Re: fetchmail for POP3 on WinDOS 2000

[Torsten Bronger <bronger@physik.rwth-aachen.de>]

Halloechen!

Michael R. Wolf <MichaelRWolf@att.net> writes:

> kai.grossjohann@gmx.net (Kai Großjohann) writes:
>
>> Torsten Bronger <bronger@physik.rwth-aachen.de> writes:
>>
>>> I now did the following:
>>>
>>> (file :path "/var/mail/bronger"
>>>       :prescript "fetchmail acaxp.physik.rwth-aachen.de; formail -bs spamc < /var/mail/bronger > /tmp/foo; mv /tmp/foo /var/mail/bronger")
>>>
>>> For this I had to configure fetchmail first (create .fetchmailrc).
>>>
>>> It works.  However, I don't like this solution at all.  What if
>>> fetchmail fails?  Are "&&" better than ";"?  Can this be as robust
>>> as my old pop3 way at all?
>>
>> It seems a good way.  Yes, "&&" might be better than ";".
>
> Or could you do a pipe?
>
>     fetchmail blah-blah | formail > /var/mail/bronger

This would be wonderful!  But I couldn't find a way to make fetchmal
send the mail to stdout.  It wants to sent it to port 25 where it
expects some other program that listens, or it starts the MTA by
itself.

>> Maybe you can also tell fetchmail to invoke formail directly. Or to
>> invoke spamc directly. Yes, invoking spamc seems like a good thing
>> to do, and it ought to be possible, even though I've never used
>> fetchmail.

As said, as far as I can see fetchmail insists on re-sending the
mail.  I didn't even find a possibility that fetchmail writes my
mail to a file!  I had to use sendmail for that.  On the other hand,
maybe one can do a little magic with fetchmail's mda option.

> This works for me:
>
>  [...]
>
> I'm still running in a biforcated world. gnus thinks I'm on a WinDOS
> machine, ergo the "C:/"-rooted path, but fetchmail is being run from
> cygwin and has a "/"-rooted path. Configure as necessary.

Hmm, you use procmail insead of sendmail.  Is this better?

Tschoe,
Torsten.

-- 
Torsten Bronger, aquisgrana