Re: Record blocking optimization

[Tony Carolla <carolla@xxxxxxxxx>]

Thanks for the info.  That 'sequential only' language had me worried.  

<snip>
> That said, on first listen, this sounds like it might lend itself to the
> old RPG cycle, which still works in RPG-IV, if the idea is to optimize.
> It's hard to beat the cycle, though I'm sure buffering optimization
> would get close.  Using primary and secondary, the compiler sets up the
> optimum block size.
</snip>

My curiosity is piqued.  My mindset has been on avoiding the cycle,
and just optimizing this thing as best as I can (read blocking, using
all DSs for input/output/updates, etc).

Now this is a little humbling, but I am a late comer to the game of
RPG (I have used IV only, and was taught to avoid the cycle at all
costs), and I have never written a pgm using the cycle AND primary AND
secondary input files.  I know that if you specify a file as input
primary, every record will be read from the file, and the calc specs
will be repeated for each record, but I am not sure how the secondary
file plays into this.  Is a record automatically read from the
secondary file also?


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