Author: guy haworth
Date: 01:04:12 06/16/00
Go up one level in this thread
I don't know what compression technique Eugene Nalimov uses ... but I think he
finishes up with a file of self-contained 8Kb physical blocks of compressed
data.
For some compression functions 'COMPRESS', I guess you could prove, because of
the way the function works that:
COMPRESS [ COMPRESS (file) ] = COMPRESS (file) ....
i.e. the COMPRESS function is 'idempotent' in matho terms.
As other replies indicate, if COMPRESS-1 is any good, you should not (by
definition) expect to get any benefit out of applying it, or another compression
function, on the resultant compressed file.
As I was just about to confirm this myself, I've just done it:
kppkp.nbw = 33,320KB = 34,119,680 bytes
winzip(kppkp.nbw) = 34,118,774 bytes
winzip(winzip(Kppkp.nbw)) = 34,050,433 bytes
What does this prove?
First, it shows that EN's compression technique is pretty good as winzip doesn't
make much of an impression.
Secondly, it shows that winzip isn't formally idempotent - but again, it can't
make much of an impression on the files it creates.
G
This page took 0 seconds to execute
Last modified: Thu, 15 Apr 21 08:11:13 -0700
Current Computer Chess Club Forums at Talkchess. This site by Sean Mintz.