Cornell University Submissions

		Coordinated by: Dennis P. Costello
				National Nanofabrication Facility
				G02 Knight Lab
				Cornell University
				Ithaca, NY 14853-5403

				607-255-2329

TPC - Author: Dennis Costello

	VAX Native Mode replacement for the TPC utility. This version fixes 
one small bug in the version submitted to the Spring 86 (Dallas) tape, where 
the tape drive specified on the command line was required to be MS:. Turns out 
that the string 'MS:' appeared in one spot in the code, but I only noticed 
this very recently, because I had defined MS (and MSA, MS0, MSA0) to be 
logicals which translate to MUA0: when I got a TU81. My apologies to any who
were confused. New versions of TPCIN.COM and TPCOUT.COM are also enclosed.

	Source, object and executable files are included. Simply change the
appropriate line in TPC.CLD to point the correct location of the image file,
then issue the following commands (the compiles and link were performed under 
VMS 4.6): 

	$ pas tpc			! optional
	$ for tapeio			! optional
	$ mar read			! optional
	$ link tpc, tapeio, read	! optional
	$ set command tpc
	$ libr /insert sys$help:helplib/help tpc

	To use TPC to copy a tape using a TU81 (TU81+, TK50) tape drive, issue
the following commands: 
	
	$ mou mu:/for/cache=tape_data/nowrite
	$ tpc mu: disk:[directory]file.tap
	$ dism mu:
	$ mou mu:/for/cache=tape_data
	$ tpc disk:[directory]file.tap mu:
	$ dism mu:

	TPCIN.COM and TPCOUT.COM (included on the tape) have been set up to do 
the two sides of the copy operation automatically. I usually submit these as
batch jobs. 

	Note that the /CACHE=TAPE_DATA speeds up TPC by a factor of about 4, 
in that TPC does not perform overlapped I/Os and therefore has no other hope
of getting the tape to stream. If the output tape should be 6250, include 
/DENS=6250 on the MOUNT command for the output tape. This has not been tested 
with long file names (I usually use .TAP as a filetype with TPC). The output 
tape need not be the same density, nor written on the same drive, as the 
input tape, but note that there is no attempt to handle premature end-of-tape 
or other I/O errors with any element of grace at all.