I've used that site for a while now. If you give him your email for his newsletter he will give you access to more info.
Oh and just because he says something doesn't mean its true.
I've been reading 1 topic at wilders lately and its this one:
http://www.wilderssecurity.com/showthread.php?t=164982Just to see if there is anyone there with enough commen sense to look at all of this stuff from the perspective of the average user and I found exactly what I was thinking in this quote:
QUOTE
Franklin,
Any user, casual or advanced, may still be faced with the question of whether downloaded content is potentially malicious or not. Given that malicious and non-malicious programs can exhibit similar behaviors, the control of behavior is not necessarily a panacea. Ultimately a user needs to be able to assess whether file X is malicious or not and there's really only three paths to that answer:
* Run it and observe over an extended timeframe. This is ambiguous if the malware is a sleeper.
* Learn enough coding to reverse engineer the executable. Not practical.
* Take advantage of the expertise of professionals who can reverse engineer the executable to make an independent assessment. In other words, use a classical blacklist AV as a screening tool.
The last option is really the only one feasible in a large and heterogeneous user base and as long as content can be downloaded and used, my personal opinion is that this will always be true. Now, that classical blacklist may need to be augmented to reflect current challenges floating around, and that short list of viable additions you mention are certainly potential avenues for anyone to explore.
Proclamations such as that leading off this thread or the Robin Bloor piece, that the classical AV is effectively dead, are ludicrous and something that even extensive user education will not change.
Finally, there are no silver bullets now, and there never will be as long as the motivation to perform malicious deeds is present.
Blue
Basically he is saying how will you know what is bad and what isn't?
Its not practical to run in a "sandbox" all of the time so you still need to know whats valid and what isn't. If virtualization fits in with your needs then by all means go for it but it isn't a one size fits all glove.