Re: change password API -- MIDRANGE-L

Tim, you do not need to be able to recover one on the
other end.  Consider.  I encrypt the password on the PC
using the same encryption IBM uses in OS/400 (if I knew
what it was).  I then send the result of this, the encrypted
passed, to the AS/400 through the network.  On the AS/400
I then compare this string to the string stored un my user
profile for my encrypted password.  I never had to recover
it on the AS/400.  The AS/400, in theory, does not know
what I originally typed in, just what it encrypted to.

Now, as for "nothing that can't be decrypted".  Lets take
a very simple encryption scheme.  You give me any word,
and for every letter in it I will put a 1 or a 0.  A 1 if it is
an odd number of the alphabet, a 0 if it's even.

so CAT = 111
DOG = 001
etc...

As you can see, there are flaws to this simple scheme, as
many different words will produce the same encrypted
string.  But, the point is, you tell me, what password is 111 ???

If something is never meant to be decrypted, but only compared,
it is possible to make an unencryptable string that will produce
the same result from the given text.

The flaw comes in from the password comparison.  I just keep
trying different strings and encrypting them until they compare
to the original.  But, the fact of the matter is, I never did really
decrypt the original encrytped string.  I just used a brute force
method and tried every possible string until one matched.

If I am allowed enough characters in my password, say 100
or so (not sure the length of a PGP password is) it would take
you enough time to try every possible combination that it can
be considered to be "un-decryptable".  And that is not "garbage".

Regards,

Jim Langston

Tim McCarthy wrote:

> Jim,
> Firstly, whether OS/400 ever decrypts the password or not is irrelevant
> to the situation in hand. If I need to avoid sending a plain text
> password over an unprotected channel then I need to encrypt either the
> actual password or the hash and I need to be able to recover one of
> these values at the other end. Secondly, the basis for any cryptosystem
> is that some standard plain text produces a certain output as the result
> of a known transformation, otherwise it's plainly useless. And as any
> cryptographer will tell you there's nothing that can't be decrypted,
> it's just a matter of the time and effort required to do so.
> Thus..."Anything that can't be decrypted is garbage."

+---
| This is the Midrange System Mailing List!
| To submit a new message, send your mail to MIDRANGE-L@midrange.com.
| To subscribe to this list send email to MIDRANGE-L-SUB@midrange.com.
| To unsubscribe from this list send email to MIDRANGE-L-UNSUB@midrange.com.
| Questions should be directed to the list owner/operator: david@midrange.com
+---