On Fri, 21 Aug 2015 09:53:49 -0400, Thomas David Rivers wrote:
>
>So - perhaps I'm confused... but are you saying you want
>a procedure that potentially reads the entire data set to determine
>if you need to read the entire data set?
>
If the hash is sufficiently strong, you needn't read the entire data set
more than once, nor preserve the older version(s), but only the hash.
If a collision is unlikely within the lifetime of the universe, don't worry.
OTOH:
.... Ron Rivest estimated in 1977 that factoring a 125-digit
number would require 40 quadrillion years, ...
but with technological advances such a problem was solved 17
years later.
https://en.wikipedia.org/wiki/The_Magic_Words_are_Squeamish_Ossifrage
-- gil
----------------------------------------------------------------------
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to [email protected] with the message: INFO IBM-MAIN
