The performance of Twistee in the VB100% Test(April 2009)

[ftfans]

Click here to see the original page(need registering).

Failure reason: 53 wildlist misses, 21 false positives

There is a long way to go!

However, I am very happy to see that Filseclab began to be in this test!

The performance of Twistee in the VB100% Test

vb.jpg (122.64 KiB) Viewed 13662 times

[Tony]

At least it is a step in the right direction being in these tests.
Surely Twister can only get better now.

[3DFireStarteR]

Its a good result even if it failed, I would bet my money that the failure was because they use all US/UK type malware where as Twister is more likely to have China/Japan/Thai - Malware files, And the FP result was quite good I don't know if they tested FDDS but if they did not It would be a reason why there was "Wild" Misses, And could also be why there are so few FP's

[Fuzzfas]

Its a good result even if it failed, I would bet my money that the failure was because they use all US/UK type malware where as Twister is more likely to have China/Japan/Thai - Malware files, And the FP result was quite good I don't know if they tested FDDS but if they did not It would be a reason why there was "Wild" Misses, And could also be why there are so few FP's

I agree. Twister needs to be more "in contact" with western malware. VB has been "passed" by far worse avs in the past.

[renegade]

Thread was updated with review.

VB100 April 2009 - Windows XP SP3

2009-04-01

Filseclab Twister AntiVirus 7.3.2.9971

The first of the newcomers in this month’s test, Filseclab’s Twister has picked up a bit of a reputation as a strong up-and-comer on various web forums and discussion boards, and has put in some excellent performances in independent tests run in China. An initial trial version we looked at impressed us with simplicity, stability and better than expected scanning performance, and a later version submitted for the test showed even more promise. With a slick and professional-looking installation process and a clear, attractive and well laid-out interface, the product certainly looks the business and has a very good level of fine-tuning available, as well as a behavioural monitoring system that is given as much importance as the more traditional detection in the layout of the interface.

Running through the tests proved a little less straightforward than hoped thanks to some slightly unusual behaviour: on-access scanning, while triggered on read, seemed not to block access instantly, instead waiting a little before alerting on and taking action against detected items. This meant that our standard opener tool, which logs items it cannot access, recorded having successfully opened everything. Thus, detection data could only be gathered from the product’s own logs and the on-access scanning speeds, recorded in the same manner, may not quite reflect the full picture.

Detection rates were not unreasonable, particularly for a product that is entirely new to our testing system and test sets. Fairly good scores were achieved in some of the standard sets, including a surprisingly excellent handling of W32/Virut samples in the polymorphic set, with a little less coverage of older polymorphic items, and a fairly decent showing in the trojan and RAP sets. Several items in the WildList set were not covered, most of which were from the latest batch of additions, and a sprinkling of false alarms were raised in the clean sets (no big surprise on the product’s first look at their diverse content), so Twister does not qualify for a VB100 award on its first attempt, but it looks like being a strong contender in the very near future.

ItW: 86.85%
ItW (o/a): 86.85%
Trojans: 66.77%
Worms & bots: 83.44%
Polymorphic: 30.25%
False positives: 21