Re: Record blocking optimization

<snip>
> Secondary files are optional.  In general it works like this, it reads all
> records from the primary file, and then reads all records from Each
> secondary file in the order they are listed.  
</snip>
I think I understand.  When you say it reads all recs from primary,
and then from secondary, I assume you mean that all files are read
from on each loop through.  A free format equivalent (minus record
blocking) would look like this:

Dow Not %EOF(primary)
     Read primary;
     Read secondary;
     // Calc specs
EndDo;

Right?  This might be interesting to try, although having seven pairs
of files to compare/synchronize, I would have to write seven programs.
 But if it is faster, I might try it.

One last question.  There is a possibility that a record might be
missing from either file, causing the synchronization of the input
streams to be out of whack.  If this is the case, Can I issue a READ
op in the calc specs to either file to re-synch the files?


-- 
"Enter any 11-digit prime number to continue..."