Multics > Sites
15 Mar 2024

Site History: CISL

 
Multics Close ⊗
Loading

The Multics contract was signed in 1964, and General Electric committed to contribute programming resources to the joint Multics OS development project. In June 1964, the MIT Project MAC Multics team moved to the fifth floor of the new 545 Technology Square "Alpha" building.

In 1966, GE set up the Cambridge Information Systems Laboratory (CISL) on half of the seventh floor of the Tech Square building. Al Dean was the original manager, reporting to Ed Vance in Phoenix.

CISL made major contributions to the success of Multics by taking over EPL support and creating the Version 1 and Version 2 PL/I compilers. The Documentation group at CISL produced many software manuals for each Multics release. CISL developers used the MIT 645 for development runs until about 1974, when CISL got its own 6180.

In 1969, CISL moved to the third floor of the new 575 Tech Square "Beta" building across the courtyard. In 1970, GE sold its computer business to Honeywell (the "merger"), and CISL became a Honeywell location. About this time, the late Charlie Clingen became the manager of CISL.

Project MAC and GE/Honeywell staff used the MIT 645 Multics service for programming, documentation, and mail. In the 1970s, several people from MIT joined CISL, including Jim Mills, Steve Webber, Noel Morris, Paul Green, and Tom Van Vleck. Major CISL projects in the 1970s included improving the scheduling, virtual memory, system security, and storage systems, and projects with Honeywell in Phoenix on Multics hardware plans and improvements, including designing and implementing Multics for the new model 6180 Multics machine, and assisting with Honeywell marking and customer benchmarks.

In 1980, Charlie Clingen moved to Phoenix, as Manager: Multics Development Center, and John Gintell became Manager: Multics System Development and led CISL. The CISL group made major contributions to the B2 security project, the Multics Data Management facility, and many communications protocols.

The CISL machine was moved to the Cambridge Center building in Kendall Square when CISL moved there in 09/83. Multics development was canceled by Honeywell in July 1985, and CISL was closed. Some CISL personnel transferred to Honeywell Billerica to work on the Opus project.

Timeline, Locations and Moves

CISL Timeline 1965 - 1986
TYPEYEARLSTYLEHEIGHTTEXTTSTYLE
back1966#dddddd1969545 TS 7th fl
back1969#ffdddd1983575 TS 3d fl
back1983#ddffdd1986Kendall Sq
axis1966
axis1970
axis1980
axis1989
data19701Merger
data19741CISL 6180
data19861CISL closed 06/86

CISL Development Machine

mainframe

Configuration

At some time in the 70's, a second CPU (an 8/70M), more memory, and paging device were added at the Tech Square site.

Operations

The people who operated the CISL machine included Don Fidler, Angelo Grieco, and Ed Winslow.

Application areas

The CISL Multics machine was used at CISL to test new versions of the operating system, especially hardcore modifications. It was available to system programmers who wished to do a development run on a signup basis.

Eventually, it was run as a "service system" at CISL during mid-day, running the experimental-system-de jour on a beta-test basis. This also gave systems programmers the opportunity to build and preload software for development runs.

The CISL machine was used for development of MCS, the GIOC replacement software. It was also connected to the ARPANet, using an ABSI board. See Networking and Communications for more discussion of communications and network development at CISL, including TCP/IP and Hyperchannel support.

The article on the Multics Development Process describes the activities of Multics developers.

Anecdotes

The People Pictures page has a picture gallery of CISL Multicians.

Ron Riedesel's story, Visiting CISL, describes his trips from Phoenix to visit CISL in the early 70s, and has a CISL group photo.

Keith Loepere Story: CISL Farewell.

Ed Ranzenbach Story: CISL Wake.

Also see Keith Loepere's Multics Rug from CISL in the Memorabilia section.

The CISL Machine

[JWG] One point about the CISL machine at both sites is that it was shut down and powered off every single night! I believe it is the only GCOS or Multics system that was ever run this way. The specs said that you could, but Field Engineering always complained, and I'm sure this contributed to some of the disk failures.

[JWG] One other thing that always blew my mind is that we had a 355 with maximum of everything that was 12 feet long! I believe we abandoned this when we moved to Cambridge Center.

[BSG] As configuration and setup of new Multics site, including the registering of projects and users is a non-trivial procedure to learn and follow, a "development hierarchy" with a small site all set up was loaded from a backup tape as part of the setup for a run. One of the endearing memories of this was the use of the IDs of a handful of prominent, senior Multicians of the era as the registered users in this environment. Junior programmers testing software at that time would say, "Gee, do I want to be Steve Webber or Noel Morris today?" (When CISL began to run its own machine as a service during the day and all CISLians had accounts and real files there, this practice ended).

[Richard Shetron] The CISL 6180 had 2 memory boxes in the late 70's. One core, one semiconductor. The CISL people said they had a 2 page program and could tell by the run time which page was in which type of memory.

Final Shutdown

The CISL machine was moved to Honeywell Billerica when CISL closed, in 06/86.

Internet Announcement of CISL shutdown

[JSLove@MIT-MULTICS.ARPA.UUCP] Posted in mod.protocols.uucp on 5/23/86:

Postmaster Alert -- CISL-SERVICE-MULTICS.ARPA => MIT-MULTICS.ARPA

Yesterday, 22 May 1986, CISL-SERVICE-MULTICS.ARPA (10.3.0.31) went off the air. The Cambridge Information Systems Lab closed its doors today.

At some future time, probably in about 2 months, a new Multics will come on the ARPAnet at a different Honeywell location in Billerica, MA. The name and address are not definitely known at this time.

For the moment, all mail intended for recipients at CISL-SERVICE-MULTICS.ARPA should be sent to MIT-MULTICS.ARPA (10.0.0.6) instead. This includes mail destined for other non-Internet sites which uses CISL-DEVELOPMENT-MULTICS.ARPA as a gateway. For example, Multics users in Phoenix, AZ, receive mail as User%PCO-MULTICS@CISL-DEVELOPMENT-MULTICS.ARPA.

Written 20 Sep 1995. Thanks for many updates since then.