You don’t have to pay for it. Paying for it gives you removal, but that can be done manually. As for HiJackThis, in addition to the logs it can generate, it also provides tools for manually removing malware.
“Security is debatable. Therefore, speed would be my basis. OpenDNS and Google DNS (although I doubt people trust Google) is much faster.”
I have to disagree. OpenDNS is the best for antiphising protection while NortonDNS is the best for malware protection see tests on it. Yes Google is a dodgy one.
Well, I was explaining why I discourage multiple engine antiviruses, but anyway, it still answers the question concerning tests…
https://forums.comodo.com/empty-t80797.0.html
Like naren and I have said, any number of engines will not secure the 2% difference. Allow me to elaborate on naren’s explication.
Technically, the 2% is a huge set of malware since each has a family of more than a dozen variants . Common sense would tell us that the use of two engines should increase your chances of getting an update for the 2%. But this is improbable both mathematically and realistically speaking. Detection of the 2% is actually based on the frequency of the release of signatures rather than the quality of the engine itself (see paper by GFI provided by Seany007). And these tests fail to consider that and the following:
Probability relies on finite measurable space . Therefore, real-time cannot be measured being infinite. Probability can tell you that the chances of picking a whole number between -7 to 7 is 8/14. But it cannot give any statistic for picking a real number except 1/∞ between the same range. If interpreted, the probability is indeterminate and imagined .
The tests were conducted at a given period of time. But the results were posted at the end of the tests. Given that the theory relies on the frequency of updates, the results are rendered invalid upon release unless none of the products tested released a signature update during the tests (unlikely since antiviruses release hourly updates).
If applying “Every little bit helps” principle, we must take a look at the complement of the event. Figuratively put in a weighing scale, the disadvantageous events outweigh the advantages (see above link).
On a side note, so what are these tests useful for?
Market share. These tests can provide data to establish a trend plotted on a diagram. From that scale, the user may extrapolate and choose a “reliable” company. (Of course, this, too, is imagined.)
Gauging removal efficiency. Depending on the context, this may involve the system parts or the malware traces themselves. In either case, the set is finite and measurable; therefore, efficiency can be estimated.