4. What and Why.
Projects and PRJ codes.
What is in the SAILDART Archive ? Perhaps surprising to some folk, who did not work inside the A.I. labs in the 1970s, the bulk of the data on the SAIL backup tapes can not be considered as central to Artificial Intelligence Research. So first let us digress on what work was supposed to have be done under the banner depicted as Artificial Intelligence, followed abruptly by a technical dive into project and PRJ codes, which are the lowest level metadata labels on each SAILDART item.
In the SAIL file system each file belonged to one programmer PRG and to one project PRJ as recorded in the master file directory, the MFD. For a primitive file system that lacked nested folders, the PRJ codes were used to partition an individual's work into sub areas such as [1,BGB] and [2,BGB]. On occasion the PRJ code identified a common group such as the Stanford course, "LISP 206" would have PRJ code 206 for students [206,ABC] and [206,XYZ]. In addition, the accounting system had its own set of Project codes, to track time sharing and disk space usage.
SAIL project code
The default login PRJ code was the numeral one, I would often login as 1,BGB. Most programmers were free to enter whatever PRJ code they liked at login to create a new PRJ directory areas to group their files like we do today using folders. Sometimes work groups of people used the PRJ code as you would a GNU/Linux group name. For example, when a Stanford class was issued accounts for their course work, their login PRJ number was their course number.
The official SAIL Topics Overview
In 2007 Les Earnest wrote citing a 1973 summary - "Before getting into the sordid history of data archiving problems encountered in SAIL, let me [G] [G] Les Earnest review some of the reasons for saving their records, specifically listings of research and accomplishments that we like to brag about. Here are the main research topics undertaken at SAIL in the late 1960s through ‘70s and some of the resulting spin-off technologies."
SAIL Research Topics [Earnest73]
* Robotics
+ Vision
+ Mechanical assembly
+ Vehicle navigation & guidance
* Heuristic programming
+ Theorem proving
+ Automatic program generation
+ Symbolic computation
+ Board games
+ DENDRAL
* Theory
+ Mathematical theory of computation
+ Representation theory
+ Grammatical inference
* Natural language
+ Speech recognition
+ Semantics
+ Machine translation
* Planetary image processing
* Computer music synthesis
The remainder of Les Earnest’s Y3K paper is here, and his Y3K video talk is there.
SAIL file system project codes
SAIL accounting system project codes
Project areas with archive specialists.
In recent years the following SailDart project areas have received attention as computer history. Here are the active areas and the names of the specialists and enthusiasts. Consider this section as an invitation to contact people who have expressed an interest in a narrow SailDart history topic that you are interested in. Be warned that the SailDart archive is just a small splinter in the history of the big topics such as Artificial Intelligence, LISP or Robotics. The SailDart is only of middling size within the subject of DEC 36-bit software.
4.5.1 PUB
Larry Tesler
4.5.2 TEX
Arthur Keller
4.5.3 LISP
Paul McJones
4.5.4 Music
Andrew Nelson
4.5.5 PARRY the Paranoid Simulation
Adrian Cornforth
4.5.6 Spacewar
Most recently the authors.
4.5.7 Mathematical Theory of Computation
David McQueen
4.5.8 System and WAITS
Baumgart is working on SYSTEM, Gorin and Frost are working on WAITS.
Stanford University Drawing System
Foonly
I am surprised at how many people were touched by Foonly, since in 1979 when I worked near the F1 Foonly at III it was at a low point with just five people (if you include Dave Dyer and myself). Before and after that time there were more Foonly people. The three Foonly principals are Poole, Petit and Holloway; aka Dave, Phil and Jack; aka DWP, PMP and H for Holloway. Single character PRG codes were an MIT A.I. Lab fashion for the innermost wizards such as G Greenblatt and H Holloway. The [F,*] SailDart files that are Foonly people include PRG codes DWP, PMP, H and then AK, FW, BO, TAG, TVR. AK is Allan Kotok, who wrote a six page, Foonly Blurb, dated 30 August 1972. The document in HTML is at FOONLY.BLB[F,AK] and as a scanned PDF from line printer LPT paper is here. The latter illustrates the characteristic appearance of horizontal character glyph alignment due to the drum hammer timing, study the line numbering on the left, columns one and three hammers fired earlier than column two.
36-bit computers
Before DEC there were IBM 36-bit computers such as the IBM-7094 from which LISP got its CAR and CDR, as well as the one of its kind TX2 at Lincoln Labs on which my mentors Les Earnest and Ivan Sutherland started their careers. Although I now have several books concerning 36-bit computers prior to the KA10; I do not think this document is the proper place for reviewing them; except to point out the photograph of Thomas Watson,Sr. at the IBM-701 in 1952. figure Pictures/people/ibm-701-watson.jpg Figure 4.1 Thomas Watson at the console of the IBM-701 in 1952
The Stanford CART
Earnest, Schmidt, Baumgart and Moravec.
Earnest : FINGER and FONDLE.
Frost : AP news.
Gorin : DART, DSKUSE, ACCT, RALPH, LOGIN, SPELL.
Baumgart : GEOMED, CRE, EDFONT, XIP, XAP.
Stanford University Course Work
CS206 LISP CS222 Music CS220 Music CS204 Programming CS227 Robotics
Orphan Project Areas
The S1 people haven’t surfaced yet to look at or comment on their SailDart material. S1 SUN MICRO SYSTEMS CISCO MARS image processing
Social Studies
Volley Ball Spring Orgy Prancing Pony Sauna Zotts
Gender, Race, Politics and Economic Class Topics
On the edges of the SailDart collection are remarks on the issues of gender, race and class. At the SAIL reunions it is obvious that we are almost all old white males and that we are adequately rich. Looking at the archive there is a bulletin board thread on date rape, but little or no discussion of race or economic class; except for the Les Earnest story about writing mongrel when applying for a security clearance.