[comp.virus] Comparing virus scanners

frisk@rhi.hi.is (Fridrik Skulason) (05/14/91)

What follows is a table comparing several virus scanners.  I am
obviously not unbiased, being the author of one of the programs, but
the table was not compiled by me, but rather Franz Swoboda in Austria.

He does not list all anti-virus programs known today - only those on
the market in Austria.

Also, he is testing the programs against his personal collection,
which may not represent accurately the world's virus scene.

The table lists the programs, the "hit rate" and the time required for
scanning.

Findviru 2.0 (S&S)       97%    75 sec
F-FCHK 1.14A             93%   259 sec
Virusdet 3.00 (Puls)     85%   223 sec
Watchdog 4.16 (KDT)      82%   512 sec
Scanv75 (McAfee)         80%   260 sec
Unvirus 3.08 (Elia Shim) 73%    65 sec
Stopvir 2.31             73%    96 sec
Virutest 1.0 (JP Landen) 66%   518 sec
Pro-scan 2.1             65%   184 sec
Norton 1.00              62%   130 sec
VU-Advanced 1.03         61%   479 sec
Viru-Spy 4.0             58%   178 sec
AVsearch 2.23            57%   750 sec
TBscan 1.7               46%   105 sec
TNTvirus 7.00A (Carmel)  45%   190 sec
Virscan 1.45             38%   200 sec

Note that in many cases the products compared are not the latest
versions, but according to Swoboda, they were the latest available in
Austria.

- -frisk

Fridrik Skulason                 Technical Editor of the Virus Bulletin (UK)
(author of F-PROT)               E-Mail: frisk@rhi.hi.is    Fax: 354-1-28801

CHESS@YKTVMV.BITNET (David.M.Chess) (05/14/91)

> From:    frisk@rhi.hi.is (Fridrik Skulason)
>
> ...
> Virscan 1.45             [very bad numbers]
> ...

I hope that's not IBM's VIRSCAN?  If it is, it's a version from *last
June* (and an internal, not a product, version at that).  If that's
"available" in Austria, it shouldn't be.  On the other hand, there are
a few different products in the world called "Virscan", so perhaps
this line is about something else.  It might be helpful if the
manufacturer were listed for all the programs named in lists like
this.  (It's also in general not a good idea to do timing-tests on
infected files; IBM's scanner, for instance, stops for a good
half-second to "beep" when it finds an infected file, which will add
greatly to the timings if infected files are used for tests!  Real
users, of course, probably don't care how long it takes to scan
infected files, just clean (normal) ones.)

DC