Author        : Hamid Buzidi
Name          : lzturbo
Version       : 17 Apr 2009 
Native OS     : win32 + linux
E-mail        : do-not-disturb@hotmail.de
Licence       : Use only for testing and for benchmarks.
Homepage      : 
		
************ USE ONLY FOR TESTING. DO NOT DELETE ORIGINAL FILES ***************************

New: - improved compression/speed
     - new optimal parsing (level 9).

Description:
Lzturbo is a high performance lz77 based compressor featuring parallel compression/decompression and overlapping i/o.
Lzturbo is specially designed for modern processors and multicore systems.

Features:
- Compress/decompress as easy as the copy command. No archive/solid files needed.
- Parallel and linear scalable compression/decompression on multiprocessor/multicore systems.
- Overlapping i/o with compression/decompression.
- Parallel compression/decompression of multiple files.
- State of the art optimal parsing with high compression ratio and fast compression.
- Very fast compression/decompression (No other compression program compress and decompress faster with better compression).
- lzturbo is the only compressor that can use modern multiprocessor/multicore systems at 100% load.

Installation:
Download and unpack the archiv http://lzturbo.web.officelive.com/Documents/lzturbo.zip.

Usage: 
Compression:
   lzturbo -ml [-r] [-p#] <filename1> .. <filenameN> DESTINATION_DIR
	m:compression method (1..5)		1:(fastest decompression) ... 5:(best compression)

        l:compression level (0..9)		0:fastest compression,...,9:best compression
						decompression time same for all levels
						level 8/9 uses optimal parsing.

        p#:					#: number of processors/cores (default=autodetect). 0=disable multithreading
                                                   
        b#                                      #: block size in mb (default 32)

Decompression:
   lzturbo -d -p# <filename1> .. <filenameN> DESTINATION_DIR
					
Compression examples:
      lzturbo -11 enwik9 .		     
      lzturbo -42 enwik9 .
      lzturbo -43 enwik9 backup_dir
      lzturbo -59 file.jpg ab*.txt backup_dir     
      lzturbo -59 A10.jpg english.dic FP.LOG MSO97.DLL rafale.bmp vcfiu.hlp AcroRd32.exe FlashMX.pdf ohs.doc world95.txt backup_dir
      lzturbo -59 sfc\* backup_dir
      lzturbo -59 -r mydir\* backup_dir
      lzturbo -59 -r c:\users\myusr\mydir\* backup_dir
          
Decompression examples:
	lzturbo -d enwik9.lzt restore_dir
	lzturbo -d *.lzt restore_dir
        lzturbo -d -r backup_dir\* restore_dir

replace '\' by '/' for linux.



Lzturbo uses ~15N/2N Mb memory in level 9 for compression/decompression per processor/core (N=block size, default=32mb).

recommended options for benchmarking:
  -10,-11 :         compression time benchmarks (slow hardware)
  -21,-22,-31,-32 : compression time benchmarks (modern hardware)
  -19,-29,-39 :     decompression time benchmarks.
  -43,-44 : 	    Efficiency benchmarks
  -59 : 	    Maximum compression benchmarks (larger block size -> better compression)

FILES:
lzturbo.exe	: win32+Linux version. 
lzturbo		: 32-bits linux (tested on Ubuntu 8.10)
lzturbo_static	: 32-bits linux with static linking  
lzturbo.txt: this file

