Multics > Sites
14 Jan 1998

Site History: NWGS


US Naval War College, War Gaming Division, Newport, Rhode Island.

Software was developed for this application at a CSC development site: SDF - Software Development Facility, Moorestown NJ

Software was maintained and upgraded at the Software Support Activity (SSA) site at the U.S. Naval War College, Sims Hall, Newport, RI.

There were three production systems:

First Installed


SDF - Software Development Facility, Moorestown NJ

A 2 CPU Level 68 System. This system was originally the ONLY system, installed at the Naval War College in Newport, RI. When the War College upgraded to DPS-8/70s, the SDF got the old Level 68 as a development machine for ENWGS.

SSA - Software Support Activity, Naval War College, Newport, RI

A 2 CPU DPS-8/70 System. This system was used for software debug and testing activities related to the delivered systems (NWC, TTGL, TTGP). This site was controlled by CSC as an on-site support facility. The KST was upsized from 1024 to 1536 by CSC programmers (Robert Matern) in order to support the ENWGS software. Probe was also modified to access the enlarged KST, and add a few unique commands for ENWGS. The lp driver was modified to add optional line numbers to print listings (change done in ALM)! The KST change and all the other modifications were eventually installed at all ENWGS sites.

NWC - U.S. Naval War College, SIMS HALL, Newport, RI

Original Level 68 system supported 64 ICCU Workstations (PDP-11/04 SBC with Sanders Graphic-7 (G7) non-raster display). Upgraded to a 2 CPU DPS-8/70 System that supported 64 SMS Workstations (Intel 286 with Multibus running iRMX-86 OS) and later WISC Workstations (Intel 386 with ISA running iRMX-286 OS). This system was used for actual War Gaming using ENWGS. It had more memory, and a different configuration due to the need to support more X25 networked workstations than the other sites. This system was the last to shut down, in 1996.

The NWC system also had twice the disk storage to support more and larger wargames. The only 2000 track wargames (the annual GLOBAL games) were played on the NWC system. Those games may have dozens of carriers, and hundreds of ships, subs, and aircraft.

TTGL - Tactical Training Group Atlantic, Dam Neck, VA (Near Virginia Beach)

A 2 CPU DPS-8/70 System -- this system was used for actual War Gaming using ENWGS. A small system, only 20 workstations, with 8 more connected via remote link at Little Creek.

TTGP - Tactical Training Group Pacific, San Diego, CA

A 2 CPU DPS-8/70 System -- this system was used for actual War Gaming using ENWGS. A small system, only 20 workstations, with 8 more connected via remote link at Coronado.

Remote Sites

Historical remote sites were at Honolulu (linked to TTGP) and London (linked to NWC).


The Enhanced Naval Warfare Gaming System (ENWGS) is the U. S. Navy's official wargaming system providing multiplatform, multiwarfare, multisite, real-time wargaming and training. The Space and Naval Warfare Systems Command (SPAWARSYSCOM) designed, developed, and maintains ENWGS.

ENWGS is a computer-based simulation system designed to provide realistic wargaming in all areas of naval warfare, including tactical, strategic, and theater operations. The system provides tactical wargaming and decision-making training to battle group staffs, and is used as a strategic and operational wargaming and planning tool. At the strategic level, ENWGS applications range from strategic mobility to logistics and force-level planning for Commander-in-Chief (CINC) naval component commanders. ENWGS has also been used to investigate and examine strategic and operational applications of naval forces within existing war plans, crisis action planning, and Joint Task Force (JTF) contingency operations.

Naval warfare mission areas represented in ENWGS include Air Warfare (AW), Anti-Air Warfare (AAW), Anti-Submarine warfare (ASW), Anti-Surface Warfare (ASUW), Mine Warfare (MIW), Strike Warfare (STW), Amphibious Warfare (AMW), Electronic Warfare (EW), and Operational Deception (OPDEC). Representation can be achieved at a variety of levels from individual units to multi-carrier, theater-level battle forces up to global, multi-theater war. The various ENWGS models consider detailed sea and atmospheric conditions as they apply to each modeled area such as air operations, and weapon and sensor performance. Virtually any area in the world can be represented through adjustment of the parameters describing the sea and atmospheric environment. NWGS/ENWGS has the capability to use a Fictitious Map, a non-real map designed by the Game Sponsor. This was sometimes done for political reasons (to play a game with participants from countries with border disputes, for example) or simply to prevent the purpose of a wargame from being too obvious.

ENWGS provided realistic combat scenario training support prior to, and during, Operation Desert Storm.


1-64 player workstations (alphanumeric display, tactical situation display, workstation processor, printer and automatic status boards).


Written for NWGS in PL/I by Computer Sciences Corporation (CSC), prime contractor since 1979.

Robert Matern adds: There were over 1 million lines of code in PL/1 for the Multics host and 250,000 lines of PL/M for the iRMX workstation. ENWGS is one of the largest programs ever written for Multics (no other site ever had to upsize the KST, to my knowledge!). The new version is the largest program ever written for HP-UX (according to HP!) and the largest for the Rational VADS Ada Compiler (according to Rational!). In porting to HP-UX, virtually every kernel parameter had to be upsized... far more changes than we ever had to make on Multics!


Order of battle, platform characteristics, acquisition and engagement probabilities. Input may be accomplished in the interactive or batch mode.


Intelligence reports, detection reports, unit tracks, battle damage assessment, and summary logistical reports.

Security Classification

(N)ato Secret
(T)op Secret

Only the original NWC NWGS (Level 68 / ICCU) had the Top Secret option; it was dropped when the SMS/WISC workstations were adopted.


The initial purchase was through the Federal Systems Division of Honeywell, Inc.

Site Analysts

Ed Ranzenbach (SDF), Steve Carlock, Jim Falksen.

Field Engineers

Dave Eastman, Dave Pickett, Peter Ames.

[Robert Matern] The Honeywell hardware support at NWC was provided by Peter Ames and Dave Eastman, first onsite then later oncall. I shared an office with Peter and Dave for nearly a year... it had a great view of the bay!

System Administrators

LCDR Mike McCord, Ed Tomasch (SDF), Robert Matern (NWC), Bob Cooper (SDF, NWC, TTGL, TTGP), Don Humbert (SDF), Chuck Liggett (SDF).

Developers (CSC)

Robert Matern, Ed Tomasch, Jim Falksen, Nick Grimes, Joe Sowers, Mike Suman, Jon Buser, Len Birns, Bob Blacker, Dave Campbell, Tony Cavallero, Gil Elston, Ray Heinz, Don Humbert, Bob Knobbe, Chuck Liggett, Bob Morgan, Marion Musto, Les Palinkas, Ron Randal, George Reinis, Dave Robertson, Ed Rose, Paul Rubin, Frank Rybicki, Tony Sacco, Dave Stephenson, Carl Stokes.

Customer (USN)

Lynne Vedder, Steve Vedder, Candy Conwell, Ray Heinz, ? Kruzona, Jerry Lima.

Final Shutdown

ENWGS was rewritten in ADA and rehosted on TAC-3 UNIX machines in 1996.


[Robert Matern] I started work with CSC in Dec.'85 on NWGS Release 4.1. This was the last release before the "Enhanced" NWGS (aka. ENWGS) was released using the DPS-8s. Very sorry to report that all five Multics systems involved with ENWGS (the enhanced version of NWGS) are now defunct... replaced by HPUX 9.07 running on TAC3s (HP-9000/755). Each site has two TAC3 computers to replace each two-CPU Multics machines.

Multics is only a memory on the ENWGS project now... but after working in SunOS and HP-UX for 5 years... I really miss Multics! The 40 MB of ENWGS has somehow turned into 2.1 GB of source & object... Dumping Multics PL/I for UNIX & Ada has really consumed resources... and the system doesn't run as fast, even though it went from two 1 mips processors to two 130 mips ones.

The old ICCU single-board computers were all in a rack in the computer room (to reset a station, you had to run into the CPU room and press the BOOT button on the CORRECT card) with cables running to G7 displays on the Game Floor. Those G7s had an ORANGE color knob (there were four color knobs: orange, yellow, green, blue if I remember correctly<g>); this was helpful when for political reasons the Navy decided not to use RED to represent the enemy... to this day, the color of tracks opposing BLUE (good guys) is ORANGE (enemy) with YELLOW (neutral) as the uninvolved entities. Sometimes a fourth side was used to represent non-aligned entities which might join or switch sides; the color used for this was GREEN.

ENWGS has another interesting area: Tactical Control Directives. The original NWGS had composite-verbs, a sort of multiple keyword which allowed game participants to define a keyword to invoke multiple NWGS functions in one command. This was fine, but lacked the ability to automate a participant. In 1986, Jon Buser, Joe Sowers, and their team at SDF wrote the TCD Compiler to allow the creation of automated forces. TCD was an OPS-5 type rule-based language written using the Multics Reduction Compiler, and on compilation generated PL/1 code which was then compiled. In this way, a TCD could be created which, when called from within NWGS, could invoke NWGS keywords whenever the TCDs rules specified and automate the behavior of a single platform or a carrier battle group or all the forces subordinate to a participant. One simple example was the ZMCS (Maintain Cap Station Coverage) TCD which launched a CAP Aircraft to a particular station (latitude/longitude/altitude) and then launched a replacement for it whenever began its recovery. This allowed players to spend more time thinking about the problem (the game's objective) and less time worrying about the details of NWGS. TCDs could even invoke other TCDs; ZMCS could invoke ZCEN (Cap Engage) to attack if a hostile Aircraft was detected, and then ZMCS would launch the replacement Aircraft for the station immediately. In 1988, I did considerable work improving the performance of the generated PL/1 code to speed up TCD execution and add several functional and debug capabilities. I'll have to see if the TCD design paper (which was presented at several symposiums around the country) is available online...

BTW, I don't know if you've realized it yet... but I'm FAMOUS for !!!! in my code comments... I counted how many there are in that final paragraph of the Application comments in the html... <g>

The NWC system had 3 FNPs, where the other sites (SDF,SSA,TTGL,TTGP) all had two. The NWC system had a switching array to shift workstations to backup channels if the primary channel failed. The NWC system also got extra memory around 1989 doubling its main memory size; this memory, and an extra disk drive, came from a deactivated system somewhere out west...

Robert has collected some ENWGS Humor.

Information from Robert A. Matern, Bob Cooper,, and