Summary of archivers

Robert Harper harper at NIC.FUNET.FI
Tue Jul 16 06:27:44 EST 1991

For those of you interested in archivers for compressing files on various
computer systems, here is the latest report that I have come across.

*******************************  CLIP   *******************************

     I was asked to include uncompress times along with everything else
     in this posting -- so this is it.  Column 2 of the times is uncompress
     on the 4.77 MHz 8088 machine.

           Packing Test for Pak/PKArc/PKZip/Zoo/LHArc/Hyper/ARJ/ARC

     The files used in the test are given below, and is the entire
distribution of the terminal emulation programme, QVT 3.74.  Here, A
denotes an ASCII file, B is a binary file.  The first two timing columns
given are for compression and decompression on a Zenith 151 running a V20
chip/8087 at 4.77 MHz on a 20Mb fully compressed harddisk, the third is the
compress time for a 12 MHz 80286 clone on a 40Mb fully compressed harddisk.

                       DEFAULT  VT      3159    A
                       QVT      HLP    17386    A
                       QVT_BBS  EXE   169929    B
                       QVT_KB   MAP     2219    A
                       READ     ME     12444    A
                       RELNOTES 374    20342    A

 Compression Tool    Extension       Time Taken     Final Size    % Compression
 ----------------    ---------    ----------------  ----------    -------------

  PKArc 3.61            ARC       0:43  0:36   7.6    146,507          35.0%
  Zoo 2.01              ZOO       1:41  1:05  18.1    147,819          34.4%
  Zoo 2.10              ZOO       4:03  1:26          148,250          34.3%
  tar/compress (16 bit) xxx       2:00  1:29          150,157          33.4%
  cpio/compress ( " )   xxx       2:26                149,927          33.5%
  SEA ARC 6.02          ARC       1:54  1:09          152,916          32.2%

  LArc 3.33             LZS       3:05                123,363          45.3%

  PKZip 1.10            ZIP       1:44  0:30  16.6    109,335          51.5%
  Pak 2.51          PAK/ARC/ZIP   2:35  0:55  25.7    109,019          51.7%
  Hyper 2.50            HYP       2:57  2:24  23.6    109,771          51.3%

  LHArc 2.12            LZH       3:07  0:51  26.1    104,488          53.7%
  ARJ 1.10              ARJ       5:28                104,634          53.6%
  Zoo 2.10 (high comp)  ZOO       7:34  1:20          104,804          53.5%

  ARJ 2.10              ARJ       2:41  1:01  24.0    102,692          54.5%

     There are now essentially five groupings of archivers with the new,
fast release of ARJ 2.10.  Best compression/time combination yet.  Very
impressive, especially with a 50% improvement in compression speeds over
the previous (1.xx) release.  Now, we will have to see if LHArc can catch
up on its next release, since it is seems unlikely Phil Katz will do any
more on compression.  Pak and Hyper are out of it, unless a dramatic speed
improvement can be realized.  I do not expect this.  ARC from SEA is
essentially useless compared to the others -- the old ZOO has better

     Current recommendations are PKZip and LHArc, in terms of compression,
speed, and usage.  ARJ is better, but must become more popular in usage
before I can give it an enthusiastic recomendation.  Note that LHArc and
ARJ both are capable of compressing LZEXE files even further -- PKZip is

     The new version of ZOO is a surprise, especially with the slow time
for the default and the high compression otherwise.  The speed is a bug to
me and has to be fixed.  It might be important to note that ZOO is
available across nearly all computer types -- IBM-PC, Amiga, unix, VAX/VMS.
PKZip is available on Amiga as well as PCs, unZip only on most systems.
LHARC exists on unix as well as DOS, as does ARC 5 (not 6 yet, I believe).
I have not seen ARJ outside of the DOS world.

     So, when do we reach the theoretical maximum compression?

Original post by:
Arnold Gill --- astrophysician in training

More information about the Bio-soft mailing list

Send comments to us at biosci-help [At] net.bio.net