[Mailman-Users] Re: [Mailman-Developers] Looping digest - mailman bug?

Barry A. Warsaw barry at zope.com
Wed Aug 1 23:11:40 CEST 2001

>>>>> "TW" == Thomas Wouters <thomas at xs4all.net> writes:

    TW> My question was whether the message text, at that moment, was
    TW> guaranteed to contain CR's (or guaranteed not to contain
    TW> them.) I'm not sure what kind of mangling Barry's mimelib does
    TW> :)

I don't think there are any hard guarantees, but certainly current
practice seems to be for Unix MTAs to deliver the message locally to
file or program with line endings canonicalized to LF only.

I actually think this is the right thing to do.  Let the protocol
specific module deal with massaging the data according to the spec,
but let client code deal with it in a natural, albeit platform
specific way.

mimelib doesn't do any explicit mangling of line endings.  Parser uses
fp.readline() and Generator uses the print command's implicit line
ending behavior which should both adhere to the local line ending


More information about the Mailman-Users mailing list