Lpc2000

At 06:45 PM 1/10/06 -0800, Mike Nelson wrote:
>The example given by IAR only does CRC16 (0x11021)
>with no complement and MSB first.  For any reasonable
>span of code, one should really use CRC32, as it is
>less likely by far to alias.

Well any 16 bit check has approx 256 times fewer aliases than an 8 bit 
checksum.  However a 16 bit CRC and a 16 bit arithmetic checksum have 
exactly the same number of aliases.  The big advantage of the CRC is its 
sensitivity to burst errors.  Whether that is an advantage for flash is, I 
think, an open question.  At least I haven't seen anything that shows the 
error distribution in flash.  And for uploads I could argue either way 
depending on the character of the changes.

Robert

" 'Freedom' has no meaning of itself.  There are always restrictions,   be 
they legal, genetic, or physical.  If you don't believe me, try to chew a 
radio signal. "  -- Kelvin Throop, III
http://www.aeolusdevelopment.com/