S
Shane
Hi, Chris,
Been wondering what happened to you.
Yes. Detecting various items as being CoolWebSearch. Most notably detecting
the Microsoft IE5 Web Accessories as CoolWebSearch components.
No. I don't run edgy apps.
Haven't seen anything like that in a long time (lol!, you could probably say
since I stopped using NAV2001!).
There's a big list somewhere - that I'm 99% certain Ron knows about. I don't
really see why he's so positive about it. Anyone serious about testing MSAV
should know about these false positives and I can only imagine the knowledge
is being repressed in a lot of cases.
http://groups.google.com/group/micr...lse+positive+silj+shane&#doc_519f496d39602241
http://tinyurl.com/7e4kg
Shane
--
The Sugitive
Chapter One: http://tinyurl.com/bcevp
Chapter Two: http://tinyurl.com/ag92o
Chapter Three: Coming to an URL near you soon!
Been wondering what happened to you.
False positives as in detecting something as being a part of a named
malware when it was not?
Yes. Detecting various items as being CoolWebSearch. Most notably detecting
the Microsoft IE5 Web Accessories as CoolWebSearch components.
Or in rating an edgy app as malware when you wouldn't judge it that?
No. I don't run edgy apps.
Or raising general heuristic alerts on things that were harmless?
Haven't seen anything like that in a long time (lol!, you could probably say
since I stopped using NAV2001!).
I'd consider only the first as a true false positive, e.g. detecting
signature material within an av product as if it were that malware.
If you know of any cases of this, please list them here, as these are
serious flaws we would wish to act on. In the harder test traditions
of the av industry, one false positive can disqualify a product.
There's a big list somewhere - that I'm 99% certain Ron knows about. I don't
really see why he's so positive about it. Anyone serious about testing MSAV
should know about these false positives and I can only imagine the knowledge
is being repressed in a lot of cases.
The second is always going to be controversial when you expand malware
beyond viruses and criminal trojans to generally unwanted commercial
software, as one must do in the 21st century. If anything, MSAS have
been accused of being too weak; they detect things OK, but the advice
on what to do about these detections may be too "soft" for comfort.
The last is simply the way risk management works. Malware scanners
use the blacklist approach when they tell you they blocked a named
malware recognised as such; they use the whitelist approach when they
block all risk behaviors until you indicate it's OK for that
particular program to initiate that particular risk.
So if by "false positive", you mean "I ran a batch file I wrote, and
MSAS popped up a dialog warning me a script was trying to run", then
welcome to the world of risk management - that's not a "false
positive" because it's not a malware detection as such.
http://groups.google.com/group/micr...lse+positive+silj+shane&#doc_519f496d39602241
http://tinyurl.com/7e4kg
Shane
--
The Sugitive
Chapter One: http://tinyurl.com/bcevp
Chapter Two: http://tinyurl.com/ag92o
Chapter Three: Coming to an URL near you soon!