Sunbeltblog comments
I have been in the business of testing for over 30 years, and it is simply not feasable to create anything that is dangerous to do testing with.
Testing is about ethics, and doing something unethical in the name of testing a product just is wrong.
Suppose that I was testing devices that sample for poisonous gasses, would I create new poisonous gasses to test these products with. Funny but I see a strange similarity here.
Alex, I have to disagree a bit here. the point of using new variants of existing viruses is to test the fall-back heuristic rules that detect unknown viruses, retrospective testing ALMOST accomplishes this, but there is one difference, the virus writing community has had say three months to pound on your old heuristic rules and have learned a thing or two. Granted the issue of these virus strains being leaked is a valid one, but nowhere near insurmountable. After all, nobody is leaking Mcafee development source code. shutting down code leaks is a function of electronic and physical security which of course is a function of the number of possible leak points and the reliability of the security measures used against them. For a short time, and in an isolated network with a very few people having physical access to the system, code containment can be accomplished.
I can unserstand your argument, however, normally as part of the testing all methods are documented. Now the software vendors software that is tested, should have access to the methods and items that wer used, so that they can do some 'failure analysis,' otherwise the purpose is, just to say which is better. That may be what is wanted, but it won't accomplish any better processes or code.
So in that case, if it is released for evaluation, there would be a much better chance of it escaping into the wild.
I personally would want to know which types of viruses were caught and which were missed.
The last point being about the test virus developers, are they active in the hacker community, that they can guess where the trends are really going? So then the validity of their viruses is also in question.
As an aside this is virtually the same problem that USAMMRID and CDC have in testing vaccines against genetically engineered biological viruses or viruses that have mutated in the wild. Ever wonder why your flu vaccine doesn't work sometimes? It's because it was designed to optimally target one specific strain. it may have limited functionality to similar strains, but as the strains get less and less aloike, the vaccine looses effectiveness and eventually becomes worthless.
PennWrite's analogy is bogus because a poison gas does not mutate. Humans have to modify it and they would likely attempt to modify it in only two ways, make it more potent, or make it harder to detect. If you are talking about a detector for nerve agents where you may encounter gasses that are similar to known organophosphates but may be different enough to be considered a new one, your detector will have to have broad enough detection window to detect unknown gasses, but that may mean false positives when experiencing organophosphate based insecticides (which depending on which one, might as well be considered to have military value since some of them are VERY potent and long lasting.).
Aren't the computer viruses mdified by humans to be either more potent or to be harder to detect. There are several ways this can be done, so if the samples are modified one way and the 'real' gases are modifed using a different method, they may not be detected.
And the same result is accomplished, where the samples used during the test should be supplied to help correct any errors. Which of course spreads the availability of new gases or viruses(man made).
To the average consumer (like me) the response by McAfee and others just sounds like sour grapes or being over defensive. Got caught with your pants down? Suck it up and develop beter heuristic tools to keep us from being infected by the next bad virus or worm.
The Consumer Reports article raises awareness about the limitations of software and should awaken the reader to the need for consistent vigilance: just because you have the latest anti-spyware / anti-virus software installed, it doesn't mean you can let your guard down and open any web site or attachment without worry.
hmmmmm.... 5500 new viruses, does this mean ConsumerReports is a VIRUS....
I support Consumer Reports and find it disengenous that McAfee would whine so long and loudly. Consumer Reports has no ax to grind. It does not accept advertising and it buys those things it tests, just as Joe Average would. I find no problems with its findings and I suspect the longer companies whine about CR, the more suspect those companies will seem to the general public. I know that I'm watching and will base my own future purchases on just such.
Consumer Reports has published bogus "analysis" for years. I first became aware of it back a few decades ago when they downgraded the professional grade Nikon F2 because the image it showed in the viewfinder was exactly what would be captured on the film instead of what would be printed on a cropped 3x5 "drug store" print.
They haven't improved with age.
Unswayed by both sides on this one.
Commenting by HaloScan
