I think that the moderators currently use some sort of Un*x shell scripts to assist the moderation process. I am no shell script guru, but I am supposing it would not be too difficult to modify them for an automated check if the moderators did not want to keep track of posting frequency and size by hand.

I'm sorry, but as well as being technically difficult it would be numerically unsound. Our moderators use shell scripts which have patterns of 19 which prevent tampering.

Attached at the end is the shell script sent to me by our moderator Ahmad, exactly as I received it.

A "word" is a Unix file-name, an address component, an English word appearing in a comment, or anything else containing at least one letter.

Some of the shell script is commented out, using the character '#'. We will show that the script is correct even including comments.

A1. There are 228=19*12 words,
A2. of which 38=19*2 are file-names.

Lines containing an even number of words have A3. 95=19*5 letters appearing in file-names,
A4. 76=19*4 letters appearing in non-csh command names, and
A5. 323=19*17 letters in words other than csh keywords.

Lines containing an odd number of words have A6. 437=19*23 letters altogether.

In addition, we note that the important keywords if-then-else-endif appear altogether 38=19*2 times.

Is there a typo?

Unix users will immediately recognise that the script appears to have an error:

echo Respond yes, no, hold, edit, show, m>ail, shell, or quit It is syntactically legal, but from the meaning it is clear that the "&t;" character should not be there. However, removing it decreases the number of words by one, and also decreases the number of file-names by one. This line also changes from having an even number of words to having an odd number. Consequently "fixing" the apparent error will destroy all the features A1-6 listed above. We conclude that this "error" is INTENDED TO BE.

Brendan McKay

Ahmad's shell script, exactly as received.

#!/bin/csh
# split

set mod = "ahm@smi.med.pitt.edu"

foreach i ($*)
  ./filter <$i >$i.out
  set ident = "`grep From: $i.out` `grep Subject: $i.out`"
  if ("$ident" == " ") then
    echo $i has no From: line
  else
#    fgrep "$ident" logfile
#    if ($status == 0) then
#      echo "$ident"
#      echo "Message appears to be a duplicate"
#    endif
  endif
  echo --- $i ---
#  more $i.out
  while (1)
#    echo -n "OK to send? "
#    set ok = $<
   set ok = "y"   #remove with previous line's comment
   if ($ok == "y" || $ok == "yes") then
      /usr/local/bin/inews -h <$i.out
      ack $i.out 
      cp $i.out sent
      echo `date` "$ident" >>logfile
      /bin/mv -f $i.out $i.sent
      /bin/rm -f $i
      break
    else if ($ok == "n" || $ok == "no") then
      echo `date` "$ident" >>rejectlog
      ./rej.sh $i
      /bin/mv -f $i $i.reject
      /bin/rm -f $i.out
      break
    else if ($ok == "h" || $ok == "hold") then
      /bin/mv -f $i $i.hold
      /bin/rm -f $i.out
      break
    else if ($ok == "p") then
      /usr/lib/sendmail $mod < $i
      /bin/rm -f $i $i.out
      break
    else if ($ok == "p2") then
      /bin/mv -f $i $i.pvt
      /bin/rm -f $i.out
      break
    else if ($ok == "q" || $ok == "quit") then
      exit
    else if ($ok == "e" || $ok == "edit") then
      emacs $i.out
    else if ($ok == "s" || $ok == "show") then
      more $i.out
    else if ($ok == "m" || $ok == "mail") then
      onintr -
      cat prolog $i.out >$i.mail
      mm get $i.mail
      onintr
      /bin/rm $i.mail
    else if ($ok == "|" || $ok == "shell") then
      echo Article is $i
      $SHELL
    else
      echo Respond yes, no, hold, edit, show, m>ail, shell, or quit
    endif
  end
end

Overview on numerical features in different scriptures
Answering Islam Home Page