I'm a totally wrong when i think of the dictionary size as the amount of data a compressor searches for compressable patterns/repetitions?
a: identical bytes sequences father away then dict size does not get compressed together
b: there is no need for a bigger dict then the data you are compression (512mb dict on 64mb file)
If this is correct should REP automatically adjust dict size down of the file is to small to avoid higher then necessary resource usage ?
or is there no differences as the dictionary size is not fully used ?
e.g 1gb dict on 512m file vs only using 512m dict.