The Perl Toolchain Summit needs more sponsors. If your company depends on Perl, please support this very important event.
Revision history for Perl module Alt::Crypt::RSA::BigInt

0.06  2017-04-26

  - Typo fixes from dsteinbrunner.

  - Use new Math::Prime::Util with a CSPRNG, remove Bytes::Random::Secure.

  - Add check in key gen for p/q distance (FIPS, ANSI), for size over 256.

  - Use BPSW + Frobenius + 3 random-base M-R instead of BPSW + 46 fixed M-R.
    No practical difference, but a little faster.


0.05  2013-02-22

  - Fixed some perlcritic issues -- mainly returning undef and changing to the
    three argument open using a lexical file handle.

  - Doh -- Alt's package name was Bigint not BigInt.

  - Speed up testing for Calc backend (only do 384-bits in test 11 if not
    using GMP or Pari backends).

  - Switch to Bytes::Random::Secure's OO interface, and allow plugging in a
    different irand function to the generator.

  - Use the new generator interface in all tests.  The suite went from
    3328 bits of blocking randomness to 1536.


0.04  2013-01-14

  - Change from random_maurer_prime to random_nbit_prime.  This is much faster
    and yields more possible primes (Maurer's algorithm only outputs about 10%
    of possible primes in the range).  The downside is that the primes are not
    proven prime, however they will have passed a strong BPSW test and a set of
    Miller-Rabin tests for the first 46 prime bases.  The BPSW test alone is
    really enough for these sizes (~600 digits or smaller).

  - Ask for Math::BigInt 1.78 to ensure we have 'try'.
    Thanks to brian carlson.

  - Reduce key sizes in 06 and 11 when not using GMP or Pari.  Some build
    systems were timing out (e.g. ActiveState).


0.03  2012-12-30

  - Add tests for SHA384 and SHA512.

  - Add optional support for RIPEMD160.

  - Add more data to meta file, including git repository.

  - All tests converted to Test::More.

  - Rewrite Key/Private/SSH.pm, and add tests.  I don't think it ever worked.
    Now supports IDEA, DES, DES3, Blowfish, Twofish, CAST5, Rijndael, RC6, and
    Camelia (only the first 4 are guaranteed -- the others depend on having a
    module installed).  Multiple alternative modules are allowed for some.
    This fixes:

     RT 69376  Bug in Crypt::RSA::Key::Private::SSH serialize function

  - FindBin is no longer used.  You'll need to do:
       perl -Iblib/lib -Iblib/arch ...
    to get the version in this directory.  'make test' does this for us.
    use 'prove -l -b' to prove using the local version.

    I think FindBin is great for example files and such, but I don't think
    it is appropriate in the .pm files or in tests.  If we do use it, it has
    to be added to the dependency list.


0.02  2012-12-28

  - Fix a decrypt issue setting keys

  - Add SHA224 support

  - Added strict and warnings to all .pm files.

  - Converted RSA.pm to UTF-8 and added POD directive

  - Make pretty maurer prime generation output with Verbosity.


0.01  2012-12-28

  - Add test for h2osp, pss sign and verify with salt.

  - Faster versions of some of the helper functions.

  - Switch from Digest::SHA1 to Digest::SHA, and add SHA256.

  - The following Crypt::RSA 1.99 defects have been fixed or made irrelevant
    due to no longer using Math::Pari:

     RT 52689  consider moving Crypt::RSA to Math::BigInt?
     RT 76655  Version 1.99's META.yml says version is 1.97
     RT 63007  Move to later version of Math-PARI
     RT  4877  Crypt::RSA test 11 fails with a segmentation fault
     RT 61392  Crypt::RSA::Key::Private::check not checking that q is an integer
     RT 64883  Make errstr from check after write more intuitive
     RT 45533  Crypt::RSA bug in mgf1 function

    These fixable possible defects remain, but should be resoved soon.

     RT 69376  Bug in Crypt::RSA::Key::Private::SSH serialize function

    as well as these that aren't clear:

     RT 50356  Crypt-RSA-1.99
     RT 59923  changing the passphrase on key

  - Makefile.PL will look for GMP and Pari, and install the Math::BigInt::GMP
    or Math::BigInt::Pari modules if possible.  It complains otherwise, but
    everything should still work -- it just runs really slow.

  - Completely removed Pari.

  - Replaced modules:

       Crypt::Random  =>  Bytes::Random::Secure
       Crypt::Primes  =>  Math::Prime::Util

  - Started with Vipul's Crypt::RSA.