This document provides a summary of the August 2010 issue of Communications of the ACM, including the following key points: - The cover story is about the Singularity System, a microkernel-based operating system being developed by Microsoft Research. - Other featured articles discuss predicting the popularity of online content, network virtualization, and the 50th anniversary of the pioneering computer system PLATO. - Departments include letters, blogs, news, viewpoints, education matters, emerging markets, code vicious, interviews, and puzzles. - Reviews cover memory models in parallel programming and mechanism design in computer science. - The issue also includes contributed articles, research highlights, and information about ACM events

1. COMMUNICATIONS
ACM
cAcM.AcM.org OF THE 08/2010VoL.53No.08
The Singularity
System
Memory Models
Predicting the Popularity
Of Online Content
CTOs on Network
Virtualization
Has China Caught Up in IT?
Mechanism Design
Meets CS
Association for
Computing Machinery

3. ABCD springer.com
Springer eBooks
Supporting You in Your Research
7 9,800 eBooks in Computer Science
7 More published every day
7 Unrestricted printing and
downloading
7 Unlimited access to
reference works,
textbooks, monographs,
LNCS
VISIT
TODAY!
014726x

4. communications of the acm
Departments News Viewpoints
5 JACM Editor’s Letter 11 Mechanism Design 24 Economic and Business Dimensions
JACM at the Start of a New Decade Meets Computer Science Is the Internet a Maturing Market?
By Victor Vianu A field emerging from economics is If so, what does that imply?
teaming up with computer science By Christopher S. Yoo
6 In the Virtual Extension to improve auctions, supply chains,
and communication protocols. 27 Education
7 Letters To The Editor By Gary Anthes Preparing Computer Science
CS Expertise for Institutional Students for the Robotics Revolution
Review Boards 14 Looking Beyond Robotics will inspire dramatic
Stereoscopic 3D’s Revival changes in the CS curriculum.
8 BLOG@CACM Researchers working in vision and By David S. Touretzky
The War Against Spam; and More graphics are attempting to develop
Greg Linden asks if spammers have new techniques and technologies 30 Emerging Markets
been defeated; Michael Bernstein to overcome the current limitations Has China Caught Up in IT?
discusses Clay Shirky’s keynote in stereoscopic 3D. An assessment of the relative
speech at CSCW 2010; and By Kirk L. Kroeker achievements in IT infrastructure,
Erika S. Poole writes about how the firms, and innovation in China.
digital world can help parents cope 17 Making Sense of Real-Time Behavior By Ping Gao and Jiang Yu
with the death of a child. Data captured by sensors worn on
the human body and analyzed in 33 Kode Vicious
10 CACM Online near real-time could transform Presenting Your Project
Print is Not Just Ink Anymore our understanding of human The what, the how, and the why of
By David Roman behavior, health, and society. giving an effective presentation.
By Sarah Underwood By George V. Neville-Neil
37 Calendar
19 Celebrating the Legacy of PLATO 35 Privacy and Security
125 Careers The PLATO@50 Conference Remembrances of Things Pest
marked the semicentennial Recalling malware milestones.
of the computer system that was By Eugene H. Spafford
Last Byte the forerunner of today’s social
media and interactive education. 38 Viewpoint
128 Puzzled By Kirk L. Kroeker Rights for Autonomous
Figures on a Plane Artificial Agents?
By Peter Winkler 21 Gödel Prize and Other CS Awards The growing role of artificial
Sanjeev Arora, Joseph S.B. Mitchell, agents necessitates modifying
and other researchers are legal frameworks to better
recognized for their contributions address human interests.
to computer science. By Samir Chopra
By Jack Rosenberger
41 Interview
An Interview with Edsger W. Dijkstra
The computer science luminary,
in one of his last interviews
before his death in 2002, reflects
on a programmer’s life.
By Thomas J. Misa
Association for Computing Machinery
Advancing Computing as a Science Profession
2 communicaT io nS o f T he ac m | AU g U ST 201 0 | VO l . 5 3 | NO. 8

5. 08/2010 vol. 53 no. 08
Practice Contributed Articles Research Highlights
48 Software Development 104 Technical Perspective
with Code Maps Attacks Target Web Server Logic
Could ubiquitous hand-drawn And Prey on XCS Weaknesses
code map diagrams become By Helen Wang
a thing of the past?
By Robert DeLine, Gina Venolia, 105 The Emergence of Cross
and Kael Rowan Channel Scripting
By Hristo Bojinov, Elie Bursztein,
55 Moving to the Edge: and Dan Boneh
A CTO Roundtable on
Network Virtualization
Leading experts debate how 114 Technical Perspective
virtualization and clouds impact Large-Scale Sound and
network service architectures. Precise Program Analysis
By Mache Creeger By Fritz Henglein
63 Seven Principles for Selecting 72 The Singularity System 115 Reasoning About the Unknown
Software Packages Safe, modern programming in Static Analysis
Everything you always wanted to languages let Microsoft rethink By Isil Dillig, Thomas Dillig,
know but were afraid to ask about the architectural trade-offs in its and Alex Aiken
the decision-making process. experimental operating system.
By Jan Damsgaard and Jan Karlsbjerg By James Larus and Galen Hunt
Virtual Extension
Articles’ development led by 80 Predicting the Popularity
queue.acm.org of Online Content as with all magazines, page limitations often
Early patterns of Digg diggs and prevent the publication of articles that might
YouTube views reflect long-term otherwise be included in the print edition.
to ensure timely publication, aCM created
user interest.
Communications’ Virtual extension (Ve).
By Gabor Szabo and Ve articles undergo the same rigorous review
Bernardo A. Huberman process as those in the print edition and are
accepted for publication on their merit. these
articles are now available to aCM members in
Review Articles the digital library.
90 Memory Models: A Case for Intelligent Service Machine
Rethinking Parallel Languages Wei-Feng Tung and Soe-Tsyer Yuan
and Hardware
Solving the memory model problem Thinkflickrthink: A Case Study
will require an ambitious and cross- on Strategic Tagging
about the cover: disciplinary research direction. Eugenio Tisselli
the singularity system,
Microsoft research’s effort By Sarita V. Adve and Hans-J. Boehm
to build a microkernel- Plat_Forms: Is There One Best
based operating system, is
the focus of this month’s Web Development Technology?
cover story as told by the Lutz Prechelt
projects’ lead researchers
James larus and Galen
IllustratIon by st udIo to nne
Hunt. the system draws How a Service-Oriented Architecture
parallels to the singularity
model where physical laws May Change the Software
as we know them are no
longer valid. rendering this
Development Process
concept for the cover is Marc N. Haines and
Paul Farrington from studio tonne, an illustrative agency
based in brighton, england. For more on the studio’s work,
Marcus A. Rothenberger
see http://www.studiotonne.com.
AU g U ST 2 0 1 0 | VO l. 53 | N O. 8 | c om m u n ic aT ion S of T he acm 3

6. communications of the acm
trusted insights for computing’s leading professionals.
Communications of the ACM is the leading monthly print and online magazine for the computing and information technology fields.
Communications is recognized as the most trusted and knowledgeable source of industry information for today’s computing professional.
Communications brings its readership in-depth coverage of emerging areas of computer science, new trends in information technology,
and practical applications. Industry leaders use Communications as a platform to present and debate various technology implications,
public policies, engineering challenges, and market trends. the prestige and unmatched reputation that Communications of the ACM
enjoys today is built upon a 50-year commitment to high-quality editorial content and a steadfast dedication to advancing the arts,
sciences, and applications of information technology.
ACM, the world’s largest educational STA F F eDITOr IAl bOA rD
and scientific computing society, delivers
resources that advance computing as a DIRECTOR OF GROUP PU BLIS h I n G E DITOR- I n- C hIE F
science and profession. ACM provides the scott e. delman Moshe y. Vardi ACM Copyright notice
computing field’s premier Digital Library publisher@cacm.acm.org eic@cacm.acm.org Copyright © 2010 by association for
and serves its members and the computing Executive Editor nE W S Computing Machinery, Inc. (aCM).
profession with leading-edge publications, diane Crawford Co-chairs Permission to make digital or hard copies
conferences, and career resources. Managing Editor Marc najork and Prabhakar raghavan of part or all of this work for personal
thomas e. lambert Board Members or classroom use is granted without
Executive Director and CEO Senior Editor brian bershad; Hsiao-Wuen Hon; fee provided that copies are not made
John White andrew rosenbloom Mei Kobayashi; rajeev rastogi; or distributed for profit or commercial
Deputy Executive Director and COO Senior Editor/news Jeannette Wing advantage and that copies bear this
Patricia ryan Jack rosenberger notice and full citation on the first
Director, Office of Information Systems Web Editor VIE W P OIn TS page. Copyright for components of this
Wayne Graves david roman Co-chairs work owned by others than aCM must
Director, Office of Financial Services Editorial Assistant susanne e. Hambrusch; John leslie King; be honored. abstracting with credit is
russell Harris Zarina strakhan J strother Moore permitted. to copy otherwise, to republish,
Director, Office of Membership Rights and Permissions Board Members to post on servers, or to redistribute to
lillian Israel deborah Cotton P. anandan; William aspray; lists, requires prior specific permission
Director, Office of SIG Services stefan bechtold; Judith bishop; and/or fee. request permission to publish
donna Cappo Art Director stuart I. Feldman; Peter Freeman; from permissions@acm.org or fax
Director, Office of Publications andrij borys seymour Goodman; shane Greenstein; (212) 869-0481.
bernard rous Associate Art Director Mark Guzdial; richard Heeks;
Director, Office of Group Publishing alicia Kubista rachelle Hollander; richard ladner; For other copying of articles that carry a
scott delman Assistant Art Director susan landau; Carlos Jose Pereira de lucena; code at the bottom of the first or last page
Mia angelica balaquiot beng Chin ooi; loren terveen or screen display, copying is permitted
ACM CO U N C I l Production Manager provided that the per-copy fee indicated
President lynn d’addesio in the code is paid through the Copyright
Wendy Hall Director of Media Sales P R AC TIC E Clearance Center; www.copyright.com.
Vice-President Jennifer ruzicka Chair
alain Chesnais Marketing Communications Manager stephen bourne Subscriptions
Secretary/Treasurer brian Hebert Board Members an annual subscription cost is included
barbara ryder Public Relations Coordinator eric allman; Charles beeler; david J. brown; in aCM member dues of $99 ($40 of
Past President Virgina Gold bryan Cantrill; terry Coatta; Mark Compton; which is allocated to a subscription to
stuart I. Feldman Publications Assistant stuart Feldman; benjamin Fried; Communications); for students, cost
Chair, SGB Board emily eng Pat Hanrahan; Marshall Kirk McKusick; is included in $42 dues ($20 of which
alexander Wolf George neville-neil; theo schlossnagle; is allocated to a Communications
Co-Chairs, Publications Board Columnists subscription). a nonmember annual
Jim Waldo
ronald boisvert, Holly rushmeier alok aggarwal; Phillip G. armour; subscription is $100.
Martin Campbell-Kelly; the Practice section of the CaCM
Members-at-Large
Michael Cusumano; Peter J. denning; editorial board also serves as
Carlo Ghezzi; ACM Media Advertising Policy
shane Greenstein; Mark Guzdial; the editorial board of .
anthony Joseph; Communications of the ACM and other
Mathai Joseph; Peter Harsha; leah Hoffmann; aCM Media publications accept advertising
C O nTR IB U TE D A RTIC LES
Kelly lyons; Mari sako; Pamela samuelson; in both print and electronic formats. all
Co-chairs
bruce Maggs; Gene spafford; Cameron Wilson advertising in aCM Media publications is
al aho and Georg Gottlob
Mary lou soffa; Board Members at the discretion of aCM and is intended
Fei-yue Wang C O N TACT P O IN TS to provide financial support for the various
yannis bakos; Gilles brassard; alan bundy;
SGB Council Representatives Copyright permission activities and services for aCM members.
Peter buneman; Ghezzi Carlo;
Joseph a. Konstan; permissions@cacm.acm.org Current advertising rates can be found
andrew Chien; anja Feldmann;
robert a. Walker; Calendar items by visiting http://www.acm-media.org or
blake Ives; James larus; Igor Markov;
Jack davidson calendar@cacm.acm.org by contacting aCM Media sales at
Gail C. Murphy; shree nayar; lionel M. ni;
Change of address (212) 626-0654.
sriram rajamani; Jennifer rexford;
PU blI CATI O N S bOA r D acmcoa@cacm.acm.org
Marie-Christine rousset; avi rubin;
Co-Chairs Letters to the Editor Single Copies
abigail sellen; ron shamir; Marc snir;
ronald F. boisvert and Holly rushmeier letters@cacm.acm.org single copies of Communications of the
larry snyder; Veda storey;
Board Members Manuela Veloso; Michael Vitale; ACM are available for purchase. Please
Jack davidson; nikil dutt; Carol Hutchins; W e b S IT e contact acmhelp@acm.org.
Wolfgang Wahlster; andy Chi-Chih yao;
ee-Peng lim; Catherine McGeoch; http://cacm.acm.org
Willy Zwaenepoel
M. tamer ozsu; Vincent shen; COMMU N ICATION S OF THe ACM
Mary lou soffa; ricardo baeza-yates AU T H O r g U ID e l IN eS RES E A R C h hIGhLIGh TS (Issn 0001-0782) is published monthly
http://cacm.acm.org/guidelines Co-chairs by aCM Media, 2 Penn Plaza, suite 701,
ACM U.S. Public Policy Office david a. Patterson and stuart J. russell new york, ny 10121-0701. Periodicals
Cameron Wilson, director A DVe rT IS IN g Board Members postage paid at new york, ny 10001,
1828 l street, n.W., suite 800 Martin abadi; stuart K. Card; deborah estrin; and other mailing offices.
Washington, dC 20036 usa ACM ADVERTISInG D EPARTM E n T shafi Goldwasser; Monika Henzinger;
t (202) 659-9711; F (202) 667-1066 2 Penn Plaza, suite 701, new york, ny Maurice Herlihy; norm Jouppi; POSTMASTer
10121-0701 andrew b. Kahng; Gregory Morrisett; Please send address changes to
Computer Science Teachers Association t (212) 869-7440 Michael reiter; Mendel rosenblum; Communications of the ACM
Chris stephenson F (212) 869-0481 ronitt rubinfeld; david salesin; 2 Penn Plaza, suite 701
executive director
lawrence K. saul; Guy steele, Jr.; new york, ny 10121-0701 usa
2 Penn Plaza, suite 701 Director of Media Sales
Gerhard Weikum; alexander l. Wolf;
new york, ny 10121-0701 usa Jennifer ruzicka
Margaret H. Wright
t (800) 401-1799; F (541) 687-1840 jen.ruzicka@hq.acm.org
WEB
Association for Computing Machinery Media Kit acmmediasales@acm.org
Co-chairs
(ACM)
James landay and Greg linden
2 Penn Plaza, suite 701
Board Members SE
REC
Y
new york, ny 10121-0701 usa A
Gene Golovchinsky; Jason I. Hong;
E
CL
PL
t (212) 869-7440; F (212) 869-0481
E
Jeff Johnson; Wendy e. MacKay Printed in the u.s.a.
NE
TH
S
I
Z
I
M AGA
4 communicaTio nS o f The ac m | AU gU ST 201 0 | VO l . 5 3 | NO. 8

7. jacm editor’s letter
DOI:10.1145/1787234.1787235 Victor Vianu
JACM at the Start of a new Decade
It has been almost a year since I assumed the
editor-in-chief role for Journal of the ACM (JACM).
The move coincided with the start of a new decade.
For both reasons, it seems the right time to share
some thoughts on how ACM’s old- published in this period, and none were us say, hypothetically, that we aim to
est publication is doing and where it accepted in software engineering. Even publish annually the three best papers
might be headed. areas with very strong theoretical sides in 10 subfields of computer science. At
First published in January 1954, have minimal representation, includ- an average of 40 pages per paper, this
JACM plays a special role among ACM ing cryptography, logic in computer sci- would quickly consume the current an-
publications. Transactions publish ence, machine learning, and computer- nual page budget of 1,200 pages. This
high-quality research in a specific sub- aided verification. makes the notion of forgoing the print
field of computer science, aiming for There are several possible expla- edition in favor of an online-only publi-
comprehensive coverage of the area. nations for the difficulty in attracting cation increasingly tempting. However,
Communications of the ACM provides top-quality papers in some areas. Con- this remains a controversial idea.
magazine-style content appealing to ference publications are increasingly fa- JACM papers are currently published
all ACM members—academics and vored over journal publications in many in e-form in ACM’s Digital Library. Be-
practitioners—and includes select re- subfields. There also seems to be a (mis) sides advantages of cost and scalability,
search articles showcasing the “best of perception that some topics are not wel- the DL provides substantial added val-
the best” results originally presented come to JACM. One way to counteract ue in the form of cross-links and search-
in ACM proceedings. JACM fulfills a this is to ensure visible representation able metadata. Another advantage is the
complementary role by publishing in a of such areas on the editorial board. ability to post additional content on
single, highly selective venue, the best Indeed, several editors have now been home pages of articles, such as errata,
research in all areas of computer sci- appointed representing such areas as appendices, notes, even slides or videos
ence, allowing researchers to keep up bioinformatics, Web systems and algo- of talks. I believe such content is appeal-
with the state of the art across the entire rithms, software engineering, and com- ing to both readers and authors, and we
discipline. As such, JACM is the flagship putational economics. will aim to provide it on a regular ba-
scientific publication of the ACM. A proactive approach to ensuring sis. JACM’s new information director,
Overall, I believe JACM is going top-quality representation from a wid- Pierre Senellart, will play a central role
strong. It is a widely respected publi- er spectrum of areas, initiated by Prab- in shaping JACM’s online presence. A
cation; according to at least one bib- hakar Raghavan, consists of inviting a new Web site for JACM has also been
liometric authority (http://www.eigen- small number of papers selected from launched (http://jacm.acm.org/).
factor.org/map), it is the top-ranked top conferences in targeted subfields. One of the often-cited drawbacks of
journal in computer science. Yet JACM We currently have or are exploring journals versus conferences is the long
is facing nontrivial challenges. With the such arrangements with annual sym- wait from submission to publication.
field expanding and becoming increas- posiums, including STOC, FOCS, Prin- In the case of JACM, the publication
ingly diversified, its charter of publish- ciples of Database Systems (PODS), queue has hovered around three issues,
ing the best research across computer Principles of Distributed Computing or six months, for quite some time. This
science—as broadly construed—is a tall (PODC), Principles of Programming is reasonable, since it is considered
order. Much of JACM’s focus has been Languages (POPL), Logic in Computer risky to have much fewer papers in the
on theory of the flavor found at the Sym- Science (LICS), and conferences such publication pipeline. I am working with
posium on Theory of Computing (STOC) as Research in Computational Mo- the board to keep the processing time
and Foundations of Computer Sci- lecular Biology (RECOMB) and Com- under a year for most papers eventually
ence (FOCS). Past editors-in-chief have puter Aided Verification (CAV). This accepted, and shorter for papers that
worked to expand the scope of JACM be- approach is appealing because it no are eventually rejected.
yond this core. However, publications in longer leaves coverage of underrepre- In summary, JACM is facing some
some of the emerging or cross-boundary sented areas entirely up to the chance growing pains but it is thriving. I am
areas have been slow to follow. Of the 93 of spontaneous submissions. confident the journal will pursue its up-
articles published in JACM over the past An orthogonal obstacle to compre- ward trajectory in the coming years.
three years, approximately 35 are still in hensive coverage is simply due to the
Victor Vianu (vianu@cs.ucsd.edu) is a professor of
core algorithms and complexity. In con- proliferation of areas to be covered, computer science and engineering at university of
trast, only one bioinformatics paper and coupled with the physical limitations California, san diego.
one computer architecture paper were and cost of JACM as a print journal. Let © 2010 aCM 0001-0782/10/0800 $10.00
AU g U ST 2 0 1 0 | VO l. 53 | N O. 8 | c om m u n ic aT ion S of T he acm 5

8. in the virtual extension
DOI:10.1145/1787234.1787237
in the Virtual extension
Communications’ Virtual Extension brings more quality
articles to ACM members. These articles are now available
in the ACM Digital Library.
intelligent Service machine Plat_forms: is There one Best
DOI:10.1145/1787234.1787268 Web Development Technology?
Wei-Feng Tung and Soe-Tsyer Yuan DOI:10.1145/1787234.1787271
Machine is a metaphor that can be used to Lutz Prechelt
expand the capability of service systems and Plat_Forms is a contest in which
‘think’ for innovative service system design. three-person teams of professional
In this article, the notion of service machine programmers competed to implement
is defined as a socio-technical system with the same requirements for a Web-based
the shared reality of customer and provider system within two days, using different
aiming for the joint optimization of technology platforms. Three teams
productivity and satisfaction. An intelligent used Java EE, three used Perl, and three
service machine (ISM) moves beyond service used PHP. The resulting systems were
machine by modeling and automating thoroughly evaluated with respect to many
the cognitive process and knowledge criteria, such as completeness (reflecting
representations of the machine’s embodied productivity), maintainability, robustness
theory, enabling a systematic and (hinting at security), and size. This article
quantitative delivery of service operation reports on the setup of the contest and
using self services. The authors present some results of this study. Readers should
a machine-aware service-system design expect to see some prevalent prejudices
ACM’s framework (iDesign) and an ISM-supported confirmed and others firmly refuted.
service system to demonstrate the notion
interactions and the framework. how a Service-oriented
magazine explores architecture may change the
critical relationships Thinkflickrthink: a case Study Software Development Process
on Strategic Tagging DOI:10.1145/1787234.1787269
between experiences, people, DOI:10.1145/1787234.1787270
Marc N. Haines and
and technology, showcasing Eugenio Tisselli Marcus A. Rothenberger
emerging innovations and industry A tag can be created and disseminated The service-oriented approach to IT
for strategic purposes, including online architecture has become an important
leaders from around the world protest. The research presented in this alternative to traditional software
across important applications of article analyzes one particular protest development. Adding to this impetus are
strategy adopted by a number of users the efforts related to open standards and
design thinking and the broadening of Flickr: the use of anti-censorship tags open source products. But one key question
field of the interaction design. to make the protest visible within the remains: Are service-oriented architecture
site itself. The study of the dynamics (SOA) adopters ready for this change and
Our readers represent a growing of uncoordinated semantic strategies will they be able to provide a technical
community of practice that within dense online communities is of and organizational environment in which
enormous importance to gain a greater SOA-related technologies can be leveraged
is of increasing and vital understanding of how social and linguistic to their full potential? This article explores
interaction takes place in a networked the development process and methodology
global importance. environment, and how it can augment the that may require adjustments in order to
users’ potential for direct action. provide a good fit for SOA development.
Coming Next Month in COMMUNICATIONS
e
ib
cr
Medical Informatics: Point/Counterpoint on the
s
ub
Why So Slow? Future of Internet Architecture
/s
rg
.o
Research in Computing: Injecting Errors
cm
The Myth of Rapid Obsolescence for Fun and Profit
a
w.
w
Performance Evaluation and Thinking Clearly
w
://
tp
Model Checking Join Forces About Performance
ht
And the latest news on computational neurobiology, applying technology
to education, and MIT’s Big Wheel and smart bikes.
6 comm unicaTio nS o f T he ac m | AU gU ST 201 0 | VO l . 5 3 | NO. 8

9. letters to the editor
DOI:10.1145/1787234.1787236
cS expertise for institutional Review Boards
I
RBs n e e d com pu teR scien- Two such anecdotes involved research find some of them. Four to eight par-
tists, a point highlighted by on phishing, an intrinsically decep- ticipants should be included if the aim
the Viewpoint “Institutional tive phenomenon. Deception research, is to drive a useful iterative cycle: Find
Review Boards and Your Re- long used in social sciences, typically serious problems, correct them, find
search” by Simson L. Garfinkel takes longer to review because it runs more serious problems.
and Lorrie Faith Cranor (June 2010). counter to the ethical principle of Never expect a usability test to find
Not just over the nature of certain CS- “respect for persons” and its regula- all problems. CUE-studies1 show it
related research but because social sci- tory counterpart “voluntary informed is impossible or infeasible to find all
entists (and others) administer online consent.” Before developing a techni- problems in a Web site or product;
surveys, observe behavior in discus- cal solution to perceived IRB delays, the number is huge, likely in the thou-
sion forums and virtual worlds, and the typical causes of delay must be sands. This limitation has important
perform Facebook-related research. In established. Possibilities include inef- implications on the size of a test group.
this regard, the column was timely but ficient IRBs and uninformed and/or So go for a small number of partici-
also somewhat misleading. unresponsive researchers. Moreover, pants, using them to drive a useful iter-
First, the authors created a dichot- as with any deception research, some ative cycle where the low-hanging fruit
omy of computer scientists and IRBs, proposals may just be more ethically is picked/fixed in each cycle.
saying IRB “chairs from many institu- complex, requiring more deliberation. Finally, the number and quality of
tions have told us informally that they michael R. Scheessele, South bend, IN usability test moderators affects re-
are looking to computer scientists to sults more than the number of test
come up with a workable solution to the participants.
difficulty of applying the Common Rule authors’ Response: In addition, from a recent discus-
to computer science. It is also quite Scheessele is correct in saying an increasing sion with the authors, I now under-
clear that if we do not come up with a number of social scientists use computers in stand that the published research in the
solution, they will be forced to do so.” their research and is yet another reason IRBs article was carried out in 2004 or earlier
However, any institution conduct- should strive to include a computer scientist and the article was submitted for pub-
ing a significant amount of human- as a member. Sadly, our experience is that lication in 2006 and accepted in 2008.
subjects research involving computing most IRBs in the U.S. are understaffed, lack All references in the article are from
and IT ought to include a computer sci- sufficient representation of members with 2004 or earlier. The authors directed
entist on its IRB, per U.S. federal regu- CS knowledge, and lack visibility among CS my questions to the first author’s Ph.D.
lations (45 CFR 46.107(a)): “Each IRB researchers in their organizations. dissertation, which was not, however,
shall have at least five members, with Simson L. Garfinkel, Monterey, CA included in the article’s references and
varying backgrounds to promote com- Lorrie faith cranor, Pittsburgh, PA is apparently not available.
plete and adequate review of research Rolf molich, Stenløse, Denmark
activities commonly conducted by the
institution. The IRB shall be sufficient- how many Participants Reference
1. Molich, r. and dumas, J. Comparative usability
ly qualified through the experience and needed to Test usability? evaluation (Cue-4). Behaviour Information
expertise of its members…” No usability conference is complete Technology 27, 3 (May 2008), 263–281.
Though CS IRB members do not without at least one heated debate on
have all the answers in evaluating hu- participant-group size for usability correction
man-subject research involving com- testing. Though Wonil Hwang’s and “CS and Technology Leaders Honored”
puting and IT, they likely know where Gavriel Salvendy’s article “Number of (June 2010) mistakenly identified the
to look. It would also mitigate another People Required for Usability Evalu- American Academy of Arts and Sci-
problem explored in the column, that ation: The 10±2 Rule” (Virtual Exten- ences as the American Association for
“many computer scientists are unfamil- sion, May 2010) was timely, it did not the Advancement of Science. Also, it
iar with the IRB process” and “may be address several important issues con- should have listed Jon Michael Dunn,
reluctant to engage with their IRB.” In- cerning numbers of study participants: Indiana University, as one of the com-
deed, if an IRB member is just down the Most important, the size of a par- puter scientists newly elected as an
hall, computer scientists would likely ticipant group depends on the purpose American Academy 2010 Fellow. We
find it easier to approach their IRB. of the test. For example, two or three apologize for these errors.
Second, the authors assumed the participants should be included if the
length of the IRB review process rep- main goal is political, aiming to, say, Communications welcomes your opinion. to submit a
resents a problem with the process demonstrate to skeptical stakeholders letter to the editor, please limit your comments to 500
words or less and send to letters@cacm.acm.org.
itself though offered only anecdotal that their product has serious usabil-
evidence to support this assumption. ity problems and usability testing can © 2010 aCM 0001-0782/10/0800 $10.00
AU g U ST 2 0 1 0 | VO l. 53 | N O. 8 | c om m u n ic aT ion S of T he acm 7

10. The Communications Web site, http://cacm.acm.org,
features more than a dozen bloggers in the BLoG@cacm
community. in each issue of Communications, we’ll publish
selected posts or excerpts.
follow us on Twitter at http://twitter.com/blogcacm
DOI:10.1145/1787234.1787238 http://cacm.acm.org/blogs/blog-cacm
The War against Spam; point that even massive and complicat-
ed spam efforts like the Storm botnet
and more
generate surprisingly low revenues for
what appears to be the work required.
What is your own experience with
email spam? Do you think the spam
Greg Linden asks if spammers have been defeated; Michael Bernstein war been won? Or are these declara-
discusses Clay Shirky’s keynote speech at CSCW 2010; and tions of victory premature?
Erika S. Poole writes about how the digital world can help parents
cope with the death of a child. Reader’s comment:
I would argue that your declaration of
victory is premature, but not for the obvious
Greg Linden’s “has the trying to customize each email sent, reason that spam in the inbox has been
Spam War Been Won?” and many other tricks to evade detec- reduced. In that respect, spam in my inbox,
http://cacm.acm.org/ tion, but their increasingly complicat- and my spam folder as well, has gone way
blogs/blog-cacm/78121 ed efforts have not been able to outwit down in recent years, and that is welcome.
A decade ago, email the filters. However, two problems remain. First
spam was a dire problem. My own experience is that email and foremost, there is still the problem
Annoyances flooded most inboxes. spam has become a non-issue. Despite of false positives. I still have to check
Any attempt to read your email started prostituting my email addresses un- my spam folder because the filters will
with deleting the crud that had leaked disguised across the Internet, despite occasionally falsely flag a legitimate email.
through your defenses. receiving hundreds of spam messages Once I do find a desired email, I can flag it
Many predicted the problem would daily, nearly zero make it to my inbox. as non-spam to teach the filter and add the
only get worse. A few predicted that The ones that I do see typically are sender to a white list, but that is reactive. If
email would be dead in just a few years, borderline spam, from companies and you’re receiving hundreds of spam emails
the filters would be overwhelmed, the small businesses sending to a small a day, as you wrote, I imagine more than a
war would be lost, and email readers list rather than the mass splattering of few false positives slip by you.
would be buried under an avalanche true email spam. Additionally, as an entrepreneur, sending
of spam. Amazingly, the drop in response email from a new company like mine that
Today, email spam appears to be a rates from 2003 to 2008 may be close hasn’t established itself with the many
solved problem. A 2003 study put re- to making spam an unprofitable en- filters out there [is] very time consuming
sponse rates at 0.005%. A 2008 study terprise. There is a substantial amount and inefficient. If there was a proactive way
where the authors infiltrated a major of effort required to attack and man- for a legitimate sender to register itself and
spam botnet found response rates age a botnet of 1 million compro- either post a bond or pay e-postage, I think
had fallen to under 0.00001%, with mised machines that can cheaply send that would clean up a lot of email inboxes.
only 28 sales out of 350 million mes- 12 million messages per day. Huge I know e-postage proposals haven’t
sages sent. Spam filters appear to have email campaigns that attempt to work gotten very far in the past, but if the spam
forced down response rates three or- around spam filters require sophistica- response rate is now down to 0.00001%
ders of magnitude in five years. Spam- tion to devise and run. Email address then the postage can be a lot lower as well.
mers have fought back with misspell- lists have to be purchased and main- The second reason we can’t declare
ings, adding additional text to emails, tained. It appears to be getting to the victory just yet is the very fact that the
8 communicaTio nS o f T he acm | AU g U ST 201 0 | VO l . 5 3 | NO. 8

11. blog@cacm
spammers and their resources, both botnets um, an event in which new scholars dis-
and humans, remain alive and well. If email cuss their work with a panel of experts.
spam continues to become less and less “usability is an In addition to being a great opportunity
profitable, then they will simply send spam important refinement for students, the Doctoral Colloquium
in other forms such as on Twitter, Facebook, highlights some of the most exciting
etc. Individual computers continue to technique when you work in the field from promising young
get infected and people still foolishly have a good idea, scholars. In particular, I couldn’t help
click on requests from Nigerian princes. but notice that the students invited to
Unfortunately, we have to continue to apply but it is a horrible this year’s event presented work high-
technological fixes to our networks and determiner of utility lighting the deeply human side of infor-
teach people not to be so gullible. mation technology.
Believe me, I wish we could declare on a grander scale.” For example, imagine you’re a par-
victory, but we’re not there yet. ent who has suffered the unthinkable:
—Will Hartmann Your child has died. How do you cope
with such a traumatic, painful, and
michael Bernstein’s disorienting experience? For some
“clay Shirky: Doing parents, information technologies can
work, or Doing Work?” tivated to do it in the first place?” Excel play an important role in the grieving
http://cacm.acm.org/ needs usability testing because people and mourning process. Yet how are be-
blogs/blog-cacm/72609 are forced to use it for Work; technology reaved parents using technologies to
MSN usability research- for work instead needs to understand grieve and mourn? If we were to design
ers were stumped. Their usability lab users’ underlying motivations. technologies that help people cope
had tested just about every aspect of Extrapolating on my own here: with grieving and loss in meaningful
its MSN portal and had been pleased to Usability is an important refinement and respectful ways, what would they
find that it consistently scored higher technique when you have a good idea, look like?
than its competitors. Yet a user base but it is a horrible determiner of utility I had the opportunity to speak with
didn’t flock to MSN—the portal simply on a grander scale. (Sure, pay me $10 Mike Massimi, a Ph.D. candidate at
could not attract and retain as many for a lab study and I’ll use anything University of Toronto who’s examin-
users as it wanted. Then Clay Shirky for an hour!) Usability is a local hill- ing these questions for his thesis work.
relayed the million-dollar question: climbing algorithm. We need tech- To understand how technology plays a
Were these tasks that users actually niques to make and evaluate that mi- role in modern grieving, Mike is work-
wanted to do? Or were these highly us- raculous motivational leap, whether ing extensively with two community
able aspects of the site going to remain it’s derived from the design process or organizations in the Ontario area that
unused because nobody wanted to use social science. Develop that and you provide social support to parents who
them? There was a gaping hole be- could save thousands of man-hours have suffered the loss of a child. His
tween usability and usefulness. developing tools that nobody will ever next step is to create meaningful, ap-
In a keynote delivered to this year’s want to use. propriate, and respectful technologies
ACM Conference on Computer Sup- that help bereaved parents mourn and
ported Cooperative Work (CSCW), Reader’s comment: remember their lost children.
author and academic Clay Shirky cap- Jared Spool, for years, talked about Death and dying are experiences as
tured this question in a distinction be- compelled shopping tasks in which people old as humanity, and have been stud-
tween Work and work. Work (Capital W) were actually given money to buy things ied by scholars in other disciplines for
is what we have considered for years; on Internet sites, and could not. He really centuries. Yet technology researchers
your boss tells you to do something, wanted to solve the usability problem, and designers are just now starting
you do it, and you get paid. By contrast, but also realized that decoupling the to come to grips with how to design
work (little w) is motivated by inherent motivational issues with usability is difficult. with end-of-life experiences in mind.
interest and is generally unpaid. Think Little ‘w’ work vs. Big ‘W’ Work suggests If you’re interested in learning more
of the difference between an Encyclope- that we are going to have to dig much about this topic, Mike co-hosted the
dia Britannica editor doing Work and deeper into this issue than we had before. first workshop (ever!) focused on death
a Wikipedia editor doing work during —Ed H. Chi and the digital world at the ACM CHI
spare hours. Big Work drives the econ- 2010 conference. You can see more
erika S. Poole’s
bernsteIn PH otoGra PH by Jason d orF M an
omy; little work drives the Internet. Big info about the workshop at http://www.
Work builds skyscrapers; little work “Death and the dgp.toronto.edu/~mikem/hcieol/.
generates a half-million fan fiction sto- Digital World”
ries about Harry Potter. http://cacm.acm.org/ Greg Linden is the founder of Geeky Ventures. Michael
Clay argued that user-testing tech- blogs/blog-cacm/72837 Bernstein is a Ph.d. student in the Computer science and
artificial Intelligence lab at Massachusetts Institute of
niques developed over the past 25 years At the Computer Support- technology. Erika Shehan Poole is an assistant professor
for Work no longer apply for work. We ed Cooperative Work 2010 conference, at the school of Information sciences and technology at
Penn state university.
shouldn’t be asking, “Can you com- 13 Ph.D. students received invitations
plete the task?” but rather “Are you mo- to participate in the Doctoral Colloqui- © 2010 aCM 0001-0782/10/0800 $10.00
AU g U ST 2 0 1 0 | VO l. 53 | N O. 8 | c om m u n ic aT ion S of T he acm 9

12. cacm online
ACM
Member
News
DOI:10.1145/1787234.1787239 David Roman WiLLiam J. DaLLy WinS
eckeRT-mauchLy aWaRD
Print is not Just ink anymore Acm and Ieee
computer
society jointly
presented the
eckert-mauchly
The world of Communications is not contained in the pages of a monthly maga- Award to
William J. dally,
zine. Like other publications, Communications has expanded over time into a va- chief scientist and senior vice
riety of electronic formats for e-connected members and readers. Each format de- president of research of nVIdIA
livers something its counterparts do not. Digital Editions (http://mags.acm.org/ corp., for his innovative
communications) present complete issues with familiar, flipable pages, but on contributions to the architecture
of interconnection networks and
full-screen and mobile systems. The Web site (http://cacm.acm.org) moves maga- parallel computers. dally, who is
zine content into HTML, also the Willard R. and Inez Kerr
and adds other articles, Bell professor of computer
science and electrical
daily news, blogs, plus ac-
engineering at stanford
cess to ACM’s abundant university, developed the system
member services. Articles and network architecture,
from Communications’ signaling, routing, and
synchronization technology that
Virtual Extension (VE) are is found in most of today’s large
available from the Web parallel computers.
site and ACM’s Digital Li- dally discussed his current
brary (http://acm.org/dl); research in an email interview,
saying, “At present I’m working
the print edition publishes on low-power computer
only their summaries. Digi- architecture. It’s interesting to
tal Editions, introduced in find out where the power actually
goes in a modern computer, and
January 2008, have cleared it’s very exciting that there are
the way for mobile apps opportunities where innovation
Communications is going mobile.
and a mobile Web site, now can make a large difference
in development, that will tailor content to handhelds. The goal of each format is in efficiency. It’s particularly
rewarding that this work is likely
to give users the content they want, where, when, and how they want it. to have a measurable positive
Communications’ brand began taking e-steps before the relaunch of the Com- impact on the environment.
munications Web site in April 2009. The concept of the VE, in fact, was introduced “I really enjoy seeing how
parallel computing enables
in 1996, first as a biannual collection of articles available only in e-format—a pio-
new applications that weren’t
neering step in publishing circles back then. Originally conceived as an outlet for possible before. some of the
articles that did not fit into page-constraints of the print edition, the VE is coming most exciting of these involve
into its own, having evolved as a monthly editorial fixture since September 2008. better human-computer
interfaces and interactions
Like Communications’ other formats, it will continue to evolve, and may become with the physical world. A great
a component of a digital-first publishing strategy. The VE’s status is evidenced by example is augmented reality—
the readership of its most popular articles listed here, which is on par with and in where parallel computing
some cases exceeds that of print issue cover stories. The VE is establishing itself enables realtime computer
vision to interpret the image
as a destination for authors and readers. you see, query a database, and
annotate the image with useful
information.”
article Bit.ly uRL
Asked about important
Principles for effective Virtual Teamwork http://bit.ly/dCSgY0 cs issues, dally singled out
“cs education, particularly
Capstone Programming Courses Considered Harmful http://bit.ly/bodbNO education about parallelism.
Number of People required for Usability evaluation: 10±2 rule http://bit.ly/9JHOlh We aren’t producing enough
cs graduates, and the ones we
PHotoGra PH by a ndrIJ bo rys
An Overview of IT Service Management http://bit.ly/cJlye7 do produce don’t understand
Why Did Your Project Fail? http://bit.ly/b0oaKt parallel programming. teaching
parallelism isn’t just an add-on
The requisite Variety of Skills for IT Professionals http://bit.ly/cyNF8g to the existing curriculum; every
course needs to be redesigned
A Holistic Framework for Knowledge Discovery and Management http://bit.ly/amhrsp
around parallelism.”
—Jack Rosenberger
10 communicaT io nS o f T h e acm | AU gU ST 201 0 | VO l . 5 3 | NO. 8

13. n
news
Science | DOI:10.1145/1787234.1787240 Gary Anthes
mechanism Design
meets computer Science
A field emerging from economics is teaming up with computer science
to improve auctions, supply chains, and communication protocols.
T
h e Bo u n d ARIe s se pARAtIn g
computer science and other
disciplines are blurring at an
accelerating pace. As work in
computers, biology, the phys-
ical and social sciences, and econom-
ics becomes more complex, so does
the motivation for practitioners to seek
help from each other.
Mechanism design, which emerged
from economic game theory in the
1970s, is now shaking hands with in-
formation technology. Built on a formal
mathematical base, mechanism design
expresses ideas that are elegantly sim-
ple, yet tricky to apply in the real world:
people in competition will act “rational-
ly” to meet their own selfish goals; they
have private information, and may act
in ways that can’t be observed; and they in the DaRPa network challenge, teams used mechanism design and social networking tech-
may lie. The central goal in mechanism niques to locate the defense agency’s 10 geographically dispersed, red weather balloons.
design is to devise a system by which
those people will tend to act in ways that nology, from network design to distrib- all 10 balloons.
benefit the owner of the system, or soci- uted computing to operating systems. Success would depend on a competi-
ety at large. The DARPA Network Challenge tor’s ability to assemble a large group
Information technologists are turn- from the Defense Advanced Research of geographically dispersed volunteers.
PHotoGra PH Court esy oF da rPa
ing these concepts into such disparate Projects Agency (DARPA) offers one ex- “We wanted to understand how you
applications as auction management, ample, with a social networking twist. could rapidly mobilize a very large team
supply chain optimization, and the Last December, DARPA tethered 10 red to solve a hard problem, and how to do
matching of organ donors and recipi- weather balloons at undisclosed loca- that in an adversarial environment,”
ents. Meanwhile, mechanism design is tions across the continental U.S. DAR- says Peter Lee, a DARPA director. Also,
enabling advances in information tech- PA’s challenge: Be the first team to find DARPA wanted to learn how teams
AU g U ST 2 0 1 0 | VO l. 53 | N O. 8 | c om m u n ic aT ion S of T he acm 11