SoloCDM <deedsmis(_at_)ris(_dot_)net> writes:
I have a very serious complaint with the flags in a .procmailrc file.
With the following example, I'll prove my point. To start with, I'd
like to state that everyone knows how important CPU time has on the
system. It's probably the number one factor in processing.
What's that last sentence supposed to mean? Processing _is_ the use of
CPU cycles. In my experience as a sysadmin, however, most machines are
I/O bound and have plenty of CPU to spare. Is your statement above
based on examining the output of vmstat or sar, or is it a
guess/assumption? BTW, when I say I/O bound, I mean _real_ I/O (i.e.,
SCSI, IDE, ethernet) and not interprocess data transfer (e.g., pipes).
In the following example I want to draw your attention to the flags.
This is significant, because in the Example, I must use the "h" flag
or the body of the message will travel along with the intended
message, but that doesn't come without a price. I must use the last
/dev/null recipe to destroy the body or it will try some unexpected
stunt: e.g. save itself to a file/folder, which it did several times.
(Keep in mind, it was without a header.)
Example:
:0 fh
| ($FORMAIL -rt \
-I "Message-ID:" \
-I "X-Precedence: Personal Mail" \
-I "X-Loop: $MDEST" \
-I "Subject: Message Distribution" \
-I "To: $MDEST" ;\
echo "Message distributed to $MAGENT~>Personal folder." ;\
echo; echo -e "\tRe:\n\t$MSUBJ\n\t$MFROM" ;\
cat $HOME/mail/.signature-notice \
) | $SENDMAIL -oi -t
:0:
/dev/null
I've run every experiment that I could think of to bring about a
different type of recipe, but it is impossible. Nevertheless, I
"Every recipe", "Impossible": I don't think those words mean what you
think they do. Either your testing procedures were broken, or you
failed to try the first recipe without the 'f' flag.
immensely appreciate the assistance I've received on this mailing
list. This is the reason for much of my recent inquires stemming from
the above dilemma, which is out of control -- not on my part, but the
making of .procmailrc.
Since you made the .procmailrc, it _is_ on your part.
I have a long term suggestion for you: stop, read, and think before
posting. On June 7th (less than 3 weeks ago), you posted this _exact_
recipe and a similar question. I replied, pointing out several
incorrect assumption you had made and suggesting that you remove the 'f'
flag and deal with logging issues separately. Apparently you have no
interest in paying attention to what I say, so I'm going to stop
spending my time trying to help you when you don't want my help.
Goodbye, and have a nice life.
Philip Guenther
_______________________________________________
procmail mailing list
procmail(_at_)lists(_dot_)RWTH-Aachen(_dot_)DE
http://MailMan.RWTH-Aachen.DE/mailman/listinfo/procmail