JamesSchumacher
March 28th, 2008, 11:48 AM
If I am not mistaken on this new implementation, I may have created a compression method that creates an output file that in itself can be compressed multiple times. I do not know if this is right, but it seems to be working out that way, and calling Compress() member 4 times results in a file smaller than 7-zip on it's most extreme settings!!!
I need to know if this is correct or not. Please, take the time to look at it, because if it's fine, I have stumbled upon 'dial up how much you want' - literally.
Yes, I know if it's right, I will have to output a length member in the output. That's no problem. However, the theory if this is correct follows the same principle that the dictionary is to be rebuilt on decompression, therefore the dictionary is different for each pass and only applies to that call. With that stated, I may have stumbled upon a method in which the output file is output in such manner that it repeats itself every time, allowing for more compression.
There is an algorithm to match any data set, and this simply might be the way to 'represent' a data set of any length pretty well in multiple calls. (if this is right)
If this is correct, I am going to call it the 'reducing theorem'.
I need to know if this is correct or not. Please, take the time to look at it, because if it's fine, I have stumbled upon 'dial up how much you want' - literally.
Yes, I know if it's right, I will have to output a length member in the output. That's no problem. However, the theory if this is correct follows the same principle that the dictionary is to be rebuilt on decompression, therefore the dictionary is different for each pass and only applies to that call. With that stated, I may have stumbled upon a method in which the output file is output in such manner that it repeats itself every time, allowing for more compression.
There is an algorithm to match any data set, and this simply might be the way to 'represent' a data set of any length pretty well in multiple calls. (if this is right)
If this is correct, I am going to call it the 'reducing theorem'.