Yes, that's intentional - we have changed this in 4.1, but it's not
documented in the manual yet. A quote from the developer working on that
code:
So, the PASSWORD() function is now not to be used for passwords? The
problem is that I have built at least a few applications that use
PASSWORD() as
-BEGIN PGP SIGNED MESSAGE-
Hash: SHA1
On Thu, 19 Dec 2002 [EMAIL PROTECTED] wrote:
> >Description:
> with the latest 4.1 tree (from today) the PASSWORD() function returns random
>alpha-numeric text /[a-f0-9]/
> 45 characters in length (which is too long for a password string)
