Comment by underdeserver
Comment by underdeserver 18 hours ago
Technically correct, but a better benchmark would be a known compressor with an unknown set of inputs (that come from a real-world population, e.g. coherent English text).
Comment by underdeserver 18 hours ago
Technically correct, but a better benchmark would be a known compressor with an unknown set of inputs (that come from a real-world population, e.g. coherent English text).
Yes, definitely. Alas, it's just harder to run these kinds of challenges completely fairly and self-administered, than the ones where you have a fixed texts as the challenge and add the binary size of the decompressor.