Viruses are dead. Long live viruses!

This year has been mercifully quiet on the virus front but anyone who reckons the virus problem has finally been beaten is failing to learn the lessons of history.

The problem of computer viruses has been declared "over" before, only to be "reinvented" a few months later, argues David Perry, a marketing manager at Trend Micro.

In the mid 90s, for example, when Microsoft moved to a virtualised 32-bit OS this greatly reduced the potential effects of boot sector viruses. There wasn't much relief for users though, since this threat rapidly was supplanted with the emergence of Word concept viruses.

The dominant Windows monoculture is blamed by many for the prevalence of viruses. If everybody used Macs or Linux boxes, the argument goes, the virus problem would die overnight. This argument is strengthened by much lower incidents of viruses on these platform and the security loopholes in MS applications, like Outlook, that have contributed to the spread of viruses like the Love Bug.

But Perry says that hopes that a different OS will end the computer virus problem are misplaced. A new OS will change but fail to eliminate the problems posed by computer viruses, he argues.

In the late 90s, heuristics (automatic detection) of viruses was offered as a way to eliminate the need to update AV tools by some vendors (many now deceased). Heuristics, though used by almost all vendors, have failed to fulfil such inflated claims. Ssignature updates remain a necessary component of protective measure, according to Perry.

Integrity checking technology or tougher punishment of virus writers are also incomplete solutions to the virus conundrum, Perry argues.

Mass mailers beaten back

This year has seen a marked decrease in the spread of fresh malicious code, as increased user awareness, improved technology and procedures have brought the mass mailing virus plague under partial control. A "chilling effect" post September 11 discouraging VXers from writing malicious code and the maturity of mass mailing viruses have also contributed to the decline of viruses through 2002.

Good - but this doesn't mean we've licked the virus problem, Perry argues.

He suggested some candidates for the next big thing - PDA infectors, .NET wrapper viruses or unusual transport mechanisms - but was reluctant to speculate on which poses the greater future risk.

So, whatever the future brings, it doesn't look like AV developers attending this week's Virus Bulletin conference will have to look for alternative employment in the near fuure. ®