[R] Hypatia?? [Was: Adding Two Vectors (was: no subject)]

[Martin Maechler]

>>>>> "Marc" == Marc Schwartz <MSchwartz at mn.rr.com>
>>>>>     on Sat, 05 Nov 2005 11:40:58 -0600 writes:

    Marc> On Sat, 2005-11-05 at 17:52 +0100, Martin Maechler wrote:

    ....................

    Marc> Agreed. There were a couple of other delayed posts
    Marc> that came through as well.
    >> 
    Marc> I added Martin here, as an FYI.
    >> 
    Marc> Thanks,
    >> 
    Marc> Marc
    >> 
    >> No problems with Hypatia, just with Martin ...
    >> He has been away at a nice workshop in Treviso
    >> "Robust Statistics and R" and then at the CSDA conference in
    >> Cyprus.  After coming back he has been quite busy and only today
    >> found time to act as mailing list moderator to approve those
    >> e-mails that were filtered accidentally ("False positives").   
    >> For some reasons, our spam filter recently gives e-mails from
    >> Yahoo and Hotmail a relatively too high spammyness score;  if
    >> you additionally send "HTML"ified e-mails, chances have become pretty
    >> high your mail won't get through.  And, BTW, only some of them will
    >> occasionally be manually approved by me.
    >> 
    >> Martin

    Marc> Thanks for the clarification Martin.

    Marc> Is there a benign (time sparing) way of tagging these messages when they
    Marc> come through, so that is can be known, first and foremost, that they
    Marc> required manual intervention? The goal being to modify posting behavior
    Marc> and to ultimately minimize the amount of your time that is required on
    Marc> these?

Thank you for the questions, Marc, 
but indeed, I had posed them to myself more than once in the
past.  AFAIK, it's not easily possible, at least not for those
that are caught by Mailman's filters {which for some lists I
made trigger on the spamassassin "*"s}.
Of course we could file a mailman feature request for this.. any
volunteers?

The 'much more probable' spam is not fed to Mailman at all but
put into spam-mailboxes which I skim through somewhat regularly
but not entirely systematically. Also, these (4 and more '*')
messages are very rarely false positives.  These then need even
more manual work, and are actually repostings (though not
visibly in the visible mail headers IIRC) with the advantage
that I could easily add "editorial remarks".

Martin