MOO-cows Mailing List Archive
[Prev][Next][Index][Thread]

Re: CRC

>>>>> "Richard" == Richard Connamacher <phantom@baymoo.sfsu.edu> writes:

    >> Cyclical Redundancy Check, I believe.  It generates a
    >> 'signature' for a chunk of data that can be used to easily
    >> check whether two large chunks of data are the same.

    Richard> Hrm..  how accurate is it?  I would imagine that whever
    Richard> you take a large chunk of data and compress it into a
    Richard> tiny signature, it would be possible for two different
    Richard> sets of data, or more, to have the same signature.  I
    Richard> assume it's much faster than checking if the data itself
    Richard> is identical on both sides of the connection :)

    Richard> Phantom

It is faster in that the real work is only done once at the origin of
the data.  It is an information theoretic convolution of the
inforation, so that a given CRC is capable of detecting N bits in
error.  For the purpose here, it is unlikeley that to nearly similar
MOOs would have the same CRC and the same size and the same filedate.
I would not trust the CRC alone to tell me they were the same -- it is
more of a flag to say they are different in this case.  If the crc is
different, the data is most certainly different.  If the crc is the
same, look at other factors to be sure.

-- 
-----------  "...  And the men went up and viewed Ai."  [Jos 7:2]  -----------
Robert Jay Brown III  rj@eli.wariat.org  http://eli.wariat.org  1 847 705-0370
Elijah Laboratories Inc;  759 Independence Drive;  Suite 5;  Palatine IL 60074
-----  M o d e l i n g   t h e   M e t h o d s   o f   t h e   M i n d  ------
References:
Home | Subject Index | Thread Index