Could anyone recommend a broad scoped evaluation of current compression / archiving formats / algorithms which explores their various merits and failings?