Discussion in 'Prevx Releases' started by SIR****TMG, Mar 18, 2010.
http://malwareresearchgroup.com/forum/viewtopic.php?f=18&t=339 see this test just ran
very impressive, especially in the area of what these tests were about. Kudos Prevx.
Yes indeed and in order to help those still stuck in the aftermath of the world wide economy meltdown (honest ), how about celebrating with a "Live Appearance" on G.A.O.T.D. (pretty please )
Congrats to Prevx you should be very proud
The malware research group test seems to be at odds with one done on here a few months back ?
Trusteer Rapport Free did much better in that one against the Zemana test files ? and others.
Good to see MRG doing the tests but wonder why MRG didn't also run these tests as well ?
Seven KL + two Screen Capture tests.
Thanks for the compliments everyone I've forwarded your comments onto the rest of the Prevx team, and the Prevx team would like to extend their gratitude back to our users (especially the ones here! ) for their dedication and testing - SafeOnline is not a simple product and we have found the journey to be much easier with the help of the Wilders team
And we also extend a special thanks to MRG for helping shed some light on the state of browser protection software with these thorough tests It takes quite a lot of time, dedication, and technical proficiency to run through this many tests systematically on that many products. 13 leaktests, 10 products... 130 different tests to run there
Thank you all!
We could... or if you use Facebook, you can get SafeOnline for free http://www.facebook.com/#!/pages/Prevx-SafeOnline/254680228961
I would prefer G.A.O.T.D. as i avoid facebook like the plague Huge privacy risk!
DW scores zero. Sure they did not know how to test different types of security applications.
I am surprised on the skill level of these so called testers. Their skill is less than many ordinary users even.
Strange test... Zemana Antilogger is tested with 4 of the 6 testing tools being from Zemana? Not much drama in that outcome, is there?
I'm glad Prevx prevailed without being tested by any of its own test tools.
And, I would have liked to see how Prevx did with detection and blocking (both ignored) in order to test penetration with SafeOnline.
I am quite surprised at DefenseWall's performance. And Mamutu's. It doesn't seem right, in fact.
I can't speak for DW but I know that Emsi have stated in the past that their products aren't set up to pass these kind of tests,they concentrate on the real threats.
You know, as many times as I've heard that said, to be honest, I just can't understand it, or maybe I don't believe it. Here's what I mean... if an application (I'll use Mamutu as an example) is smart enough to know that something is a test and not a real threat, then isn't that in fact detection? Plus, if Mamutu is letting something get by (because it knows it is a test and not a real threat), then what's to stop a real threat from disguising itself as a test? I am not disagreeing with you, andyman35... I just can't quite buy it when these developers say they focus on real-world stuff and don't bother with stopping test tools. I think they are making excuses, or at the very least, if they are so good that they know the difference between a test and a threat, then why not go ahead and stop the test tool?
Well done - kudos Prevx!
I understand your point and I'm not trying to speak as a Mamutu fanboy here.It's just that from my understanding of it's methodology individual events won't necessarily trigger a warning,an overall malicious behaviour threshold must be exceeded for that to happen.Since these tests don't actually do anything damaging to the system I can see where they're coming from on this.When I've tested Mamutu against real KLs it has worked as it should but who knows? Only Emsi could answer this definitively.
Good result. Well done Prevx .
Congrats to the Prevx Team and that means good for us as Clients and Security Guru's!
It also takes money. What is the business model of the Malware Research Group (MRG)? How can they afford the expense of such tests without a corresponding and offsetting revenue stream?
P.S. Congratulations to Prevx!
This is just one more feather in the cap of Prevx as far as testing and their plume is growing.
Dont look at the still picture folks, but the complete movie.
Just like to add my congrats to Prevx And we still have Prevx 4 to come
I would suspect that the tests were run as trusted, they need to be untrusted. They needed to be downloaded through DW untrusted browser.
The same applies to GesWall.
I fully agree also! But I knew that Prevx would pass any test to do with SafeOline! Joe and the Prevx Team do there own Testing https://www.wilderssecurity.com/showpost.php?p=1636649&postcount=1
indeed what a waste of time - and yes I use Prevx and DW
Good job Prevx.
If I recall correctly, DW will not block keylogging, screencaptures etc, it will merely report the event is happening via a pop-up with the option to terminate. It does not stop the malicious activity even while the pop-up is there. So on that basis, it is a fail on all tests...which I agree with in this case.
...and well done Prevx!
With reference to comments from agile, PC Gamer, Starfish and Tripplehelix, if you read the methodology, you will notice we did not specify an infection vector. This is for an important reason.
Financial malware infect systems using any number of vectors. Security applications need to prevent this type of malware from stealing data from online baking sessions, regardless of how the infection entered the system.
Prevx with its SafeOnline technology achieves this since it secures the browser itself. Defensewall only protects infections from a limited range of vectors. If the infection had entered via a p2p client or numerous other vectors Defensewall would be completely ineffective and would fail to block any of the attacks.
We are currently designing a major, groundbreaking test, which we feel will accurately model how financial malware compromise systems and steal data. We will be using custom, zero day malware which should go undetected by AV engines.
The malware used will employ the same methods that Zeus etc use when capturing and sneaking data out of the infected system, so this test will also determine which security applications are able to prevent actual data capture and also prevent data from leaving the system.
To make the test reflect the real world as accurately as possible, the test will be repeated using several common infection vectors and will therefore be able to determine the effectiveness of each security application in protecting a system from attacks via these.
This test should provide useful feedback for security vendors and we hope, enable them to revise their applications and so make a safer online environment – which is what I believe we all want.
Separate names with a comma.