Dnia Thu, 11 Apr 2013 13:18:10 +0200, Wojciech Sura napisał(a):

> Patrzyłeś tu?
> http://en.wikipedia.org/wiki/Cyclic_redundancy_check
#Commonly_used_and_standardized_CRCs

Dzieki, jakos tam predzej nie zajrzalem, a to mi temat nieco rozjasnilo.
Okazuje sie ze "moja" (DEC) implementacja korzystala z alternatywnego
wielomianu. Opisuje to komentarz w jednym z plikow z "Jedi Library":

Description:
Defines a new "generator polynomial" and starting value for the 32 bit CRC
functions. Beware that the check and the correction only works if the
polynomial is "appropriate". Two of these polynomials are predefined:
Crc32PolynomCCITT = $04C11DB7 represents the polynomial x^32 + x^26 + x^23
+ x^22 + x^16 + x^12 + x^11 + x^10 + x^8 + x^7 + x^5 + x^4 + x^2 + x + 1
that is suggested by the CCITT and used by Ethernet and ZIP.
Crc32PolynomXXXXX = $9DB11213 represents the polynomial x^32 + x^31 + x^28 +
x^27 + x^26 + x^24 + x^23 + x^21 + x^20 + x^16 + x^12 + x^9 + x^4 + x + 1
which is another "appropriate" polynomial (as far as I know). Beware that
not appropriate polynomials result in highly undesirable behaviour of the
CRC.

Nie wiem co dokladnie autor mial na mysli piszac o <another "appropriate"
polynomial> ale skoro w DEC tez z tego wielomianu skorzystali to musi byc to
jakis "inny" standard.

--
pozdro
R.e.m.e.K