a563@mindlink.UUCP (Dave Kirsch) (08/27/89)
Taking a look at the compression war:
COMPRESSION COMPARISON
----------------------
Percentages are the size of the file compared to the orginal size,
i.e. if 37% is listed, the file is 37% the size OF the orginal.
Time in seconds, Size in bytes.
Test done on file TELIX.DOC (text, size: 170,821)
normal -ex
PKPAK PAK PAK PKZIP PKZIP LHARC ZOO PKZIP
V3.61 V1.00 V2.00 V0.92 V0.92 V1.13 V2.01 V1.01
Time 3.407 8.462 14.451 3.407 13.462 21.923 6.923 12.637
Size 63,772 56,227 51,474 58,582 58,582 54,399 64,538 50,423
37.33% 32.92% 30.13% 34.29% 34.29% 31.85% 37.78% 29.52%
I find it very interesting that PKZIP 0.92 did the same ammount of
compression on both normal and EXtended compression (I double checked
and did it again to make sure, it's true).
------------------------------------------------------------------- --
Test done on file TELIX.EXE (binary, size: 279,488)
normal -ex
PKPAK PAK PAK PKZIP PKZIP LHARC ZOO PKZIP
V3.61 V1.00 V2.00 V0.92 V0.92 V1.13 V2.01 V1.01
Time 5.659 18.736 21.429 6.758 14.935 27.747 14.231 17.033
Size 182,392 166,776 132,261 180,413 145,410 135,327 182,189 134,453
65.26% 59.67% 47.32% 64.55% 52.03% 48.42% 65.19% 48.11%
It looks like PKPAK wins in the speed category, though it produces the
least ammount of compression. In text files, PKZIP 1.01 wins with
producing the smallest file, just behind PAK 2.00. But in the binary
file, PAK 2.00 just beats PKZIP 1.01 by 2K.
All tests were done on a 25MHz 80386 running DOS 3.3, no RAM Cache. Drive
was a 330MB Maxtor hard drive at 16 millesecond access time. All tests
were run 3 times then averaged. The drive was not chached. The only TSR
in memory was SideKick (so I could record the results).
Dave Kirsch -- a563@mindlink.UUCP