home
***
CD-ROM
|
disk
|
FTP
|
other
***
search
/
The Business Master (3rd Edition)
/
The Business Master (3rd Edition).iso
/
files
/
utilhive
/
battle
/
compare.doc
< prev
next >
Wrap
Text File
|
1991-08-02
|
9KB
|
289 lines
╔═══════════════════════════════════════════════════╗
║ ║
║ BATTLE OF THE COMPRESSIONS ! ║
║ ║
║ Arj, Pkzip, Lha and Arc go head-to-head ║
║ in an intense, heated battle of compression ║
║ ║
║ ║
╚═══════════════════════════════════════════════════╝
Since the release of ARJ and the new version of LHA, there has not
been an objective comparison of all the current compressions. Obviously,
ARJ has its own comparison within the file ARJ.DOC, but once reading it,
found them to be rather subjective and laxed in explanation and true
experimentation.
Because of this, I ran several extensive tests with all four of the
major compression programs:
ARC 6.02
PKZIP 1.1
LHA 2.12
ARJ 2.2
Other methods, such as PAK, ZOO, LHARC, and the remainder umteenth
compressions were not included because they are out-dated, old and not
widely utilized.
┌─ ─┐
│ARC performed bad enough, I didn't want│
│to horridly embarrass ZOO, PAK, etc ...│
└─ ─┘
The study is split into four groups:
1) Single file, regular compression methods.
2) Single file, self-extracting compression methods.
3) Multiple file, regular compression methods.
4) Multiple file, self-extracting compression methods.
NOTE: LHA and ARJ are the only SFX files included, considering the
rather large size of both ZIP2EXE and MKSARC self-extractors.
In the self-extracting mode, the following commands were used:
a) LHA S FILENAME.EXT
b) ARJ F -JE1 FILENAME.EXT
Also, speed wasn't considered in my experimentation because
all methods rated around .043 second maximum difference.
I did not consider that ratio large enough to justify a
"speed" comparison.
In the comparisons, the boxes are separated into six sections, and
four sections, respectively.
FILE refers to the type of file being compressed. A large document
file would be something like 'ARJ.DOC' or the like.
Large executable file would be 'TC.EXE' from Turbo C++.
SIZE is the original size of the file(s) being compressed.
LHA,ARJ,ARC,ZIP are the compression methods used.
------------------------------------------
THIS ARCHIVE WAS COMPRESSED WITH LHA 2.12
------------------------------------------
=============================================================================
Our first trial involves six different file types of various sizes.
Regular Compression
"""""""""""""""""""
│ File │ Size ║ LHA │ ARJ │ ARC │ ZIP │
├────────────┼────────╫────────┼────────┼────────┼───────┤
│ │ ║ │ │ │ │
│ LARGE DOC │ 87289 ║ 30553 │ 29294 │ 43062 │ 30945 │
│ │ ║ │ │ │ │
│ MEDIUM DOC │ 31399 ║ 11394 │ 11267 │ 15253 │ 11673 │
│ │ ║ │ │ │ │
│ SMALL DOC │ 13723 ║ 5554 │ 5520 │ 7227 │ 5675 │
│ │ ║ │ │ │ │
│ LARGE EXE │ 876480 ║ 438072 │ 430285 │ 616481 │ 462487│
│ │ ║ │ │ │ │
│ MEDIUM EXE │ 211168 ║ 109133 │ 107306 │ 147837 │ 113582│
│ │ ║ │ │ │ │
│ SMALL EXE │ 88834 ║ 48235 │ 47827 │ 65570 │ 51009 │
└────────────┴────────╨────────┴────────┴────────┴───────┘
1) In all single-file compressions, both document and executable were
compressed the greatest by ARJ 2.2.
2) ARJ compressed:
a) Large Doc the best (66.4 %)
b) Large Exe the worst (50.1 %)
Now, once the above files are converted to self-extractors:
Self-Extracting File
""""""""""""""""""""
│ File │ Size ║ LHA │ ARJ │
├────────────┼────────╫────────┼────────┼
│ │ ║ │ │
│ LARGE DOC │ 87289 ║ 32186 │ 34639 │
│ │ ║ │ │
│ MEDIUM DOC │ 31399 ║ 13027 │ 16612 │
│ │ ║ │ │
│ SMALL DOC │ 13723 ║ 7225 │ 10981 │
│ │ ║ │ │
│ LARGE EXE │ 876480 ║ 439705 │ 435743 │
│ │ ║ │ │
│ MEDIUM EXE │ 211168 ║ 110803 │ 112765 │
│ │ ║ │ │
│ SMALL EXE │ 88834 ║ 49904 │ 53286 │
└────────────┴────────╨────────┴────────┴
1) In self-extracting files, LHA 2.12 clearly outperformed ARJ 2.2,
especially in SMALL DOC compression. The docs were 34% tighter
in the LHA files than the ARJ.
2) In the MEDIUM EXE files, both compressions rated about the same.
3) LHA compressed:
a) Large Doc the best (63.1 %)
b) Small Exe the worst (43.8 %)
===========================================================================
Our second trial involves multiple files being compressed using the
regular compression method. Note that the multiple document files
involves 190 text files, and multiple executable files involves
19 binary files.
Regular Compression
"""""""""""""""""""
│ File │ Size ║ LHA │ ARJ │ ARC │ ZIP │
├────────────┼────────╫────────┼────────┼────────┼───────┤
│ │ ║ │ │ │ │
│ MULTI DOC │ 1344961║ 476168 │ 474106 │ 715585 │ 511631│ 190 TEXT FILES
│ │ ║ │ │ │ │
│ MULTI EXE │ 552145 ║ 344284 │ 343849 │ 418390 │ 356818│ 19 BINARY FILES
│ │ ║ │ │ │ │
└────────────┴────────╨────────┴────────┴────────┴───────┘
1) Again, ARJ beats out the others with:
a) Multi Doc 64.8 %
b) Multi Exe 37.7 %
Now, once the above files are converted to self-extractors:
Self-Extracting File
""""""""""""""""""""
│ File │ Size ║ LHA │ ARJ │
├────────────┼────────╫────────┼────────┼
│ │ ║ │ │
│ MULTI DOC │ 1344961║ 477224 │ 474106 │
│ │ ║ │ │
│ MULTI EXE │ 552145 ║ 345863 │ 349194 │
│ │ ║ │ │
└────────────┴────────╨────────┴────────┴
1) And again, LHA defeats ARJ with executable files,
whereas ARJ does best with multiple documents.
=============================================================================
The conclusion of this test is obvious:
ARJ works best when compressing files with the regular .ARJ extension.
However, if you want to create a self-extracting file, LHA clearly
outperforms ARJ 2.2 in all the trials.
╒══════════════════════════════════════════════════════════════╕
│ Using the conclusion: IS ARJ WORTH THE EXTRA 3% DIFFERENCE ? │
╘══════════════════════════════════════════════════════════════╛
Noticing that ARJ 2.2 is superior to LHA in its compression, we
come to the question of "Is it worth it to change ?".
And to answer that, we could take several viewpoints:
1) Any difference is better ! If ARJ gives me 5% better compression,
and the next compression method gives me another 5%, then all that
adds up, especially if I use it to save on disk space.
2) Changing compressions is a hassle, who cares about the 5% difference.
3) I'll change my compression when I get better than a 5% difference,
maybe a 10% or 20% better compression.
and finally, other observations:
1) ARJ 2.2 jams up once compressed by an executable file compressor,
whereas LHA 2.12 does not.
2) ARJ 2.2 is radically larger than LHA 2.12.
3) ARJ 2.2 has many more features than LHA, like sorting files inside,
searching for a text string within the archive, etc.
4) In the .H files, it has M. ADLER and R. JUNG as the creators of
ARJ. Is that a joke ? I'm a psychologist, and think it's kinda
cute, but...
Comments or suggestions ??
Write me:
DAVE SMITH
Compuserve ID 71441,2723