[PARPORT] Re: Iomega Ditto 3200

Claus-Justus Heine (claus@momo.math.rwth-aachen.de)
18 Jun 1998 15:07:25 +0200

Claus-Justus Heine <claus@momo.math.rwth-aachen.de> writes:

> Claus-Justus Heine <claus@momo.math.rwth-aachen.de> writes:
> > Mmmh. Well. Actually: it should work to use gzip's way of computing
> > things, and then multiply the resulting CRC by x^15 and take the
> > remainder mod x^16.
> The three lines above are complete nonsense. I'm sorry.

Ok, next attempt. Of course, reverting the order of powers in a
polynomial ring IS an isomorphism. And the connection between the two
point of views (i.e. LSB is the highest power of x <-> MSB is the
highest power of x) w.r.t. to computing a CRC sum really is only a
matter of reverting the order of bits. I.e. a generating polynomial
0x1021 (that is x^0 + x^5 + x^12 + x^16) where the LSB corresponds to
the coefficient of x^0, generates the same CRC values as the bit
reverted polynomial 0x8408 (which also is x^0 + x^5 + x^12 + x^16),
only that the order of bits in the resulting CRC sums is reverted
(w.r.t. to the other point of view).

The included sample program below computes both case for some 32 byte
sample input streams.

Compile with gcc -o crc crc.c

and run with

crc 0x8408


crc reverse 0x1021

to get the result the bpck tape drive uses.



  Claus-Justus Heine

  PGP Public Key:

  Ftape - the Linux Floppy Tape Project
  WWW : http://www-math.math.rwth-aachen.de/~LBFM/claus/ftape
  Mailing-list: linux-tape@vger.rutgers.edu

('binary' encoding is not supported, stored as-is)

This archive was generated by hypermail 2.0b3 on Wed 30 Dec 1998 - 10:17:52 EST