FAQ
Greetings. On a 2.1.6 era mailman, I've discovered that any mime input
messages are having their mime boundaries and related mime information
stripped out from the message body. All that remains is the
Content-type header line. As one would expect, this causes all sorts of
problems for mime-aware UAs, since the various parts of the input
messages are all cat'd together without any boundaries.

Filters are OFF.
Scrubbing is OFF.

Any ideas? Thanks.

--Lauren--

Search Discussions

  • Mark Sapiro at Nov 10, 2007 at 2:26 am

    Lauren Weinstein wrote:
    Greetings. On a 2.1.6 era mailman, I've discovered that any mime input
    messages are having their mime boundaries and related mime information
    stripped out from the message body. All that remains is the
    Content-type header line. As one would expect, this causes all sorts of
    problems for mime-aware UAs, since the various parts of the input
    messages are all cat'd together without any boundaries.

    Filters are OFF.
    Scrubbing is OFF.

    Please give us more information. If filter_content is No and
    scrub_nondigest is No, messages sent to the list should have all the
    MIME headers, boundaries, and structure intact except for possibly
    wrapping the entire message structure in a higher level
    multipart/mixed message in order to add msg_header and/or msg_footer
    parts.

    Is this Mailman 2.1.6 installed from source or a package, and if a
    package, which?

    Under what circumstances are you seeing this?

    Do you have an example of a raw message as sent to the list and the
    corresponding raw message as received from the list or found in the
    archives/private/listname.mbox/listname.mbox file.

    --
    Mark Sapiro <msapiro at value.net> The highway is for gamblers,
    San Francisco Bay Area, California better use your sense - B. Dylan
  • Lauren Weinstein at Nov 10, 2007 at 4:25 am
    Mark, et al.,

    Hi. This is a source install that I've been using for a long time
    mostly for locally generated messages and digests, but am now using
    for large numbers of externally generated (moderated) message
    distributions. I should add that I also have removed (including all
    white space -- and I set the configuration default for both to "")
    the headers and footers for direct distribution (I've left in the
    header for digests, on the assumption that won't be an issue). I
    had *thought* this had solved the main problem, and I have a test
    message that apparently came through correctly, but later tests are
    failing again. I feel like the guy in the old movie "The 4D Man" who
    kept trying to replicate the one time his tricky experiment worked.

    What's happening is that messages coming through (and they look this
    way in the moderator's preview as well) are being stripped of all
    MIME boundaries -- I'll include an example below. The boundaries
    are clearly present (and the messages display correctly) on cc'd
    versions that don't go through mailman. During testing, I also had
    an incident where it appeard that a newly arrived test message was
    deleted unseen when a bunch of older messages were being deleted thru
    the moderator interface, but I'm not ready to swear to that yet.
    Different issue.

    Example of the Mime problem follows. Note that the crypto portion
    immediately follows the main text with no boundary, nor is there
    a boundary at the end of the crypto section. Any ideas appreciated!
    From the moderator interface:
    HEADER SECTION:



    Received: from chrome.vortex.com (chrome.vortex.com [67.119.61.37])
    by neon.vortex.com (8.12.6/8.12.6) with ESMTP id lAA4JcRI001368
    for <test@[67.119.61.35]>; Fri, 9 Nov 2007 20:19:38 -0800 (PST)
    Received: from [67.119.61.44] (nebula.vortex.com [67.119.61.44])
    by chrome.vortex.com (8.12.6/8.12.6) with ESMTP id lAA4JcGm021330
    for <test>; Fri, 9 Nov 2007 20:19:38 -0800 (PST)
    Message-ID: <4735315A.1020605 at vortex.com>
    Date: Fri, 09 Nov 2007 20:19:38 -0800
    From: Lauren Weinstein <lauren at vortex.com>
    User-Agent: Thunderbird 2.0.0.6 (Windows/20070728)
    MIME-Version: 1.0
    To: test
    Subject: Test - signed MIME message
    Content-Type: multipart/signed; protocol="application/x-pkcs7-signature";
    micalg=sha1; boundary="------------ms050600000604090002090408"




    BODY SECTION:




    TEST
    TEST
    TEST
    Signed section should follow
    MIAGCSqGSIb3DQEHAqCAMIACAQExCzAJBgUrDgMCGgUAMIAGCSqGSIb3DQEHAQAAoIIRYTCC
    BN0wggPFoAMCAQICEHGS++YZX6xNEoV0cTSiGKcwDQYJKoZIhvcNAQEFBQAwezELMAkGA1UE
    ...
    mFLEUZnF8jaU9MvAPdNmRsulow+SySznPvYnkLUv2WdY3yVhTXN+ljxFLSebHegMAIRAbnX1
    PS2Ed260hBueoXTUuY4woaaea5SmDHOInb8LoPPfYEiXlQO4jYDQL/Br/oWOEsB4ZHx0bVyG
    IGZiEmK+3pPqetzbReRqRR19VyBYlA4mu/5lVi2kZagSjYLo4+V81rzTFsUShe0pca/kaHkg
    Flw3W7KrMxXMviMb0JXsTC907o629w+2toyuaQRWbeZ0dkGiZ5I8Y/n2mSlN48IAAAAAAAA=





    Lauren Weinstein wrote:
    Greetings. On a 2.1.6 era mailman, I've discovered that any mime input
    messages are having their mime boundaries and related mime information
    stripped out from the message body. All that remains is the
    Content-type header line. As one would expect, this causes all sorts of
    problems for mime-aware UAs, since the various parts of the input
    messages are all cat'd together without any boundaries.

    Filters are OFF.
    Scrubbing is OFF.

    Please give us more information. If filter_content is No and
    scrub_nondigest is No, messages sent to the list should have all the
    MIME headers, boundaries, and structure intact except for possibly
    wrapping the entire message structure in a higher level
    multipart/mixed message in order to add msg_header and/or msg_footer
    parts.

    Is this Mailman 2.1.6 installed from source or a package, and if a
    package, which?

    Under what circumstances are you seeing this?

    Do you have an example of a raw message as sent to the list and the
    corresponding raw message as received from the list or found in the
    archives/private/listname.mbox/listname.mbox file.

    --
    Mark Sapiro <msapiro at value.net> The highway is for gamblers,
    San Francisco Bay Area, California better use your sense - B. Dylan
  • Mark Sapiro at Nov 10, 2007 at 4:51 am

    Lauren Weinstein wrote:
    What's happening is that messages coming through (and they look this
    way in the moderator's preview as well) are being stripped of all
    MIME boundaries -- I'll include an example below.

    The example below from the moderator interface is irrelevant. In the
    moderator (admindb) interface detailed view, "Message Headers:" shows
    only the main message headers, not any MIME part headers and "Message
    Excerpt:" shows only body lines from the MIME parts.

    In short, you never see boundaries or part headers in the admindb
    interface, but they should still be in the message sent to the list
    members (unless removed by subsequent content filtering) if the held
    message is approved.

    The boundaries
    are clearly present (and the messages display correctly) on cc'd
    versions that don't go through mailman.

    And what about a message that is approved in the moderator interface
    and sent to the list? Note that if any messages were sent to the list
    with missing boundaries/part headers, and you have access, you should
    be able to find them in the
    archives/private/listname.mbox/listname.mbox file, and even if you
    don't have direct access to that file, you should be able to get it
    through the web via a URL like
    <http://example.com/mailman/private/listname.mbox/listname.mbox>.

    During testing, I also had
    an incident where it appeard that a newly arrived test message was
    deleted unseen when a bunch of older messages were being deleted thru
    the moderator interface, but I'm not ready to swear to that yet.
    Different issue.

    This should not happen. Even if you "Discard all messages marked
    Defer", the only messages discarded should be those on the page.

    Example of the Mime problem follows. Note that the crypto portion
    immediately follows the main text with no boundary, nor is there
    a boundary at the end of the crypto section. Any ideas appreciated!

    As I said above, this is expected in the admindb view and says nothing
    about the message that will ultimately be delivered to the list.

    --
    Mark Sapiro <msapiro at value.net> The highway is for gamblers,
    San Francisco Bay Area, California better use your sense - B. Dylan

Related Discussions

Discussion Navigation
viewthread | post
Discussion Overview
groupmailman-users @
categoriespython
postedNov 10, '07 at 2:08a
activeNov 10, '07 at 4:51a
posts4
users2
websitelist.org

2 users in discussion

Mark Sapiro: 2 posts Lauren Weinstein: 2 posts

People

Translate

site design / logo © 2023 Grokbase