ThunderBird 1.5 and Scams

Discussion in 'Firefox' started by RodneyK, Mar 1, 2006.

  1. RodneyK

    RodneyK Guest

    I'm getting a lot of scam emails lately.

    If ThunderBird marks an email as a scam I can click a button to say
    "not a scam" if I need to.
    But there seems to be no way that I can mark a scam email that has been
    missed; and that is most of the scam emails.

    How does ThunderBird identify and treat scams?

    RodneyK
     
    RodneyK, Mar 1, 2006
    #1
    1. Advertisements

  2. RodneyK

    Olgierd Guest

    Świat wstrzymał oddech bo osoba znana jako *RodneyK* wypowiedziała te
    słowa:
    Right-click on the email and "mark as spam" or whatever it's in English
    version ;-)
     
    Olgierd, Mar 1, 2006
    #2
    1. Advertisements

  3. He is talking about the 'scam' feature in TB, not the Junk Mail Controls

    The 'scam' feature is new in Thunderbird, and currently is not
    trainable. You can mark a message that Thunderbird thinks is a scam as
    "Not a Scam" - and thats it. There is no means of marking a message as a
    'Scam' by the user.

    The 'Scam' and 'Not a Scam' markings arnd settings are actually
    irrelevant, at the moment anyway. If TB thinks its a scam, then the
    message is marked as one, and you can un-mark it if you wish. There is
    no training, no learning, the next time you get the message, it will
    again be marked as a scam by Thunderbird

    The only control you actually have over the 'feature' is to turn it off.
    It doesnt 'learn', it doesnt 'adjust' for 'Not a scam' selections, nor
    can you teach it what is a scam and what isnt.
     
    Moz Champion (Dan), Mar 2, 2006
    #3
  4. RodneyK

    gwtc Guest

    and what is the file name that this Scam uses.
     
    gwtc, Mar 2, 2006
    #4
  5. The detector is triggered if the URL has a numeric Internet Protocol
    address rather than a domain name, or if the URL does not match the
    address displayed in the link text.

    As implemented, if the message has been marked Junk, then it wont be
    checked for phishing
    if the user selects the "Not a Scam" then the message loses its 'Scam'
    marquee and is reloaded (for viewing)
    NB, it only applies to the specific message in view

    This is what happens.
    When you open a message, the software checks to see if the message is
    junk, if it is, then it wont be processed for scam features.
    If the message has not been marked as Junk, then if the message contains
    urls, they are checked. If the url in the message is a numeric one, OR
    if the url in the link text doesnt match the url in the link itself,
    then the message is given the 'Thunderbird thinks this is an email scam'
    label, complete with a "Not a Scam" button.

    Pressing the "Not a Scam" button clears the label for the current message.


    Once more, the 'scam' feature is not trainable, not configurable, doesnt
    'learn' and never changes, at least in the current implementation.
     
    Moz Champion (Dan), Mar 2, 2006
    #5
  6. RodneyK

    gwtc Guest

    from what I'm understanding is that its a Label that is placed on the
    messages. Right? Therefore, there is no file like the Junk Controls
    has with training.dat.
     
    gwtc, Mar 2, 2006
    #6
  7. Correct. There is no training.dat, no file, nothing

    If the message is not junk, has urls within it, but the url is a numeric
    one OR if the url is different from that in the link text, then the
    label is generated. Pressing the "Not a Scam" button clears the label
    for THAT specific message

    There is no means of users marking a message as a scam. Use of the "Not
    a Scam" button simply removes the label from the specific message

    No training, not configurable, no learning, no changes at all are
    possible with the current implementation of the feature. Users can turn
    the feature off, or leave it on. If on, they can remove the label from a
    message they determine not to be a scam. Thats it.
     
    Moz Champion (Dan), Mar 2, 2006
    #7
  8. RodneyK

    gwtc Guest

    Then, if its a "label", then wouldn't it part of the *.msf file.
     
    gwtc, Mar 2, 2006
    #8
  9. RodneyK

    RodneyK Guest

    Yes, thank you for the info. Phishing is worry.
    I'll keep that in mind.,-)

    RodneyK
    Enjoy
     
    RodneyK, Mar 2, 2006
    #9
  10. Yes and no.

    When a message is opened the phishing scan is done.
    As before, if the message has been marked as Junk, then the phishing
    scan does not take place
    Additionally, if the message was marked "Not a Scam" then the scan
    doesnt take place either

    'Scam' labels are applied to the message when opened, there is no
    recording of such, no actual changes to the message, except if the
    message is marked "Not a Scam" in which case it gets a notaPhishMessage
    header.

    So the value to 'mark' the message as scam is derived each and everytime
    the message is opened, unless it has been marked 'Not a Scam' in which
    case the scan is suppressed (not done).

    So, Scam Messages themselves are not changed to any degree, but messages
    marked 'Not a Scam' will have the custom not a phish header - which of
    course will be in the .msf file

    Ergo, the only property in the .msf regarding scams would be if you had
    explicitly marked that message as not a scam
     
    Moz Champion (Dan), Mar 3, 2006
    #10
  11. RodneyK

    gwtc Guest

    gotcha, thanks Dan
     
    gwtc, Mar 3, 2006
    #11
    1. Advertisements

Ask a Question

Want to reply to this thread or ask your own question?

You'll need to choose a username for the site, which only take a couple of moments (here). After that, you can post your question and our members will help you out.