ml> in what way? there is no zone 6 because there is no one there... the
 ml> bad decision was in not simply reading until the end of the file OR
 ml> until the next ZONE keyword ;)

 PQ> Cheeky!  You do realise how long a sequential read on a nodelist
 PQ> takes using batch stuff, don't you.  I found a shortcut method that
 PQ> worked excellently.  Of course I'm upset.  :)

hahaha, yes, i know that some .bat processing can be slow... that's one
reason i run 4dos ;) ;)

you could always build a (new) zone index for each new nodelist and then
jump to the zone being searched to do your sequential scan, couldn't you?
that's what i'd do in pascal, asm or C if i was to (try) to write something
like this in a compiled language...

part of the problem could lie in the tools being used, too... grep is a
powerful tool that one can put to many textfile processing jobs but some
textfile database processing jobs can be a real b!tch to do ;)

)\/(ark