"Binsort - sort files by binary similarity": Thanks for the info! Yes, this seems to do something similar to what I implemented, on a file level (not chunk leve

"Binsort - sort files by binary similarity": Thanks for the info! Yes, this seems to do something similar to what I implemented, on a file level (not chunk level). It seems to be O(n^2), while my algorithm is O(sort(n)). But having a standalone too is great! lrzip / rzip use 900 MB a window, which used to be huge at that time it was written (1998?). But nowadays it is considered medium size.