Houston Automatic Spooling Program, origins at NASA (which used IBM systems) to avoid having to dedicate a line printer to a single application which might be crunching data and generating output even slower than the printer could print.  Later became Job Entry Subsystem 2, which as a name seems more oriented to the 'input' side of things (card reader machine) instead of the 'output' side (printers) which is where I always heard the focus had been - but "spooling input" and "spooling output" so that the physical devices are not dedicated to applications was the basic idea.

That was all in the 1960's and my data center still had a working card reader/punch machine on the mainframe until around 1984 or so.  I edited the sysgen file to remove the definition myself after the machine was gone.

>>> Pete Weiss <[log in to unmask]> 02/12/09 2:41 PM >>>
At 14:35 2/12/2009 Thursday, Andrew Bosch wrote:
>The DD parameter and the associated DD line is derived from the old IBM 
>JCL (Job Control Language). It's purpose was to provide data to batch 
>commands for processing within the batch job script.

HASP running under IBM MFT or MVT (VM/CMS hadn't been invented, though predecessor had) on an IBM 7000 series e.g., IBM 7074.  HASP then was take over by IBM and productized as JES2.  Users' group called SHARE.

I'm sure others will chime in :-)

/Pete