WHY_ARJ.DOC September 1991
This document describes the benefits of ARJ. Pardon the commercial.
Compression benchmark results are at the end of this document.
You can find reviews of ARJ in the following magazine articles:
Computer Personlich, June 12, 1991, Leader of the Pack, Bernd
Wiebelt and Matthias Fichtner. In this German magazine, ARJ 2.0 was
named Test Sieger (Test Winner) over six other archivers including
PKZIP and LHA. Compression, speed, documentation, and features were
compared.
PC Sources, July 1991, Forum, Barry Brenesal, "A new challenger, ARJ
2.0, not only offers the speed of PKZIP, but also has the best
compression rate of the bunch."
Computer Shopper, September 1991, Shells, Bells, and Files:
Compressors for All Cases, Craig Menefee. fast
"ARJ ... is extremely
and produces excellent compression; it ... has a rich set of options.
... This is a mature technology, and any of these programs will do a
fine and reliable job."
PC Magazine, October 15, 1991, Squeeze Play, Barry Simon. has
"Jung
combined that foundation with academic research to produce an
impressive product. ... If your main criterion is compressed size,
ARJ will be one of your two main contenders, along with LHA."
1) ARJ provides superior size compression to the other products
currently available on the PC. In a few particular cases, other
archivers may produce slightly smaller archives than ARJ.
ARJ is particularly strong compressing databases, uncompressed
graphics files, and large documents. One user reported that in
compressing a 9.0 megabyte database, PKZIP produced a compressed
file of size 1.8 megabytes, and ARJ produced a compressed file of
size 1.1 megabytes.
2) Of PKZIP, LHArc, PAK, ARC, ARJ, and HYPER, only ARJ provides the
capability of archiving files to multiple volume archives. In
other words, ARJ can archive files directly to diskettes no matter
how large or how numerous the input files are.
It is possible to archive a 10 megabyte file to several diskettes
and to recover the file directly from the diskettes. Other
archivers require that you compress the large file to hard disk
or large RAM drive and then slice the compressed file to fit on
diskettes. To recover the original file involves reassembling the
compressed file on the hard disk from the diskettes and then
extracting the original file from the reassembled compressed file.
This option is not even possible if you lack the hard disk space.
This feature makes ARJ especially suitable for distributing large
software packages without the concerns about fitting entire files
on one diskette. ARJ will automatically split files when

necessary and will reassemble them upon extraction without using
any extra disk space.
This multiple volume feature of ARJ makes it suitable as a "poor
man's" backup utility. ARJ saves pathname information, file
date-time stamps, and file attributes in the archive volumes.
ARJ
can also create an index file with information about the contents
of each volume. Files contained entirely within one volume are
easily extracted using just the one volume.
3) ARJ provides the facility to store EMPTY directories within its
archives. This makes it easier to do FULL backups and also to
distribute software products that come with EMPTY directories.
4) Both ARJ self-extracting modules provide default pathname support.
That means that you can build self-extracting archives of software
directories containing sub-directories. The end user of the
self-extracting archive does not have to type any command line
options to restore the full directory structure of the software.
This greatly simplifies software distribution.
5) The ARJ archive data structure with its header structure and 32
bit CRC provide excellent archive stability and recovery
capabilities. In addition, ARJ is the only archiver that allows
you to test an archive during an archive process. With
other archivers, you may have already deleted the input files
with a "move" command before you could test the built archive.
6) ARJ provides a security envelope facility to "lock" ARJ archives.
A "locked" ARJ archive cannot be modified by ARJ. This provides
some level of assurance to the user receiving a "locked" ARJ
archive that the contents of the archive have not been tampered
with. Data integrity checks contribute to the security of the
ARJ "lock".
7) The myriad number of ARJ commands and options allow the user
outstanding flexibility in archiver usage. This also means
that ARJ requires fewer support utilities compared to other
archivers.
8) ARJ has MS-DOS 3.x international language support. This makes
ARJ more convenient to use with international alphabets.
9) You will also receive strong technical support from a software
author with many years of experience in software technical
support.
COMPRESSION COMPARISON TEST RESULTS September 21, 1991
This benchmark archiver test uses the original PKZIP 1.10 distribution
archive, PKZ110.EXE, as the base data to compress. The PKZIP 1.10
distribution archive totals 302196 bytes. Only compression results
better than 50 percent are included.

ARCHIVER PACKED SIZE COMPRESS TIME EXTRACT TIME
---------------- ----------- ------------- ------------
8088 8MHz PC 20MB 65msec HD
ARJ 2.21 -m4 143981 1:07.7 0:46.8
PAK 2.51 138324 2:26.2 0:53.0
LHARC 1.13c 137450 4:05.7 1:46.1
PKZIP 1.10 136245 2:12.0 0:34.5
ARJ 2.21 -m3 131725 1:30.4 0:48.5
ARJ 2.21 -m2 130994 1:49.5 0:48.1
LHA 2.12 130606 2:39.1 0:50.4
ARJ 2.21 [-m1] 129002 2:14.0 0:47.7
ARJ 2.21 -jm1 128114 2:33.0 0.47.6
386 25MHz PC 64K SRAM CACHE 130MB 15msec HD
ARJ 2.21 -m4 143981 0:07.9 0:06.4
PAK 2.51 138324 0:20.4 0:07.3
LHARC 1.13c 137450 0:26.4 0:12.8
PKZIP 1.10 136245 0:14.8 0:05.0
ARJ 2.21 -m3 131725 0:09.6 0:06.5
ARJ 2.21 -m2 130994 0:11.2 0:06.5
LHA 2.12 130606 0:15.4 0:06.5
ARJ 2.21 [-m1] 129002 0:13.5 0:06.5
ARJ 2.21 -jm1 128114 0:15.5 0.06.5
ARJ 2.21 produced the best COMPRESSION SPEED and COMPRESSION SIZE
results in this test. PKZIP had the fastest EXTRACTION SPEED.
The files were stored on the hard disk without using disk caching.
The archives were created on and extracted from a ram disk.
end document