Let's not scrimp on bit space

From: Karl Auerbach <karl_at_cavebear_dot_com>
Date: Wed May 05 2004 - 12:47:59 CDT

On the bar code issue:

I was in Mexico City when Yogan Dalal made the presentation of the
Ethernet addressing scheme - 48 bits. We all thought - 48bits! that's
huge. Today the IEEE is moving to 64 bits.

I remember when we thought that IPv4's 32-bits were more than enough. Now
we are moving to IPv6 with 128 bits, and there are those who consider that
too small.

I find it unwise to scrimp on bits - I'd use a 128 bit number to represent
ballot "style" identifier to indicate the exact date, election, precinct,
ballot-within-precinct. And I'd use at least 64 bits for a digitallly
signed message digest of the encoded results. And I'd use at least
another 32 bits for a message digest/CRC to protect against end-to-end
software encoding-decoding errors.

With numbers like these, a 1-D bar code is exhausted even before election
data is inserted.

There seems to be a belief that 2D bar codes are significantly more
expensive. I've been researching the issue and that seems not to be the
case.

On the printing side there is no cost difference.

On the scanning side the cost difference is not worth hassling over. Once
you get past the cue-cat level of scanners and get into things that are
robust and reliable, the cost difference of scanning 1D versus 2D is
small. And as equipment is driven to become both 1-D and 2-D capable, the
price difference will become negligable.

                --karl--
==================================================================
= The content of this message, with the exception of any external
= quotations under fair use, are released to the Public Domain
==================================================================
Received on Mon May 31 23:17:16 2004

This archive was generated by hypermail 2.1.8 : Mon May 31 2004 - 23:18:15 CDT