🚀 go-pugleaf

Submitted-by: gwyn@smoke.brl.mil (Doug Gwyn)

In article <1ip51dINNp41@ftp.UU.NET> atkinson@tengwar.itd.nrl.navy.mil (Ran Atkinson) writes:
>  Doug Gwyn <gwyn@smoke.brl.mil> has been using such generalised batch
>software on UNIX systems.  I think his stuff might even be freely
>distributable.

Actually, the BRL Multiple-Device Queueing System was originally developed
some 10 years ago by Douglas P. Kingston III with assistance from Michael
J. Muuss.  When DPK moved on to greener pastures, as his last team leader
I inherited the maintenance responsibility for his software, and have kept
it up to date with bug fixes, portability enhancements, and support for a
variety of additional "devices" (some of which are actually interfaces to
services such as FTP, remote lpd, and remote MDQS).  The current release
is kept available for anonymous FTP on host FTP.ARL.ARMY.MIL (aka
FTP.BRL.MIL) as a compressed tar archive named arch/mdqs.tar.Z.  (Remember
to set BINARY transfer mode before fetching.)

The advantages of MDQS are that it supports multiple servers per queue,
and each server is expected to simply deliver the data to the "device"
using whatever protocol the device demands -- extra garbage should be
added before spooling the data.  There is no presumption that the
spooled data is text nor that the device is a printer, although there
is optional support for that mode of operation (banner pages, etc.)
There are some shell scripts included to emulate UNIX "lp" and "lpr"
interfaces (also QMS/Imagen's "ipr"); the "lpr" one is being overhauled
and a much improved emulation of modern "lpr"s should be available (in
an updated MDQS distribution) soon.

For further information, look in some really old USENIX Conference
Proceedings, or retrieve the MDQS distribution, unpack it, and read
the documentation (which includes the USENIX paper and an
administration/internals guide).

My personal opinion from having worked with/on MDQS for years is that
it is a good UNIXy spooling system that can and should be extended to
meet evolving requirements.  It covers pretty much the same functions
as the more common "lp" and "lpr" (lpd-based) spoolers and on several
of our systems it totally replaced them.  ("lpr"/lpd was resurrected
unnecessarily on some local systems when their system administrators
didn't know how to get some commercial software to work with MDQS.)


Volume-Number: Volume 30, Number 24