While trying to learn what an "anti-virus" program does (specifically), I came across this:

Known computer viruses grew by 28,327 in 2004 to bring the number of old and new viruses to 112,438, according to IBM.
(from https://www.aladdin.com/home/csrt/statistics/statistics_2005.asp)

How many viruses exist today in 2006? Do "anti-virus" programs check for all of them?

(or...how many virus "signatures" are in the Clam db?)

Thanks!

Please open clamwin and go to help-about - this information is listed there.

Is there a reason why you can't answer directly?

Because this number is constantly increasing and it is relatively easy to find it out using clamwin application.

Checking mine it says it protects from 65656 viruses, taking into account todd's 112,438 viruses in 2005 and probably and increase this year possibly offset by patches to fix vunerabilities and there is still a pretty big gap.
Not that I have had a virus in my testing environment that clam hasn't found (well actually there was one, a real new one, and by the time I sent the email off it had already been included in the latests sig).
So why the huge disparity, is it because IBM counted variants and clam doesn't, is it because some variants can be found with the same signature?

if other av detect the file as virus while clamav not, you can directly submit the sample to the clamav db team using this form:
https://cgi.clamav.net/sendvirus.cgi https://cgi.clamav.net/sendvirus.cgi

be sure you have the latest db definition

Let's cross our fingers...

I can just say that the the number of detected virus variants is not a good indicator for an antivirus effectiveness. Just a few reasons:
* some virus variants counted as one in ClamAV are counted as many in other products and vice versa
* ClamAV database is being populated since 2002 and othe comnmercial AV vendors do that since 1996 or even earlier. They count old DOS viruses that are harmless in Windows.

THE most important factor is quickly the new vireus is included in the database and identified. CalmAv is quite good at that. Some googling will give the stats.

How so exactly? If I can create a batch file to delete your hard drive, you're telling me it's better then them?

the scanning time decreases when number of signatures increases so you should also take this in count

I never implied that a a batch file is a DOS only virus. I was talking about a DOS executable format that does not run or does no harm on windows.

Other than system files (e.g. command.com, keybd.com, etc.), I've never seen simple DOS files that won't launch in Windows (you know, in the same way Windows files don't launch in Dos). I say simple because a virus isn't a game or a full scaled application. It's just meant to cause havoc by messing with or deleting files and I don't see how that wouldn't work in Windows.

most DOS viruses use direct interrupt access that does not work in Windows, you get a VDM error instead.

Why don't you create an executable batch file to del c:*.* or format c: then, and see if it wipes out your own hard drive? If it does, then you can tell us you've added one more virus to the 112,000!

As you get farther away from the early Windows versions (and DOS), batch files aren't needed. I do notice, however, that my AVG seems to go into DOS when it initially boots up.

I haven't seen a boot virus in some time.

Regards,

I don't really care about the effectiveness of viruses on windows, I tend to use Linux and don't run anything as root, so good luck to any virus out there.
But some of my colleuges use that most holey of operating systems, windows... I just found it interesting.
Yeah a lot of things won't work on windows that worked on DOS, simply put DOS was 16-bit winxp is true 32, plus other improvements and a strue of patches. Patches beat a large number of viruses even ones target at windows, so keep your os up to date, wether it be Linux, BSD, MacOSX or the lesser windows.