Text 7679, 49 rader
Skriven 2010-05-10 07:36:18 av Robert Bashe (2:2448/44)
   Kommentar till text 7664 av Ward Dossche (2:292/854)
Ärende: No blanket right for press to protect sources
=====================================================
Ward Dossche wrote to Bob Ackley on Sunday May 09 2010 at 19:52:

BA>> Given sufficient resources, anything that has been encrypted by a
BA>> human (and intended to be decrypted) can be decrypted by someone it
BA>> wasn't intended for ...

WD> I beg to differ.

And I think there must be a misunderstanding here.

WD> The US cryptograph people have never been able to break Russian code,
WD> not even once. Their resources were virtually without a limit.

Their _current_ resources. A cypher is only considered secure if it cannot be
broken using _current_ technology within a "reasonable" period of time,
"reasonable" being subjective and depending on the importance of the
information one is trying to obtain by decryption.

But _no_ cypher is "unbreakable", given enough resources and time - and
considering the advances in computer technology that are made year for year. A
cypher that was considered "secure" in 1990, with the technology available at
the time, is probably not considered secure nowadays, with current technology.

I'll give you an example: I bank via the Internet using a German system known
as HBCI (Home Banking Connection Interface). This is a public key cipher, wth
the secret key stored on either a floppy disk or a chip card. The cipher is
used to identify the customer to the bank's server and to encrypt and decrypt
the traffic between the customer's computer and the bank's server.

The public and private keys were 256 bytes long for many years, but last year,
in view of the fact that advances in technology had made this key length
insecure, the length was increased to 1024 bytes.

This will be secure in the sense that the time and equipment needed to break
the cypher just to access my personal accounts will not be justifiable or
possibly even practicable in at least the next few years. But it won't be
secure for all time - the next jump, when it comes, will probably be to 2048
byte key lengths.

No encryption method is secure for all time. It's only safe to talk about the
immediate present in such matters, and even then the question is open as to
whether more effective equipment and methods are available than one currently
realizes - governments (and criminals) don't tend to advertise their
capabilities, for good reasons.

Cheers, Bob

--- GoldED+/W32 1.1.5-0613
 * Origin: Jabberwocky System - 02363-56073 ISDN/V34 (2:2448/44)