Skip to content

Mark Weiser and Ubiquitous Computing

Abstract

Mark Weiser (1952–1999) never took a bachelor’s degree, joined the computer science faculty of the University of Maryland in 1979, and wrote the 1981 paper that founded program slicing, a debugging technique still taught today. At Xerox PARC, the laboratory that had built the personal computer, he coined the term ubiquitous computing in 1988 and argued that the PC was a stage to be passed through: computers would come in hundreds per room, in the sizes of sticky notes, notepads and blackboards, and would have to stop demanding attention. He led PARC’s Computer Science Laboratory, became its first chief technology officer, and played drums in Severe Tire Damage, the band whose 1993 gig at PARC went out live over the internet’s multicast backbone. He died of cancer at 46 in April 1999.

Chicago, Michigan, Maryland

Mark David Weiser was born in Chicago on 23 July 1952 and named after his father, David. The family moved to Mount Carroll, Illinois, when he was four, where as a boy he played in a garage band called the Outsiders; at 14 they moved to Ithaca, New York, and a year later to Stony Brook on Long Island, where he finished high school. His first paid programming job, in 1967, was for the chemistry department of Cornell University.

He studied philosophy at New College of Florida in Sarasota and left in his second year when the money ran out. He never received a bachelor’s degree. He worked as a programmer for Omnitext from 1970 to 1975 and co-founded two small companies in Ann Arbor, Cerberus Inc. and Portable Information Systems, which he ran from 1973 to 1976. The University of Michigan admitted him directly to its master’s program in Computer and Communication Sciences: he took the MS in 1976 and the PhD in 1979.

The University of Maryland hired him as an assistant professor of computer science in 1979. He directed a departmental laboratory from 1981, was promoted to associate professor in 1984 and became associate chairman of the department in 1986. He presented the Maryland Window System at the USENIX conference in 1984.

Program Slicing

Weiser’s 1979 dissertation, “Program Slices: Formal, Psychological, and Practical Investigations of an Automatic Program Abstraction Method”, started from how programmers debug: they cut a program down in their heads to the statements that could affect the variable they are looking at. He published the technique as “Program Slicing” at the fifth International Conference on Software Engineering in March 1981 and in IEEE Transactions on Software Engineering in July 1984, and the experiment behind the psychological claim as “Programmers Use Slices When Debugging” in Communications of the ACM in July 1982. A slice is that reduced program: given a statement and a set of variables, it keeps every statement that can influence those variables’ values at that point and drops the rest, and it still runs. The 1981 paper gave a dataflow algorithm for computing an approximation of a slice automatically.

The idea spread slowly. A later survey by David Binkley and Keith Gallagher counted about ten citations a year for the paper from 1984 to 1989 and about twenty a year in 1990 and 1991. In 1993 the ICSE community gave it its “Best Paper Ten Years Later” award. Slicing became a standard technique in program analysis, testing and maintenance tools.

PARC

Weiser joined Xerox PARC in May 1987 as a member of the research staff and became head of its Computer Science Laboratory in November 1988, the lab where Butler Lampson and Chuck Thacker had built the Alto. He kept doing systems research. With Hans-J. Boehm he wrote “Garbage Collection in an Uncooperative Environment” (Software: Practice and Experience, September 1988), which described a collector for C programs that needs no help from the compiler: it treats any word that looks like a pointer as one. The Boehm-Demers-Weiser collector built on that design was later used by the GNU Java compiler’s runtime, the Mono .NET implementation, the GNU Objective-C runtime and Mozilla’s leak detector (see Garbage Collection). With Alan Demers and Carl Hauser he published the Portable Common Runtime in 1989, a layer that gave several languages shared threads, garbage collection and I/O across operating systems.

Ubiquitous Computing

In 1988 Weiser started a research program at PARC that he called ubiquitous computing. The premise was that the personal computer, one person in front of one screen, was the second of three stages: mainframes had been shared by many people, the PC belonged to one, and in the third stage each person would use hundreds of computers without thinking of them as computers. In a 1989 talk to the MIT Club of Northern California he said the approach would “strive for a maximal mixing of humans and computation.”

He set it out for a general audience in “The Computer for the 21st Century” in the September 1991 issue of Scientific American. The article opens: “The most profound technologies are those that disappear.” His example was the electric motor: a factory around 1900 had one engine driving dozens or hundreds of machines through shafts and pulleys, while the shop manual of a typical car listed twenty-two motors and twenty-five solenoids. Looking around a room, he found more than a hundred inch-scale writing surfaces (labels, wall notes, book spines), ten or twenty foot-scale ones and one or two boards, and set PARC’s goal accordingly: “hundreds of computers per room.”

PARC built all three sizes. Tabs were inch-scale, palm-sized devices; pads were foot-scale, closer to a sheet of paper; boards were yard-scale wall displays. The Liveboards built by Richard Bruce and Scott Elrod measured about 40 by 60 inches with 1024×768 black-and-white pixels, and enough were built that they stood in ordinary conference rooms without a booking sheet. The ParcTab, designed by Roy Want’s group, weighed 215 grams, had a 128×64 touch screen and three buttons, and talked by infrared to a transceiver in each room, so that the building knew which room each tab was in. The first ParcTab system went live in March 1993 with 20 users; the second, in April 1994, had about 41 (the full account is in How We Use Computers). Weiser set out the engineering problems, from wireless bandwidth to power and location tracking, in “Some Computer Science Issues in Ubiquitous Computing” in Communications of the ACM in July 1993.

The 1991 article positioned the idea against virtual reality, which Weiser called “only a map, not a territory”: it built a world inside the computer, where he wanted computers drawn out into the world, a process some at PARC called “embodied virtuality”. He also named the main risk. Hundreds of networked computers per room, all able to sense the people near them, “have the potential to make totalitarianism up to now seem like sheerest anarchy.” His proposed safeguard, borrowed from Jim Morris of Carnegie Mellon, was to give computer systems the same privacy protections as the physical world and no more, so that a break-in would leave the digital equivalent of fingerprints.

The article’s last paragraph promised that computers that fit the human environment “will make using a computer as refreshing as taking a walk in the woods.” Bill Gates listed it as required reading.

Calm Technology

By the mid-1990s Weiser was asking what such devices should feel like. With John Seely Brown he wrote “Designing Calm Technology”, dated 21 December 1995, which defined calm technology as technology that “engages both the center and the periphery of our attention” and moves between the two: most of the time it stays at the edge of awareness, the way an engine’s sound does while driving, and it comes to the centre only when something changes.

The paper’s example was the Dangling String by the artist Natalie Jeremijenko, an eight-foot piece of plastic spaghetti hanging from a small motor in the ceiling of a PARC hallway. The motor was wired to a nearby Ethernet cable, so each packet made the string twitch: light traffic gave a small movement every few seconds, heavy traffic made it whirl with a characteristic noise, and people in the surrounding offices knew the state of the network without looking at a screen. It used no software and a few dollars of hardware. A revised version, “The Coming Age of Calm Technology”, was reprinted in the anthology Beyond Calculation: The Next Fifty Years of Computing, edited by Peter Denning and Robert Metcalfe.

Severe Tire Damage

Outside work Weiser was the drummer of Severe Tire Damage, a rock band of computer scientists from Xerox PARC, DEC’s Systems Research Center and Apple. On 24 June 1993 the band played a gig at PARC while, elsewhere in the building, researchers were working on the MBone, the multicast backbone built on Steve Deering’s IP multicast to carry audio and video across the internet. The performance went out over it and was watched live as far away as Australia; it is usually cited as the first live band on the internet.

On 18 November 1994 the Rolling Stones broadcast part of a concert over the MBone. Severe Tire Damage went out over the MBone ahead of them as an uninvited opening act. The Stones told The New York Times that the surprise opening act “was a good reminder of the democratic nature of the Internet.”

Chief Technologist

In 1996 Weiser became PARC’s first chief technology officer. He was program chair of the ACM Symposium on Operating Systems Principles in December 1995 and sat on the board of the Computing Research Association, where he started its annual meetings for the heads of industrial research laboratories. By his death he had written or co-written more than 75 technical papers and held several patents.

He died of cancer in Palo Alto on 27 April 1999, aged 46. ACM SIGMOBILE gave him its Outstanding Contribution Award that year. He was survived by his wife, Victoria Reich, and their daughters, Nicole and Corinne. Reich gave his papers to Stanford University in May 1999; the collection fills 94 linear feet plus 71 optical discs and 49 floppy disks, and one box is closed until 2048. ACM SIGOPS established the Mark Weiser Award for creativity and innovation in operating systems research in 2001.

Dead End: The Calm That Did Not Come

The hardware arrived much as Weiser described it. The tab became the smartphone, the pad the tablet, the board the conference-room display, and the room that knows who is in it became any building with Wi-Fi location and badge readers. Ubiquitous computing grew into a research field and, under other names, into the Internet of Things. Japan’s TRON project had argued for computers in everyday objects since 1984.

The calm did not arrive. The companies that built the devices earned their money from attention, and the result is described in The Attention Economy. Critics at PARC had seen a different problem. Lucy Suchman worried about privacy and read ubiquitous computing as the wish for “a world which is always familiar”, one that spares its users the unfamiliar encounters that change them. Reviewing John Tinnell’s 2023 biography, Ben Tarnoff judged Weiser largely unconcerned with mass surveillance; the 1991 article had named surveillance as the field’s main risk on its last page. Weiser died eight years before the iPhone.

📚 Sources