George James wrote:
> 
> Tim Churches wrote:
> >Of course, future computers might be much, much faster, or a 160-bit
> >quantum computer may be able to examine all the possibilities for a
> >160-bit symmetrical key all at once and crack it instantly...but I am
> >not worried yet. The point is that if secrets need to be maintained,
> >then one must be prepared to re-encrypt them using better algorithms
> >and/or longer keys as the decryption technology advances. You should not
> >assume that once encrypted, secrets will be safe forever, although with
> >a reasonable choice of key length, you probably only have to worry every
> >10 years or so. But where are the systems to remind us of things which
> >were encrypted 10 years ago?
> 
> Of course, if you encrypt information using current key lengths, then
> re-encrypting it in the future with a stronger key would be pointless
> if the more weakly encrypted version was still available.   Once it's
> out there, it is out there.

Indeed! I think I was thinking of private medical databases - at least
that's my excuse - but the context was in fact encryption of publically
accessible databases, in which case my suggestion about re-encrytion is,
as you point out, bollocks.

Cheers,

Tim C

Reply via email to