Revision history for Perl module Data::Float::DoubleDouble

1.09
   - Add get_doubles function.
   - Fix croak() message in DD2HEX XSub. (Replace '%' with '%%' where required.)

1.08
   - Add dd_bytes function.
   - Remove valid_hex, valid_bin, valid_unpack - silly ideas in the first place

1.07
   - Change template for NV2H() and H2NV() from "D<" to "F<", and provide LD2H() and
     H2LD() - which use the "D<" template.
   - Rewrite _calculate as an XSub. (I don't trust perl to be always capable of calculating
     2 ** k correctly.)

1.06
   - Add NV2binary and express() functions.

1.05
   - Add DD2HEX and std_float_H functions.

1.04
    - Add some metadata to Makefile.PL.
    - Small rewrite of float_B to avoid new warning scheduled for 5.22

1.03
    - Add some more tests.
    - Add validate_hex, validate_unpack, validate_bin
    - Add float_class, float_is_finite, float_is_zero, float_is_nzfinite, float_is_normal,
      float_is_subnormal, nextafter, nextup, nextdown for compatibility with Data::Float.
    - Add DD_* vars to reflect the float.h values for the various LDBL_* macros (and for
      FLT_RADIX).
    - Define PERL_NO_GET_CONTEXT

1.02
    - Update version number to the correct value (1.02) ... as opposed to what
      happened wrt 1.01.

1.01
    - INCOMPATIBLE WITH EARLIER VERSIONS !!
    - Refactor, including coverage of those double-doubles whose values require more
      than 106 bits for accurate representation.

0.03
    - Alter float_H to, in list context, return (single) sign and (single) exponent, in
      addition to the hex string and binary string it was already returning in list context.

    - Add some more test files.

0.01
    - First release.