Avast! 8 Beta?

Thanks for the information

 

Good to see you here man
Thanks for sharing
Yes have worked a lot on the backend.

 

i have this issue to, it was hard to get my internet working again

 

Just finished watching it ... WOW BTW we need more tests like that (with good pieces of malwares).
Anyway, it did a perfect job indeed.

 

Good, finally
I also like the extra information in the pop-ups i.e. blocked before execution or when created/modified, and the wording doesn't sound too technical for regular users.

 

Fantastic avast! performance. This video also gets few special awards.

- Award for first ever video documented Auto Sandbox detection
- Award for largest number of Win32:Evo-gen [Susp] detections in one video sequence
- Award for largest number of FileRepMalware deetctions in one video sequence

In addition it has prevented all the malware in the test. I wish Auto Sandbox detected stuff more often but they apparently did something really well with the Evo-gen classification system and the cloud that seem to be working really well. Their machines seem to be crunching the data and churning out detections really well. Looking forward to release of avast! 8 though Evo-gen itself is not bound to avast! 8 (meaning it also works already in avast! 7).

 

I was shocked to see as much good things from V8, I'll probably change my current config which is CIS 6 to avast! 8 soon ...
(Installed it on my own VM to start testing).

 

CIS 6 is not bad, it's certainly a great step forward compared to slow, overcomplicated and annoying CIS 5, but it's still too buggy and just doesn't work like i wanted it to work. avast! on the other end is combining traditional scanning with cloud and their internal sorting and signature generating system.
And it really shows. Most of these systems are already online, remaining one will go online this month or next.

My predictions are that avast! will be among top 3 in 2013 tests that involve version 8.

 

The thing I don't really like is the submission of files from CIS UI ...
It's almost useless. They never analyse them but many users submit their FPs from there and also something : Any unknown file is sandbox and sent to CAMAS, if CAMAS detects a suspicious behavior it automatically generates a signature (UnclassifiedMalware or Heur.Suspicious) the issue is that it also generates many many FPs ...

 

And that FP's never ever get fixed if you report them through GUI. I did so many months ago and all it happened they added more idiotic false positives in the meantime and haven't fixed a single old one. Unless you report it through their forum where they constantly harass you with the god damn reports formatting.

 

My freebie prediction for 2013, 1 Avast 8, 2 Fortinet 5, 3 Bitdefender Free (has less protection as pro version, so comparing apples with apples), 4 Panda cloud free, 5 AVG free, not but being a fan of Comodo, I think CIS 6 will score high also. All products will score high (all above 97% protection)

Avira remains great, but to much advertising to my taste (fighting for top three together with new Avast and new Fortinet).

 

Good list, but missing Roboscan. Have you tried it?

 

I only try out AV's when Wilders Members post a lot about them and I can get my hands on some fres malware samples of a honey pot, See for instance Panda https://www.wilderssecurity.com/showpost.php?p=2176109&postcount=98 I think that CIS 6 will out perform Panda free. Only the false positives of unsigned programs could be a problem for CIS scoring high in tests.

I have tested RoboScan once and found to my surprise the on-demand scanner to be better than the on-access scanner. Because most AV's emulate code execution the on-access is normally stronger than the on-demand scan, so I thought it was something strange in my small test (and did not post my experience).

I somehow find that free RE-BRANDED/WHITE LABEL versions of Bitdefender and Viper engines don't perform as well as the paid (original brand) engines. So they are not top of mind, reason for not mentioning them.

 

I have not tried it yet. It seems to not be ready for prime time.

 

another review: http://www.youtube.com/watch?v=vZHjLncg0hc

again avast in top class due to evo-gen and filerep...ignore nsm's conclusion he has a history on avast forums of being banned

I started testing today..i even saw MDE heuristics somewhere in the links.

Vlk said they had a bunch of cool stuff coming in 2013...this looks like the start of it...they are working on getting webrep in threat detection..stay tuned!!

Rejzor,I have been fiddling with avast v7 and v8 lately...it seems like v7 isnt that fully exposed to evo-gen but v8 is.

In addition, files were picked by autosandbox if avast isnt allowed to use evo-gen so it looks like evo-gen overrides sandbox to protect the user as a first line of defense and now the sandbox is the last line of defense against unknown threats...and evo-gen and filerep take the load of protection...and if they fail it is passed on to the sandbox for analysis

 

Hope they can make it compatible with Win8 soon.

 

as i said v8 was doing stellar and will be one to compete with this year...

 

Good to know.

 

if the advertising was taken out of the picture, how would Avira fare in your list?

 

One more test: http://www.youtube.com/watch?v=2tmQd6nq0lI

The items detected by HMP was on the desktop that were in the malwarepack...and the one in programdata was sandboxed and not allowed to infect the machine...Mind that the test was done 2 days later the public beta...and evo-gen and other stuff came into full power later this week.

 

Yes Evo-Gen has now really started to work I think first time I have as much alerts on my own tests ..

 

3:08, autosandboxed blocked a malware

 

yep,I guess the test by PCsecurity is much mature because it was done a few days later the beta was released when evo-gen actually began to be a strong defense

 

This test is a bit weird. HMP detected stuff on desktop but the folder wasn't there anymore (unless it was hidden behind apps on a desktop). Hm. 1 thing seemed to slip through though he didn't check with Killswitch if it was actualyl running after reboot so we'll never know for sure. But either way, 1 single sample out of 250 that slipped past is still pretty damn good imo. Especially considering this was a pure local only test where Network Shield cannot do its magic with Url:Mal detections...

 

As said in the top 3