2.4

• Bug fixes
• Reliability improvements: hardened decompressor against invalid bitstreams, fuzzed decompressor, fixed all known UBs
• Support for 64 bits block checksum
• Stricter UTF parsing
• Improved LZ performance (LZ is faster and LZX is stronger)
• Multi-stream Huffman for faster decompression (x2)

2.3

• Bug fixes
• Performance improvements (ANS, ROLZ, TPAQ, ...)
• New CLI options (delete source, skip dot files or links)
• The decompressor now checks decompressed size vs original size
• Headerless mode (via API)
• Code improvements (C++ linters, portability)

2.2

• Level 4 was too slow and has been replaced: Level 1 to 3 have been promoted and a new Level 1 with no entropy codec has been added
• Rewritten Huffman codec (30% faster)
• Improved LZ codec (better compression)
• Faster UTF transform (>25% faster)
• Improved build (new build targets, fixed compilation with C++98, with VS2017 and C++11, ...)
• Many other performance improvements (ANS, ROLZ, FPAQ, CM, ...)
• Improved CLI parsing
• Bug fixes and many small code improvements

2.1

• Improved LZ codec
• Rewritten EXE codec supports X86 and ARM64
• New UTF codec
• Multi threaded mode is now the default
• New C API
• Various code improvements (portability, C++20 support, improved DNA compression L1-L4, ...)

2.0

• Performance improvements at all levels
• Compression & decompression significantly faster at levels 3, 4, 5, 6, 7
• Reduced memory usage during compression & decompression
• Better multi-threading scalability for very large blocks
• Various code improvements (portability, stability, types, etc...)

1.9

• Level 1 compresses better
• New level 2 to fill a compression/speed gap
• Level 3 compresses slightly better and faster
• Level 5, 6, 7 decompress faster
• Level 8 uses less memory (a bit weaker and faster)
• Partial decompression available (only some blocks)
• Bitstream format frozen

1.8

• Corner cases fixed and code improvements
• Level 1 compresses a lot better
• New codec for some multimedia files added to levels 2 & 3
• Multi-threading rewritten to parallelize entropy (de)coding
• Level 5 faster & level 6 faster (but a bit weaker)

1.7

• Bug fixes and code improvements
• Small compression gains throughout
• Level 1 compresses better (a bit slower)
• Level 6 has been redefined (faster for text files)
• Better handling of small files

1.6

• Bug fixes & code cleanup
• Decompression speed improvements, especially level 4, 5 and 6 (new inverse BWT)
• Better compression ratio at level 1, 2, 5 and 8.
• New Sorted Ranks Transform
• Added support for C++17

1.5

• Two new levels (2 and 8) have been introduced to remove gaps in the compression ratio/time curve.
• Many speed improvements for compression ratios similar to 1.4.
• Better text compression at level 1.
• Inverse BWT is now multi-threaded.