John R. "Jack" Callahan:
I was one of Mark's orphaned Computer Science
Ph.D. students at the University of Maryland, College Park (UMCP)
when he left to join Xerox PARC permanently in 1987. I remember
when Mark invited all of his graduate assistants to his house
one night and asked us our opinion of his plans to leave UMCP.
He cared deeply about our futures and factored us into his decision.
Without exception, we all supported Mark's decision to move to
PARC. We felt it was the best thing for him because it seemed
to us that this was an important dream that he wanted to fulfill.
We all wanted the best for our friend and mentor.
I actually met Mark Weiser at Xerox while
working for the Office Systems Division/OSBU North in the summer
of 1985 on Geng Road in Palo Alto. Mark would come over to play
Go at OSD in the evenings. He was working summers at PARC at
the time but was still a CS Professor at UMCP As an undergrad
at UMCP, I did not know him, but when I returned in the Fall
of 1985, I wanted to work for him as an Graduate Research Assistant.
He liked my experience on the Xerox platform (Mesa, Tajo, ViewPoint,
etc.) and he needed students with research skills on Xerox systems
(UMCP was a Xerox University Affiliate). As a UNIX guru, Mark
had a standard requirement of his RAs: you must be able to write
a forking program under UNIX in C. I told him I didn't know UNIX
that well, but I showed him an old program I wrote when working
for the Defense Department doing PL/1 work under Multics as a
coop student. He immediately said "You're hired!" I
later understood the relationship between UNIX and Multics.
I finished my Ph.D. in 1993 under Jim Purtilo
who joined UMCP after Mark left for Xerox (indeed, Jim Purtilo
moved into Mark's old office at UMCP). After I joined West Virginia
University, Raghu Karinthi (another Weiser orphan) and I learned
that Mark was going to present a brand new talk on Ubiquitous
Computing at Carnegie-Mellon University on a Friday morning (it
was sometime in 1993). We found out that Mark was flying in to
Pittsburgh on Thursday night and we wanted to meet him for dinner.
After some exchange of email, he agreed to fly out earlier on
Thursday so that he could come down to visit us at West Virginia
University in Morgantown, WV (90 minutes south of Pittsburgh).
I drove up to Pittsburgh to meet his plane and bring him down
to Morgantown. As soon as we were in my car, he pops open his
laptop, wireless modem, and starts to send/receive email he had
been writing his wife on the airplane. Mark was always a "first"
when it came to using new technology and was always communicating
with his family via email long before the Internet boom. (Anyone
who worked in Mark's office at UMCP knew that he had a "talking
clock" that would announce the time each minute. God, that
clock was annoying).
So, Mark gave his first presentation of his
new Ubicomp speech to a packed crowd at WVU in Morgantown, WV
the day before he presented the same talk at CMU. Given his busy
schedule, he could have easily declined to visit, but Mark was
always a giving person. We took him out to dinner and Raghu drove
him back to Pittsburgh late that evening. I guess we still owe
CMU some $$$ for hosting his airfare and hotel :-).
The next time I saw Mark was at in his office
PARC during a brief visit in the summer of 1993. He showed me
the MBone work done by Steve Deering at Stanford and Ron Frederick's
nv tool. I returned to West Virginia and immediately installed
an mrouted on our server and we became part of the early MBone.
I remember seeing Mark's name constantly appear amongst the many
listeners in the vat audio tool. One of my students, Todd Montgomery,
was inspired by Mark's previous visit to WVU and by the MBone
work. He has since implemented the Reliable Multicast Protocol
(RMP), worked in the area of multicast protocols, and is completing
his Ph.D. in 2000.
Mark has touched the lives of many "students"
both directly and indirectly. His work was truly inspired and
he inspired the same energy into those around him. He used to
say that a true programmer was someone who would develop computer
software no matter what the pay because they loved the sheer
thrill of programming. I feel extremely blessed to have known
and worked with such a genuine soul. I will miss him.
Mark Weiser played a pivotal role in the development
of distributed computing here at the University of Maryland.
While he was still an assistant professor, Mark directed our
laboratory facilities during a period when we acquired our first
workstations with grants from NSF and Xerox. Mark's own research
interests also evolved from program slicing to heterogeneous
distributed systems as he studied multiple-protocol hosts and
network-independent user interfaces.
Mark was an inspirational advisor and colleague.
He attracted many outstanding undergraduates (e.g., Chris Torek,
Steve Miller, and James O'Toole) to work with him on projects
in the lab, and collaborated with other faculty and graduate
students on projects in visual programming, program testing,
and user-interface design.
When Mark taught his computer science courses
at College Park in the '80s, he would typically begin by asking
the class "What's been good"? The responses from each
student in turn were often funny, and so each session started
with a sunny atmosphere. This was all a part of his great talent
for encouraging optimism and creativity in those he worked with,
Charles K. Kao:
When I became one the Trustees of the Institute
for the Future I had the fortune of getting to know Dr. Mark
Weiser. I was looking forward to knowing Dr. Weiser more and
working with him. It was most unfortunate that our interaction
was cut short after just one Trustee meeting. IFTF, Xerox Park,
and the world lost one of the best technology innovators. Through
this website his legacy and vision will continue to provide us
After contacts going back to 1990, our first
face-to-face meeting was supposed to happen this spring. We shared
an interest in what he coined ubiquitous computing. In 1994,
as I started my program in Computational Design at the University
of Wuppertal, he was very enthusiastic about my intention to
make ubiquitous computing the core of this new discipline. We
stayed in touch and exchanged ideas. For the class that Terry
Winograd invited me to give during my stay at Stanford University,
I planned to have a session on ubiquitous computing-- and planned
to have Mark Weiser present. As usual, he reacted promptly to
my e-mail. He was interested but also not sure that he could
make the date I mentioned to him. Not uncharacteristically, he
barely mentioned treatment. And as I arrived mid-March, he was
still considering being present, although he knew that it was
more wishful thinking than a realistic chance. In this last episode,
I read his commitment to his work and to students.
Instead of any eulogy, I prefer to limit my
words to the description of these last contacts with him. We
will continue our work in the direction I discussed with him.
The best that will come of it will be, by virtue of our contact
and his concern for others, dedicated to him.
I met Mark in the mid 1970s when we were both
working at the University of Michigan Computing Center (home
of MTS -- the Michigan Terminal System operating system). We
were both grad students, as were many of the workers there. I
remember reading a draft of his dissertation on program slicing
and thinking this is a really good idea.
Also, I think he was there when we had approximately
one terminal for every two programmers. The programmers negotiated
fluidly among ourselves for time at the terminal. But we also
constantly lobbied the administration that it would be cheaper
to put one on everyone's desk, due to the increased productivity.
Eventually they saw the light, and we all agreed that in the
future everyone would have access to a computer all the time.
Perhaps this stimulated Mark's further insight, to the concept
of ubiquitous computing.
Other comments can be found on pages devoted to Mark's ideas on ubiquitous computing, and