Attached is a pickle of a POP3 message (as returned by POP3_SSL.retr)
that generated long lines in my tracker. Load it with:

    import pickle
    with open('msg.pickle', 'rb') as f:
        msg = pickle.load(f)

I've hand-manipulated the lines of the message and replaced names of
individuals and organizations with different names of the same length
(for privacy).

Here is an excerpt of one of the long lines from msg[1]:

 b'We should have some more answers from this one vendor this week, and
if not=',
 b'hing is satisfactorily received for this one, we will keep continue
to move=',
 b' until we have a solution.',

As you can see, the e-mail payload contains continuation blocks
(trailing equals signs). Presumably, these are used construct new lines
that are far more than 80 characters. Because these are not wrapped
before they're displayed in roundup, they cause the display to be very
difficult to read.