Comodo Firewall Test Suite

Discussion in 'other firewalls' started by Coolio10, Nov 7, 2008.

Thread Status:
Not open for further replies.
  1. Alcyon

    Alcyon Registered Member

    Joined:
    Jan 16, 2008
    Posts:
    438
    Location:
    Montr?al, Canada
    Thanks for the screenshots, Rickster.

    The comparison is interesting but it ends here. It looks like that there's a problem concerning the ActiveDesktop test: even if I block it completely, it'll be shown as vulnerable. There was similar tests problems with earlier versions of CLTs! Are these Comodo tests reliable? You're the judge. I need to investigate but if there really is a coding glitch (which will not surprise me at all) with the first test, there's a high probability that many others will show you as vulnerable even if you're not.

    As an EQS user, I will not (and probably never will) rely on Comodo leaktests to make EQS rules.
     
    Last edited: Nov 10, 2008
  2. alex_s

    alex_s Registered Member

    Joined:
    Aug 13, 2007
    Posts:
    1,251
    Yep, he answered a question, but he didn't solve the problem. The problem is a user thinks he is safe, while actually he is not.
     
  3. alex_s

    alex_s Registered Member

    Joined:
    Aug 13, 2007
    Posts:
    1,251
    I can say for sure BIS test is also not reliable. For example, even if BITS job was submitted by COM interface, it doesn't mean the JOB was actually executed.

    For example, svchost can be restricted with addresses it is allowed to connect to. Then the submitted BITS job will just senslessly try to execute.

    Other option is to get and event on new BITS job and enumerate them all and suspend all of them and give a user a chance to select which jobs a good and which are bad. "Bad" jobs can then be cancelled. But the test alsways says "failed". It should surely check was job actually sucessful or not.

    Then registry autoruns. You can submit as many autoruns as you wish, but if HIPS prevents unauthorised autoruns fromstart then this is not a leak at all.

    And so on ... And I completely agree with Ilya, filedrop "test" is a bullshit.
     
  4. subset

    subset Registered Member

    Joined:
    Nov 17, 2007
    Posts:
    825
    Location:
    Austria
    Hi,

    what I really do not unterstand is why you can pass some tests with OS only and no security (Admin account; Windows Firewall, Windows Defender and UAC deactivated).

    XP SP3 20/340
    Protected:
    30. Invasion: Runner
    34. RootkitInstallation: MissingDriverLoad

    Vista SP1 100/340
    Protected:
    3. Hijacking: ChangeDebuggerPath
    6. Hijacking: UIHost
    8. Hijacking: WinlogonNotify
    13. Impersonation: OLE automation
    18. Injection: CreateRemoteThread
    21. Injection: ProcessInject
    26. Invasion: DebugControl
    28. Invasion: PhysicalMemory
    32. RootkitInstallation: DriverSupersede
    33. RootkitInstallation: LoadAndCallImage

    I have posted this at Comodo forums, no serious answer so far. Maybe someone can explain this here.

    Cheers
     
  5. IceCube1010

    IceCube1010 Registered Member

    Joined:
    Apr 26, 2008
    Posts:
    963
    Location:
    Earth
    Strange how some of us are getting differnet results with the same security setup, os, admin account etc. etc. I'm running vista home (sp1) admin account using CIS (without the AV) on Internet secuity (Safe mode on firewall and D+) and Avast Home (not that it matters in this test) and I get a perfect score. Strange.

    Ice
     
  6. Alcyon

    Alcyon Registered Member

    Joined:
    Jan 16, 2008
    Posts:
    438
    Location:
    Montr?al, Canada
    Not really strange I must say, considering that it's Comodo leaktests made BY Comodo. They probably haven't included many leaktests with "fail" results as a marketing strategy. That's how i'm seeing it.

    To see the true strength of hips and firewalls, what is needed is third-party leaktests... And i'm not talking about Matousec.
     
    Last edited: Nov 11, 2008
  7. 3xist

    3xist Guest

    No upsets.

    Sorry about before mate... No hard feelings! :) Thanks everyone for testing.
     
  8. alex_s

    alex_s Registered Member

    Joined:
    Aug 13, 2007
    Posts:
    1,251
    Nothing doing. Some tests were interesting.

    PS. Now you are much better, I'm really glad :)
     
  9. EASTER

    EASTER Registered Member

    Joined:
    Jul 28, 2007
    Posts:
    11,126
    Location:
    U.S.A. (South)
    It's totally out of character i know but i can easily readjust EQS 3.41 rules to successfully snuff off every single one of those CLT leaky tests, although the likelyhood of most of them are obviously biased in Comodos favor as being originair of it, so IMO it;s not to be taken as wide ranging gospel for sure.

    EQS 3.41 w/ Alcyon's Super Trapper (Adjustable) Rules = 100% Score in EQS favor. No Sweat, :thumb:
     
  10. mack_guy911

    mack_guy911 Registered Member

    Joined:
    Mar 21, 2007
    Posts:
    2,677
    KIS .454 score 240 out of 340 with comodo leaktest tested 3 times

    i tested it with outpost pro + KIS firewall uninstalled its shows 300 tested 3 times same result

    in administrative mode

    4. Hijacking: StartupPrograms
    19. Injection: DupHandles
    29. Invasion: RawDisk
    31. RootkitInstallation: ChangeDrvPath

    are Vulnerable

    please retest and confirm .......with your resultso_O?

    can any one confirm the results in usermodeo_O?

    core 2 duo 2.4
    2Gb ram
    windows vista home premium
    KIS+outpost pro
     
  11. pykko

    pykko Registered Member

    Joined:
    Apr 27, 2005
    Posts:
    2,236
    Location:
    Romania...and walking to heaven
    I managed to get 340 with Comodo.
    It seems there was something wrong with my initial installation: some missing protected registry keys.
     
  12. Einsturzende

    Einsturzende Registered Member

    Joined:
    Apr 14, 2008
    Posts:
    390
    Location:
    neubauten
    Best I can squeeze out of KIS (alone) is 290
     
  13. wat0114

    wat0114 Guest

    Newest Malware Defender beta release and Jetico 2 firewall w/Process Attack filter disabled:

    330/340 - only the dubious Active Desktop test failed :)

    Great work by Malware Defender's xiaolin to fix the compatibility bug with Jetico 2 so quickly :thumb:
     

    Attached Files:

  14. alex_s

    alex_s Registered Member

    Joined:
    Aug 13, 2007
    Posts:
    1,251
    There is only two options. Either the test is inconsistent (which was proven already) or product is inconsistent. Taking in account there are products that shows the same (or very similar) score despite of setup, you are free to make conclusion.
     
  15. doktornotor

    doktornotor Registered Member

    Joined:
    Jul 19, 2008
    Posts:
    2,047
    Impressive; I got 10/340... Not disappointed by the yellow company performance at all, just what I expected... :D :rolleyes:
     

    Attached Files:

  16. Leolas

    Leolas Registered Member

    Joined:
    Jun 18, 2008
    Posts:
    58
    Location:
    Modena, Italy
    Norton hasn't got an HIPS, so that's completely normal!
     
  17. Pseudo

    Pseudo Registered Member

    Joined:
    May 4, 2008
    Posts:
    193
    :cautious:
     
  18. IceCube1010

    IceCube1010 Registered Member

    Joined:
    Apr 26, 2008
    Posts:
    963
    Location:
    Earth
    I think you hit the nail on the head. What is really needed is a un-biased 3rd party test.

    Ice
     
  19. EASTER

    EASTER Registered Member

    Joined:
    Jul 28, 2007
    Posts:
    11,126
    Location:
    U.S.A. (South)
    Now that is true progress!

    Congrats!

    EASTER
     
  20. wat0114

    wat0114 Guest

    I'm encouraged by what I'm seeing with this HIPS and especially the tireless efforts of the developer. He's obviously highly skilled in this area of coding, so I hope he does not bail on the product that has such tremendous potential as others have done. I have no problems supporting this kind of effort going into a very promising security application, even if I don't actually need a license.
     
  21. 3xist

    3xist Guest

    If people are having really weird scores with various products, excluding CIS - Try and set them all up to high on every angle of the Firewall & HIPS/Behavior Blocking your using.

    This Testing Suite was designed by the AV Labs, And these tests use REAL malware techniques, including those nasty rookits!! It was designed to test the full strength of a product. That's why if you have "COMODO - Proactive Security" Configuration in CIS, And you block all Defense+ Alerts you will get 340/340, or you should anyway. Proactive Security Configuration activates all of D+ Full power, All D+ Monitor Settings & Image Execution is Enabled, Where in the default installation configuration it was just Internet Security Enabled, And eased off in D+ Places due to Antivirus, etc.

    So I advise if you could look for some "high" settings in Kaspersky, and any other product your using - and blocking ALL Alerts they show, to fully pass this test. Just a though on the top of my head. :) Goodluck!
     
  22. wat0114

    wat0114 Guest

    Thank you 3xist. Is CIS achieving this perfect result under an administrator account or limited account? I'm still of the opinion these tests should be run under a limited account, even if it's Power user.
     
  23. 3xist

    3xist Guest

    Admin Account...

    I haven't tested Limited.
     
  24. hammerman

    hammerman Registered Member

    Joined:
    Jul 14, 2007
    Posts:
    283
    Location:
    UK
    Hi Easter

    Are you saying that with Alcyon's latest rules, a 100% score is achieved against these leak tests? What do you mean by Super Trapper (Adjustable) Rules?

    Thanks.
     
  25. Yoda1953

    Yoda1953 Registered Member

    Joined:
    Nov 1, 2007
    Posts:
    163
    Location:
    Netherlands
    I tried Spyware Terminator beta on this. It left an empty frame. :blink: So score was 0/0.
     
Thread Status:
Not open for further replies.
  1. This site uses cookies to help personalise content, tailor your experience and to keep you logged in if you register.
    By continuing to use this site, you are consenting to our use of cookies.