The Multics History Project (MHP) was begun in June 2004 by Olin Sibert and Roger Roach. Roger Roach worked on CTSS, then Multics, eventually as MIT's IS director, and retired in 2005. Olin Sibert worked on Multics (initially for Roger, then at Honeywell CISL, then independently). At the Multics Reunion (June 2004), we decided to try preserving the archives of Multics historical information that Roger had maintained since the early 1970s.
The MHP began with a Tactical Goal, to preserve the MIT-Multics archives. Over time, the focus has expanded to include the preservation of other collections of Multics information. In addition, a Strategic Goal has emerged, namely to collect enough information to allow a Multics simulator to be created and operated. [Saturday 08 November: Harry Reed's simulator for the Multics CPU is able to boot and run Multics.]
We work with other people interested in complementary historical activities: The Computer History Museum in Mountain View, California, has a non-functioning Multics machine and several troves of Multics documents as part of its collection. Al Kossow maintains bitsavers.org, which has scanned many historic computer manuals, including nearly 100 Multics manuals that are part of the CHM collection. The Multicians website preserves published papers, stories, photographs, articles, and source code about Multics.
Multics Archives at MIT
The MIT Collection is a large collection of Multics-related material maintained by Roger Roach since the 1969 beginning of MIT's Multics computer service. Its documents cover MIT's participation in the development process from the beginning through the late 1980's. The MIT Collection is estimated to represent approximately 100,000 sheets (185,000 page images) of non-duplicate scannable material.
MIT-Multics was a campus computing utility, run as a service to MIT and MIT-associated customers. MIT's Information Processing Services organization played a major development and QA role under contract to Honeywell (through 1984). This organization was administratively separate from the original Project MAC / Laboratory for Computer Science (LCS) development team, but there was close cooperation through the whole time, and personnel moved from one organization to another occasionally.
The MIT-Multics archive is most complete for later (post-1975) material, with a focus on Multics as a commercial product. Some material was not saved: for example we have no Multics Administrative Bulletins (MABs), which described short-term staff assignments and activities. The MIT archives also lack post-MR11 material, since the relationship between MIT and Honeywell ended at MR11. Furthermore, old material, for example from the 645 era, is quite incomplete.
LCS Multics Archive
There is also a large collection of Multics information that was archived by Project MAC/Laboratory for Computer Science (LCS), the Original Multics development organization. This information contains both data stored by LCS, and LCS staff personal files. These archives include published documents, internal memos, listings, and personal and business files.
The existence of this collection came to light February 2006, and it is just now being investigated.
Currently, about 85% of the MIT-Multics archive documents have been scanned.
A slide presentation (900K PDF) given by Olin Sibert to the CHM on 18 Feb 2006 is available with more detail.
How You Can Help
Do you have any Multics material? Take a look at our wish list and see what we're still looking for, and please get in touch if you'd like help out.