Revision history for CGI::EncryptForm

1.02 1999-10-25

  - NOTICE: CGI scripts written for versions prior to 1.02 are incompatible
            with this release. See below.

  - Changed the way we flag an error. Instead of returning -1 we now return
    undef. The reason being CGI scripts that have the -w switch turned on
    would produce a warning saying the lvalue of the test condition
    $result == -1 isnt numeric, when there was no error.
    Thanks to Howard Lowndes for bringing this to my attention.

  - Prefixed all error messages return by error() with the method's name
    where the error occured and a unique number within [] to
    cross reference the number against the operation.
    e.g. decrypt [1] Encrypted string is inconsistent.

  - Modified _random_key method to ensure the random key is not persistent
    when run under mod_perl or CGI::SpeedyCGI.

  - Changed the way we autoescape the encrypted string. Instead of URL
    escaping the string, we now use our own character set where by an array
    reference contains from 0 to 255 elements, and each element is 2 characters.
    The elements in the array represent the translation $array->[ord($char)].
    The method charset() is provided to allow you to define your own character
    set. If no character set is defined, the default character set will apply,
    which is suitable for storing the encrypted result in form fields,
    URL's and cookies.  The autoescape method/paramater is now known as
    usecharset.

1.01 1999-10-20

  - Removed dependancy on Math::TrulyRandom because of portability issues.
    Replaced random number generator with perl's builtin rand(). Given
    the purpose of this module, I believe rand() will surfice.
    If you disagree make me understand why and I will look into it further.

1.00 1999-10-18

  - First Release