Larry is right to some degree. You can get away with being more sloppy on the gateway than on the client system. So a user doesn't get an email because of an FP -- that's not good, but it's not the same as blowing your entire desktop. Furthermore, you can always pull the email out of quarantine. Same goes for web surfing. As an example, WebWasher has a reputation in industry circles for high FPs, but consistently scores as a top gateway scanner (it is only a gateway scanner, not a client product):
Example: http://winnow.oitc.com/AntiVirusPerformance.html That being said, Dr. Solly is right. The sheer volume puts an enormous strain on the system. A month of seasoning would be wonderful to reduce FPs. But that's occuring anyway; the beta testers are the users. Just download some of the more popular antispyware apps these days and see what "trojans" are found on your system (which sure helps them with their sales... but that's another discussion). Alex -----Original Message----- From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] On Behalf Of Larry Seltzer Sent: Friday, December 21, 2007 6:36 PM To: Drsolly Cc: [email protected] Subject: RE: [funsec] Kaspersky strikes again Even so, there would be so much less testing to do, wouldn't there? After all, on an appliance users can't just arbitrarily install applications (not and expect support). Larry Seltzer eWEEK.com Security Center Editor http://security.eweek.com/ http://blogs.pcmag.com/securitywatch/ Contributing Editor, PC Magazine [EMAIL PROTECTED] -----Original Message----- From: Drsolly [mailto:[EMAIL PROTECTED] Sent: Friday, December 21, 2007 6:29 PM To: Larry Seltzer Cc: [email protected]; Richard M. Smith Subject: RE: [funsec] Kaspersky strikes again On Fri, 21 Dec 2007, Larry Seltzer wrote: > Damn, I'm going to get a good column out of this. > > Doc: What about gateway appliances? Is a signature system more > reasonable when you have a limited number of closed platforms? You've misunderstood my concern. If you update your sigs hourly, then you have less than an hour to do all the testing. It doesn't matter how many computers are running the new version; they're all running something that has had less than an hour of testing, and I don't really want to run something that has been tested for less than an hour, on my systems. A month would probably be enough. A day would probably not be enough. Flagging "Explorer.exe" puts me in mind of when Fredrik issued a sig that false-alarmed on Command.com in the Virus Bulletin publication. We called that "The mother of all false alarms". > Larry Seltzer > eWEEK.com Security Center Editor > http://security.eweek.com/ > http://blogs.pcmag.com/securitywatch/ > Contributing Editor, PC Magazine > [EMAIL PROTECTED] > > > -----Original Message----- > From: Drsolly [mailto:[EMAIL PROTECTED] > Sent: Friday, December 21, 2007 5:52 PM > To: Larry Seltzer > Cc: Richard M. Smith; [email protected] > Subject: RE: [funsec] Kaspersky strikes again > > That's one of the big reasons why it isn't possible to write a > signature-based antivirus these days. You're caught in the nutcracker > of > 1) need to update frequently and 2) need to test adequately. > > I don't see how it's possible to do daily updates, let along hourly. > Even weekly updates sounds too difficult. > > On Fri, 21 Dec 2007, Larry Seltzer wrote: > > > I remember years ago writing about the speed of updates necessary > > now for a/v vendors, and how kaspersky talked about how they do it hourly. > > It basically makes it impossible to do meaningful tests. > > Larry Seltzer > > eWEEK.com Security Center Editor > > http://security.eweek.com/ <http://security.eweek.com/> > > <http://blogs.pcmag.com/securitywatch/> > > http://blogs.pcmag.com/securitywatch/ > > <http://blogs.pcmag.com/securitywatch/Contributing> > > Contributing Editor, PC Magazine > > [EMAIL PROTECTED] > > > > > > > > ________________________________ > > > > From: [EMAIL PROTECTED] > > [mailto:[EMAIL PROTECTED] > > On Behalf Of Richard M. Smith > > Sent: Friday, December 21, 2007 9:11 AM > > To: [email protected] > > Subject: [funsec] Kaspersky strikes again > > > > > > Kaspersky false alarm quarantines Windows Explorer Accidents will > > happen > > > > By John Leyden > > <blocked::http://forms.theregister.co.uk/mail_author/?story_url=/200 > > 7/ > > 12 > > /20/kaspersky_false_alarm/> > > 20 Dec 2007 17:00 > > http://www.channelregister.co.uk/2007/12/20/kaspersky_false_alarm/ > > <blocked::http://www.channelregister.co.uk/2007/12/20/kaspersky_fals > > e_ > > al > > arm/> > > > > A faulty signature update from Kaspersky Lab on Wednesday flagged up > > Windows Explorer (explorer.exe) as infected with a low-risk virus, > > Huhk-C. As a result the core Windows component was quarantined or > worse. > > > > Kaspersky released a revised update alongside advice on how to > > recover > > > legitimate system and application files from quarantine (the default > > setting) within two hours. But that's not much consolation for users > > that had set their software to auto-delete infected files, who found > > themselves with hosed systems. > > > > Among those affected was Reg reader Carl. "A false positive caused > > the > > > deletion of explorer.exe.," he reports. "It would have only caused > > problems for companies performing their network scan during the > > hours that the dodgy update was present - which included me, unfortunately. > > I was working out of hours to fix the previous Kaspersky update > > problem. I finally finished sorting it all at 5am.". > > > > ... > > > > > > _______________________________________________ Fun and Misc security discussion for OT posts. https://linuxbox.org/cgi-bin/mailman/listinfo/funsec Note: funsec is a public and open mailing list. _______________________________________________ Fun and Misc security discussion for OT posts. https://linuxbox.org/cgi-bin/mailman/listinfo/funsec Note: funsec is a public and open mailing list.
