Hey Jame!   
      
   Perhaps this might illustrate what I am thinking about with respect to   
   FTN headers and binary data.  Just looking at one variable, say my   
   point's point number 9, then this works;   
      
     binary_origin_point=$(echo 9 | ascii2binary -t us)   
      
   which yeilds a 16 bit binary number that can be later echoed to a header   
   with the appropriate offset.  However given that most of the header   
   information is static with respect to a point and it's bossnode then a   
   one time creation of stored binary data with appropriate offsets would   
   be all that is required and that can be tacked on to the start of a   
   outbound packet followed by prepacked messages, properly terminated, and   
   then fired off to the bossnode.  Doing that results in far less   
   resources being required when sending out messages.  With a bit of   
   forsight a simular strategy could be adapted for a number of different   
   Fido addresses if indeed that is an issue as in the case of a node with   
   many different destinations, such as a bossnode of points, and in that   
   case only the destination point address is dynamic and most of the pkt   
   header is still static data wrt the bossnode.  Having said all that I   
   believe the idea could be extended beyond bossnodes and points but I   
   figure this is a good place to start this type of conversation.   
      
   Bottomline is the above idea is lean and mean and shouldn't cause grief   
   amongst anything upstream and I can keep the local messages and/or   
   scripts in the native text format, which now includes utf-8   
   capabilities if it ever matters.   
      
   Life is good,   
   Maurice   
      
   --- WeBeBashing with VIM - Vi IMproved 7.3   
    * Origin: Pointy Stick Society (1:261/38.9)