Show HN: The new compression algorithm is not like the standard ones

This compresses each line as a delta against its prior line if the delta is smaller. It is not optimal to always pick the prior line as the base.

Good afternoon, I redesigned the compression algorithm, now it is much better.

Have you tried it in realistic examples, like the first 10 likes of "Romeo and Juliet" by William Shakespeare?

Not yet, but there is a test on GitHub.