Summarized by Dave Sullivan, Harvard |
As people sat at their tables following dinner on Monday, a special session was held honoring Mark Weiser, a beloved member of the systems research community and program chair of the 1995 SOSP, who passed away from cancer earlier this year.
The session was organized and introduced by Doug Terry, who worked with Mark for many years at Xerox PARC, where Mark served as manager of the Computer Science Lab and later as chief technologist. Doug still remembers the first time that he met Mark. It was at the 1984 USENIX Conference, and Mark was raving about window systems. "He loved to rave," Doug recalled. "He had a passion for technology that I've rarely seen in someone." Doug also mentioned Mark's vision for ubiquitous computing, a term that he coined. Mark passionately pursued that vision, and it had a tremendous impact on many of the papers in recent SOSPs. Doug then brought out "the appropriate attire" for the event, a pair of red suspenders like the ones that were Mark's trademark. In an impromptu development, each of the speakers "wore" the suspenders while sharing stories of Mark and how he profoundly influenced systems research and the members of our community.
Marvin Theimer, who worked with Mark on ubiquitous computing at PARC, described what it was like to work with him. He mentioned Mark's "incredibly infectious enthusiasm and irreverence," and how he always went for the high-risk, high-payoff option. With ubiquitous computing, for example, they did almost everything from scratch, trying not to compromise on anything. While the risks did catch up with them at times, Mark's enthusiasm and risk-taking made it incredibly exciting to work with him; it felt like "working on the future."
M. Satyanarayanan spoke next, recalling the reaction to Mark's 1992 distinguished lecture at CMU on ubiquitous computing; most of the audience was "dumbfounded at the audacity of his vision." To help us appreciate the reaction, Satya reminded us that the Web did not exist at that point, laptops were slow and heavy, and there were no PDAs or wireless networks. Indeed, we still don't have the technology needed to realize Mark's vision. Given this fact, why was Mark so bold and tireless in enunciating his vision? Satya attributed it to Mark's realization that reaching for "a truly tantalizing vision" allows us to make progress that we could not make otherwise. "In this sense," Satya said, "I truly believe that Mark Weiser is in the company of people like one of the founders of our field, Charles Babbage, who enunciated a vision that took a century to realize." He spoke of Mark's courage in putting his reputation on the line in articulating his vision. Mark's legacy to all of us, Satya said, is "to never be afraid to ask the big questions, and to follow your guiding light wherever it takes you."
Maria Ebling spoke of Mark as an energetic researcher who believed in his vision of what the world would be and who loved his work. She recalled his willingness to listen to her ideas and to encourage her. Maria also remembered how Mark "bounced" with excitement after being appointed as program chair of the 1995 SOSP, asking for suggestions about how to make it the best conference it could be, as well as his excitement about the band he played with in his free time. She said that Mark's legacy still influences her work on pervasive computing at IBM.
Mary Baker recalled Mark's willingness to listen to a crazy idea, to find a salvageable part of it, and to give interesting ideas about how to pursue it. He was excellent in discussing both technical ideas and how they affected the people that used them. Mary noted Mark's role as a mentor to students throughout the country. He frequently attended research group retreats at Berkeley, and he would follow up with students by e-mail about their ideas. She shared a story about a memorable bus ride with him to one of the retreats, in which Mark and others put together a call for papers for "ICON '92, the Second International Conference on Nothing." She shared parts of the hilarious CFP, giving us "a flavor of some of his insanity."
Frans Kaashoek spoke about Mark's work as program chair of the '95 SOSP, and his clear goal of broadening systems research as much as possible. He recalled Mark's personal touch: the members of the program committee got personalized mugs that reflected their roles on the committee. Frans also noted Mark's introduction of a CD version of the proceedings and his commitment to having as much of the relevant source code as possible on the CD so that people could attempt to reproduce the results. Frans concluded by mentioning the impact of Mark's vision on his work and those of others at MIT, including the new Oxygen project that has recently been started there.
Doug Terry then opened the microphone to anyone else who wanted to speak. Eric Brewer mentioned Mark's advice to students at the last Berkeley retreat to always have an "elevator pitch" about your work. David Cheriton recalled speaking with Mark when they would pick up their daughters at youth symphony rehearsals. He mentioned Mark's tremendous pride in his children, and the importance of family in Mark's life. Stefan Savage shared several memories of Mark, including his openness to a crazy (but since almost realized) idea of obtaining energy for mobile computers from people's shoes as they walk, his willingness to put Stefan up at his home during Stefan's internship at Xerox PARC, and Mark's tremendous enjoyment of his band.
Jay Lepreau reiterated Mark's commitment to the reproducibility of results--"that we should act like a science and not just be named one." He recalled a prescient "energy-scheduling" paper that Mark contributed to the first OSDI that relied on a type of hardware that didn't then exist, addressing the important issue of reducing power consumption. Jay also mentioned Mark's firm belief that "anybody could do anything," and how this belief extended to the way that he "looked for the best in people and got it."
Andrew Black recalled the courage of Mark's opinions. Mark not only argued for the reproducibility of results, he brought enough CDs to the 1993 Asheville SOSP for everyone to have a copy of the source code for his paper. Amin Vahdat spoke of Mark's feedback to Brad Chen, a student speaker at the same SOSP, "you have set the standard for this and all future SOSPs," which inspired Amin and his classmates to work to produce papers worthy of such praise. Karin Peterson mentioned his personal impact in recruiting people for PARC, his ability to keep open a wide range of possibilities when discussing a research project, and the personal dimension that Mark brought to the workplace, including his participation at Friday night social outings and his willingness to speak about things that didn't directly involve work. Drew Dean spoke of the way that Mark could enliven the lunchtime conversations at PARC.
Carla Ellis concluded this moving tribute to Mark by announcing that SIGOPS plans to establish the SIGOPS Mark Weiser Award, an annual award for "creativity, innovation and vision in operating systems research" to be given to a researcher less than twenty years into his or her research career. Every year, members of the systems research community will have an opportunity to nominate someone for this award with "real vision and creativity, to capture some of Mark's spirit."