Doug Terry is a longtime admirer of the conferences delivered by the Association of Computing Machinery’s Special Interest Group on Operating Systems (SIGOPS), the leading professional organization for researchers focusing on operating systems and distributed computing.
His colleagues at Microsoft Research are doing their best to intensify that admiration.

Terry, a principal researcher at Microsoft Research Silicon Valley, was elected the new chair of SIGOPS for the two-year term beginning July 1. And, as somebody who has been a SIGOPS participant for 25 years, that role provides genuine satisfaction.
“I value the high quality conferences that SIGOPS supports,” he says, “the strong sense of community among its members, the forum it provides for exchanging and vetting new ideas, and its technical focus on practical problems that positively impact the computing world.”
Terry, then, must be thrilled with the 21st Symposium on Operating Systems Principles (SOSP), scheduled for Oct. 14-17 in the Columbia River community of Stevenson, Wash.
Microsoft Research has contributed four of the 25 papers accepted for SOSP, and two of the 10 SOSP sessions will be chaired by Microsoft Research personnel. In addition, two of the talks to be presented during the Works-in-Progress Session will be delivered or co-delivered by researchers from the organization’s Silicon Valley and Redmond labs.
Such a contribution signifies the importance that Microsoft Research—and Terry—place on SIGOPS and its associated events.
“As chair of SIGOPS,” Terry says, “I aim to provide an important service to the research community, namely, maintaining a healthy and vibrant organization. SIGOPS is a highly respected brand that ensures researchers of a premier venue to which they can submit papers.”
Terry was asked to run for SIGOPS chair by the previous chair, Keith Marzullo, professor and chair of the Department of Computer Science and Engineering at the University of California, San Diego. Terry prevailed in elections held in May and June, and he entered office with some specific goals in place.
“As SIGOPS chair,” he says, “one role is to initiate new conferences and journals that publish the latest state-of-the-art technology and best practices from both the research community and industry. Another goal of mine is to expand the community’s outreach into more countries, since important research is being done all over the world.
“Third, SIGOPS is an organization with a long, distinguished history of pioneering work that has shaped the computing industry, and yet little has been done to capture the history in a comprehensive way. As the luminaries in our field age, I believe that it is important for SIGOPS to help put into perspective the fundamental contributions of its early members and mine their collective wisdom while the opportunity still exists.”
In his third decade of association with SIGOPS, Terry has seen the very definition of the organization’s focus evolve in distinct, exciting directions.
“Although SIGOPS originally started with an emphasis on traditional ‘operating systems,’ “ he says, “this focus has expanded over the years to encompass the broad spectrum of systems software. Today, a computer operating system is not just the software kernel running on a single PC. Virtually all systems today consist of components that interact over a network and, hence, are designed as large, distributed operating systems.
“Moreover, new technology trends such as powerful mobile devices and scalable, data-center-based services indicate an increased emphasis on practical new forms of ‘operating systems.’ I am excited to take on the leadership of an organization that plays such a central role in advancing computing technology throughout the world.”
A list of the names of the sessions scheduled for SOSP 2007 provides a glimpse into the breadth of subjects currently under discussion in the field of operating systems:
Mike Schroeder, assistant managing director of Microsoft Research Silicon Valley, will serve as session chair for the Concurrency session, and Rich Draves, research area manager who heads the Systems and Networking group at Microsoft Research Redmond, will serve as session chair for the Works-in-Progress session.
The conference gets under way in earnest on the morning of Oct. 15, and the first session, Web Meets Operating Systems, features three papers, two of them from Microsoft Research.
One, called Protection and Communication Abstractions for Web Browsers in MashupOS, was co-authored by Helen J. Wang, Xiaofeng Fan, and Jon Howell, of Microsoft Research Redmond, along with Collin Jackson of Stanford University.
The other, entitled AjaxScope: A Platform for Remotely Monitoring the Client-Side Behavior of Web 2.0 Applications, was written by Emre Kiciman and Benjamin Livshits of Microsoft Research Redmond.
Microsoft Research also helps inaugurate the second day of the conference, with the paper Bouncer: Securing Software by Blocking Bad Input—written by Manuel Costa and Miguel Castro of Microsoft Research Cambridge, Lidong Zhou and Lintao Zhang of Microsoft Research Silicon Valley, and Marcus Peinado of Microsoft—serving as one of three to be delivered during the Software Robustness session.
On the final morning of SOSP 2007, the Storage session features a paper called Improving File System Reliability with I/O Shepherding, for which Vijayan Prabhakaran of Microsoft Research Silicon Valley is a co-author, along with Haryadi S. Gunawi, Swetha Krishnan, Andrea C. Arpaci-Dusseau, and Remzi H. Arpaci-Dusseau of the University of Wisconsin-Madison.
During the Works-in-Progress Session, to be held on the afternoon of Oct. 15, Dahlia Malkhi and Venugopalan (Rama) Ramasubramanian of Microsoft Research Silicon Valley will discuss Prediction Trees to Support Network-Aware Applications. A second discussion, entitled A Social Networking-Based Access Control Scheme for Personal Content, will be delivered by Kiran K. Gollu and Stefan Sarolu of the University of Toronto and Alec Wolman of Microsoft Research’s Redmond lab.
Microsoft Research’s substantial SOSP 2007 presence underscores the significance of such scientific endeavors to Microsoft—and to the greater computer-science community.
“The research performed by SIGOPS members,” Terry says, “explores techniques for building reliable, scalable, secure, and correctly functioning computer systems and, hence, is of critical importance to the computing industry as a whole and Microsoft in particular.”
|