Path: utzoo!utgpu!news-server.csri.toronto.edu!cs.utexas.edu!uunet!world!robjung
From: robjung@world.std.com (Robert K Jung)
Newsgroups: comp.compression
Subject: Reply to ARJ and Jung are not all they seem
Message-ID: <1991Apr24.215923.15481@world.std.com>
Date: 24 Apr 91 21:59:23 GMT
Organization: The World Public Access UNIX, Brookline, MA
Lines: 61


 This is in response to the comments from Mitsue Osanami about ARJ 2.00.
 
>From: mitsue@csd4.csd.uwm.edu (Mitsue Osanami)
>Subject: ARJ and Jung are not all they seem
 
>People should not be getting too excited about ARJ 2.00 or any other...
 
>90% of the code used in ARJ is "stolen" directly from LHA.  I do not think it
>at all appropriate for Jung to be putting a copyright and "all rights
>reserved" on code that wasnt' written by him.  The code is the SAME in
>most parts of the code down to formatting and a few comments.
 
 To clarify, the file archiver ARJ 2.00 is based upon the public domain
 work of Haruhiko Okumura released in AR002.EXE.  ARJ is NOT based upon
 the LHArc 2.1x source.  ARJ 0.20 came out weeks before Yoshizaki
 released his source code.
 
 The AR002 program consists of a grand total of 839 lines of source and
 #include files.  The source for ARJ 2.00 consists of almost 5,000 lines
 of source and #include files.  This means that AR002 can be NO MORE
 THAN 18 PERCENT of the ARJ code!  In fact, less than TEN percent of
 AR002 can be found in the ARJ code.  UNARJ looks like AR002 because
 UNARJ is stripped down to its bare bones.  The regular UNARJ is over
 3,000 lines of source.
 
 Moreover, many VERY IMPORTANT details of ARJ are different from AR002.
 
 a)  32 bit CRC
 b)  C Text mode support
 c)  Archive and file comment support
 d)  Recoverable header design
 e)  Cross platform archive support
 f)  Multiple volume archive support
 g)  Encryption support
 h)  More compression using less memory!!!
 i)  and more!
 
 It is for this original work that the word "copyright" is stamped upon ARJ.
 
>If you will go back to ARJ version .20 the results from it will be BYTE for
>BYTE the same as that from AR2!
 
 ARJ 0.20 uses the AR002 LZ77 compressor.  ARJ 2.00 DOES NOT!!!
 
>The only advancement that he has made is increasing the dictionary size to
>22k!  This guy deserves no recognition.
  
 AR002 CANNOT even be built with a 22K or greater dictionary.
 
 The UNARJ decompressor may be very similar to AR002, but the LZ77
 compressor used in ARJ 2.00 is COMPLETELY NEW, UNPUBLISHED, and
 probably PATENTABLE.  (U.S. patent rights are currently being sought).
 How else could a NOVICE PC programmer in his first shareware project
 outdo LHArc 2.12 in COMPRESSION SIZE and COMPRESSION SPEED!!!
 
 Besides, in my ARJ acknowledgements, the credit goes to God Almighty 
 and to all of the ARJ testers.
 
 Sincerely,
 Robert K Jung  (robjung@world.std.com)