If you want to move this information around and keep the error correcting properties, worth remembering that communication channels require carefully coded bit streams that (for Shannon) look like white Gaussian noise and for the electrical efficiency in the transmitter are free of long strings of 1's or 0's Can that be folded into the ideas of targeted bit error correction? --R. On Mon, Jan 28, 2019 at 6:03 AM David John Makin via math-fun < math-fun@mailman.xmission.com> wrote:
Starting from a single bit for 0 or 1 - surely no difference here…….
On 26 Jan 2019, at 01:15, Henry Baker <hbaker1@pipeline.com> wrote:
Clearly, one would like to put more redundancy into X than Y, and more into Y than Z.
Alternatively, encode X into more bits than Y, which enables "standard" ECC to better deal with errors in X than in Y.
_______________________________________________ math-fun mailing list math-fun@mailman.xmission.com https://mailman.xmission.com/cgi-bin/mailman/listinfo/math-fun