2002 at 01:44:18PM -0600, Peter Seebach wrote:
# > I haven't even figured out exactly what it's for.  Actually, that's not quite
# > true:  I have a vague belief that the idea is that I can have a read-only
# > /usr/src, NFS-mounted to hell and back, and everyone uses /home/obj for
# > builds, and it All Just Works.
#
# The big idea for the obj subdirectory stuff had to do with the belief
# that you would have your /usr/src on one drive, and /usr/obj on a
# second drive.

*sigh*.

This really needs to be worked around.  I have a good guesstimation of
how big to make /usr/src, but I have more of a problem figuring out how
big /usr/src separated from /usr/obj ought to be.

# Then when you did "make obj" at the top of the tree, the /usr/obj tree
# would be populated with a directory hierarchy, and the /usr/src tree
# would be symlinked into the /usr/obj tree.  Then as you did
# "make depend" and finally "make all", the /usr/obj tree would get all
# the .o files, along with the binaries and formated man pages.

This is sensible, but I should still be able to put /usr/obj pointing
anywhere I want it, including /usr/src/obj.  This used to work.

				--*greywolf;
--
If anyone requests a reason as to why Windows NT is inferior to UNIX,
refer them to the process scheduler, for starters.  Of course, users
don't care, and programmers try not to, even though they both should.
If that fails, reiterate that remote administration and control of a
node is a *good* thing, especially if network security is concerned.