I wouldn't get melodramatic over the fact that this course is being offered at Calgary. I have some solid doubts that any of the major US universities will be offering such a course due to concerns over substantial legal liability, if nothing else. However, a case could be made that carefully selected students should learn how to write viruses as part of a course or a series of courses on how to design anti-virus software - The easy part is killing the virus once the virus's signature has been recognized. The hard part is designing the heuristic that wiil recognize that virus's signature. Add to to that the fact that some viruses come with a built-in library of signatures, and we could have quite a potboiler.

The viruses that are hardest to detect are those that are designed to be dormant until a set of circumstances are met, at which time they wake up quietly and do their work even more quietly. Those are the viruses that I fear the most, because no one has any idea how long ago it infiltrated successfully - and thus how many backups it has successfully corrupted. I am not worried about those viruses that do immediate, spectacular damage, as thoroughly stomach-churning as the experience may be to those of us who make a living in the kill zone.

The moderator should delete this message if he or she believes that it carries a risk element. However, I believe that these concepts are already known to serious writers of viruses.

[ reply ]

Link to this comment: http://www.securityfocus.com/comments/columns/164/20274#20274