0.09
 - Fix bungled NV_IS_FLOAT128 code in Rmpfi_set_NV, Rmpfi_get_NV and overload_equiv.

0.08
 - Amend typedef of __float128 alignment (Windows only)
 - Replace SvUV/SvIV/SvNV with SvUVX/SvIVX/SvNVX where appropriate. (In typemap, too.)
 - Alter NV overloading - replace the condition (SvNOK(b)) with (SvNOK(b) && !SvPOK(b))
 - Add Rmpfi_set_NV and Rmpfi_get_NV.

0.07
 - Further tweaks for builds where nvtype is __float128

0.06
 - Accommodate $Config{nvtype} eq __float128
 - No longer define USE_64_BIT_INT if ivtype is 'long'.
 - Rename USE_64_BIT_INT #define to MATH_MPFI_NEED_LONG_LONG_INT
 - Define IVSIZE_BITS if MATH_MPFI_NEED_LONG_LONG_INT has been defined.
 - Version 3.23 of Math::MPFR now a pre-requisite.

0.05
 - Alter the way that strEQ gets hold of HvNAME(SvSTASH(SvRV(sv))) - so that -Wall
    doesn't emit "null argument" warnings. (Thanks Daniel Kahn Gillmor.)
 - _has_inttypes() no longer checks for whether USE_LONG_DOUBLE is defined.
   (Was harmless ... but didn't make much sense.)
 - Define PERL_NO_GET_CONTEXT
 - Tweak to Math::MPFI::gmp_v()
 - Add metadata to Makefile.PL

0.04
 - Fix t/tls.t

0.03
 - Remove 'not' from overloads. Apparently achieved nothing anyway, and 5.16
   warns that it's invalid.
 - Add MPFI_VERSION_MAJOR, MPFI_VERSION_MAJOR, MPFI_VERSION_PATCHLEVEL
   and MPFI_VERSION_STRING (mpfi-1.5.1 and later only).
 - In aab_basic.t, check that library and header versions match
   (mpfi-1.5.1 and later only).

0.02
 - Add documentation of gmp_v and mpfr_v (previously missed).
 - In aab_basic.t print out the mpfr and mpc library versions being
   used, as reported by the library in use.
 - In MPFI.xs change mpfr_v() to return the library version instead
   of the mpfr header version (as had previously been the case).
 - Minor doc fix.

0.01
 - First release.