On Steep Compression.

Team. We have spoken in the past on this "CABOOSE plus computing idea" weblog about a simple data compression and decompression methodology called DNA ( Distinct Number Algorithm ) Storage. A letter was written the NSF strongly suggesting that they stop wasting money on researching storage techniques using deoxyribonucleic acid and adopt this simpler approach.

While driving this day, a more practical notion of annotating a file with stop symbols for decompression occurred. The file itself is a maker with unique patterns in it. One might take samples of a given size from random parts of a file.

If one had a megabyte that he were compressing, he could sample sixty-four eight byte span and store them. This would be half a kilobyte of information for uniquely identifying the decompressed image. If the compressed image, plus any metadata composed another half kilobyte, the final descriptor for the megabyte would consume a kilobyte.

If this megabyte were the fortieth one in a gigabyte, the kilobyte descriptor would be the fortieth in a compressed description of the gigabyte. Whose final size would be a megabyte. Then, this megabyte, an image of a gigabyte, could be made a kilobyte. The process becomes recursive. So, a reasonable challenge problem for a college computing class would be producing an algorithm, based on the intuition of DNA Storage, that placed a megabyte in a kilobyte. One would imagine that the "top-notch" engineers at Google could draft such a routine in an afternoon.