Desktop Security Software – Testing Standards
Antivirus programs are constantly changing, and, like anything, simply because antivirus is on the peak of the desktop security program market one year doesn’t guarantee it will stay there. So how can it be that people decide on what makes the best antivirus programs? If you go to some review website, the allusive’evaluation’ is known as marking this year’s top-class program. But if you’ve ever tried to use these results spreadsheets to ascertain which stage you should go with, you know they do very little good at emphasizing one maker Either this means that a lot of the large players are essentially the same, or something is wrong with how the tests are conducted.
The question then is, what sort of test is it in the first place? For anybody who wants to read more posts on the topic, check out these resources from the AMSTO (Anti-Malware Testing Standards Organization). AMSTO has been discussing this question quite seriously over the past few years and has come around to publishing new testing criteria. Until today, the tests have essentially taken some substantial and impressive number of sample dangers, anywhere between five hundred thousand and one million, and isolating one layer of this the program under testing to determine how well it detects an incoming threat.
This sounds just like what you would expect, but there are a few issues that originally come to mind. 1) If a single layer of a software package is analyzed separately, how well does this really speak for how the program will actually run on a customer’s computer? (2) Can the way of throwing hundreds of thousands of threats at a system actually simulate the actual world of online security in the modern world? For the first query, we can only assume that this technique doesn’t provide a precise result for the way the program will operate when it’s functioning as it would on First of this may bypass certain security measures built into a different layer of the applications designed for specific threats, leading to the antivirus missing it because it wasn’t just because a program utilizes certain buffers doesn’t mean it needs to be reflected badly in the test.
Moreover, if the computer software is whittled down to a single layer, this will greatly impact speed. The picture that Software A is a part of a major package with plenty of features and Software B is a program rather stripped of additional features. If Software A, when focusing on just the antivirus, finds 98 percent of all viruses in quicker time than Software B that detects the identical number, Software A can As a user, I see this and I buy Software A, only to be amazed by the obvious lag in system tools.
It doesn’t appear to operate as quickly as the test stated it will. Of course, it’ll be slower when all elements of the software are operating concurrently. To answer the next question from above, in the real world, an individual won’t ever be faced with that variety of threats in a limited period of time. The evaluation procedures are an industry standard for the last twenty years. Obviously, a massive amount has changed since then.
Probably the most pertinent shift is that the mass use of social networking sites and downloading software from the Internet. Those two examples are equally real-time threats that are quite isolated; non-Armageddon such as scenarios-unlike how antivirus software is analyzed. Imagine if a message on Facebook popped up with a connection that contained some kind of malware? We do not need our applications to shield us from a million dangers; only that 1. Two percent of a million is 20,000, which makes 20,000 chances of a danger that could get through when I click on the link. If I click on it, the probability is strong that my antivirus will grab it. 1 chance from one is a bigger threat than 1 chance out of a million. The point I am trying to make is that we simply can not know. Because of this, the software should observe such messages which are a part of our online experience before we do, and advise us on what action is best.