I'd say the bigger difference is the lack of homophones in a compressor.

Languages add new words constantly, albeit slower than a computer does compressing a new file.