v0.5.3

nanoprintf v0.5.3:

Feature: Rework ssize_t definition and remove <inttypes.h> to make nanoprintf work in freestanding environments.

v0.5.2

nanoprintf v0.5.2:

Bugfix: Preprocessor logic bug was preventing the slow non-accelerated "count leading zeroes" path and causing compile errors.

v0.5.1

nanoprintf v0.5.1:

Bugfix: Restore the NPF_NOINLINE qualifier to npf_utoa_rev, fixing a v0.5.0 size regression that I introduced when trimming some function prototype definitions.

v0.5.0

nanoprintf v0.5.0:

Feature: Floating-point overhaul by @Okarss! No more range limitations, subnormal support, double-precision formatting, huge footprint reduction (up to 2KB), configurable integer width (no more uint64_t requirement).

v0.4.0

nanoprintf v0.4.0:

Feature: Floating point rounding. Costs ~100 bytes but space well spent.
Bugfix / Safety: npf_[v]snprintf truncates identically to snprintf. Removes the NANOPRINTF_SNPRINTF_SAFE_TRIM_STRING_ON_OVERFLOW flag and performs string-trim by default.
Bugfix: Don't dereference a string pointer if the precision is zero.

v0.3.4

nanoprintf v0.3.4:

Bugfix: #244 leading-zero flag canceled by precision with integer conversion.

v0.3.3

nanoprintf v0.3.3:

Bugfix: #233 using precision on strings incorrectly behaved like strlen

v0.3.2

nanoprintf v0.3.2:

Feature: Parse %e/%E, %g/%G, %a/%A. They don't format yet, but they're handy for parsing.
Cleanup: Reformat some code for readability, update the header comment to better explain the licensing.

v0.3.1

nanoprintf v0.3.1:

Bugfix: float fractional components were prematurely truncated on some architectures.
Compatibility: fix warnings on the latest MSVC.
Cleanup: Better README.md

v0.3.0

nanoprintf v0.3.0:

License update: nanoprintf is now dual-licensed under 0BSD and Unlicense to minimize encumberance.

Cleanup: One codepath for both GCC and Clang for the printf attribute stuff.