\ NATIONAL SCIENCE FOUNDATION WASHINGTON. D.C. 20550 I

June 16, 1981

Professor Edward A. feigenbaum Comouter Science Department Stanford University Stanford, C. 94305

Dear Colleagues:

Enclosed are the minutes of the May 28 and 29, 1981 meeting of the Advisory Subcommittee on Computer Science of the National Science Foundation. There are a number of significant items addressed at the meeting. These were:

1. New Investigator Program 2. Postdoctoral Program 3. Coordinated Experimental Research 4. Computer Network 5. Recommendations to the NSF on the Cryptology Issue (A supplement to the enclosed report will be issued on this subject in a few weeks.) 6. Recommendation to the NSF to promote the Computer Science Section to a Division. 7. Recommendation to the NSF re Science Education and the Behavioral and Social Sciences 8. Tribute to Dr. John R. Pasta 9. Election of Dr. Jack Minker as Chairman of the Subcommittee for FY 82.

I hope that you will read the minutes carefully and send any comments you may have concerning any of these issues to me or to Kent Curtis at the National Science Foundation.

Sincerely yours,

RECEIVED Jack Minker Advisory Subcommittee JUN 2 3 1981 for Computer Science National Science Foundation Department of Computer Science E. A. FEIGENBAUM University of Maryland College Park, Maryland 20742

Attachment

U^-^^~J

Chairman, DRAH" SUMMARY MINUTES

ADVISORY SUBCOMMITTEE FOR COMPUTER SCIENCE

May 28 and 29, 1981

NATIONAL SCIENCE FOUNDATION 1800 G. Street N. W. Washington, D. C. 20550

I Attendance:

Advisory Committee:

Membership list attached

Subcommittee members present:

Dr. Alfred V. Aho, Bell Telephone Laboratories, Murray Hill, NJ Dr. Daniel E. Atkins, 111, University of Michigan, Ann Arbor, MI Dr. Woodrow W. Bledsoe, University of Texas at Austin, Austin, TX Dr. Taylor L. Booth, University of Connecticut, Storrs, CT Dr. Michael J. Fischer, University of Washington, Seattle, WA Dr. Susan L. Gerhart, Univ. of So. California, Marina del Ray, CA Dr. John V. Guttag, MIT, Cambridge, MA Dr. Jack Minker, University of Maryland, College Park, MD Dr. Mary M. Shaw, Carnegie-Mellon University, Pittsburgh, PA

Ad Hoc Members

None

Subcommittee members absent:

Dr. Susan H. Nycum Dr. Robert W. Taylor Dr. Joseph F. Traub

Advisory Subcommittee for Mathematics:*

Dr. James G. Glimm, Rockefeller University, New York, NY Dr. Daniel Gorenstein, Rutgers University, New Brunswick, NJ Dr. Jerrold E. Marsden, University of California, Berkeley, CA Dr. Hugh L. Montgomery, University of Michigan, Ann Arbor, MI Dr. Yiannis N. Moschovakis , University of California, Los Angeles, CA Dr. Ronald Pyke, University of Washington, Seattle, WA Dr. Martha K. Smith, University of Texas, Austin, TX Dr. Frank L. Spitzer, Cornell University, Ithaca, NY Dr. William A. Veech, Rice University, Houston, TX Dr. Grace G. Wahba, University of Wisconsin, Madison, WI

*A.M. Session May 29, 1981 NSF Staff:

Dr. Donald Langenberg, 0/D Dr. William Klemperer, MPS Dr. Bernard Chern, DAR Dr. William Rosen, MCS Dr. Al Thaler, MCS Dr. Al Schutzman, ECSE Dr. Edward Weiss, IS Dr. Richard Nicholson, CHEM Ruth OGC Donald W. Frenzen, OGC Jeanne E. Hudson, 0/D Computer Science Section Staff

Public:

Professor Merrell Patrick, Duke University Professor Henry Fuchs, University of North Carolina David Dickson, Nature Magazine Ken Lees on, NTIA Jerry Sturges , NTIA John Walsh, Science Magazine Lee Hatch, NSA

II Agenda :

The agenda given in the public notice was followed.

11l Narrative: May 28, 1981 - 9:00 A.M. to 1:00 P.M. A. NSF Research Support, Policies and Budgets - Dr. William Klemperer, Assistant Director for Mathematical and Physical Sciences

(1) An effective administrative structure has been set up in the Computer Science Section.

(2) Effective relations have been established with ARPA.

(3) The mathematical, computer and physical sciences are not threatened hy the new administration attitudes or actions. The Mathematics Directorate budget for 1982 is the same as in the budget proposed by the previous administration. Although the 1983 budget will not be one of growth it should not be a dismal budget.

2

Greenstein, I

(4) The NSF is probably the most effective organization in presenting and coordinating needs to Congress. The NSF must, however, abide by administration policies in making such representation to Congress.

(5) No change is planned to separate mathematics and computer science and to place them in separate divisions.

(6) Additional shifts in funds within the Directorate to meet new initiatives in computer science (e.g. for VLSI) are not envisioned. Such initiatives will require a shift of funds within the Computer Science Section. B. Coordinated Experimental Research - Dr. W. Richards Adrion The Coordinated Experimental Research program in NSF includes Research Facilities for Experimental Computer Science, the Computer Science Network and the New Faculty Investigator awards.

Dr. Adrion reviewed the status of the research facilities effort.

1. Statistics

Average FY 80 No. Amt. Annual Rate

Proposals 7 $24. 8M $800K Awards 1 (Wash. U.) $4.1M $825K

Average FY 81 No. Amt. Annual Rate

Proposals 24 $104. 9M $875K Awards (Rec.) 4 $14M $700K/awar FY 81 - $3.6 M of special projects budget identified a 'experimental computer science".

2. Four proposals were recommended for funding to the National Science Board. Three proposals were accepted by the NSB and further review of the fourth proposal was requested by the NSB.

3. DARPA plans to fund a fifth effort

4. DARPA is planning a computer initiatives effort. The current status is that no new budget has been received for this effort through 1982.

3 ' -

5 Plans for FY 82 CER Budget - $9.06MTotal For research facilities awards:

Obligations from awards made in FY 80 and FY 81

New Awards

Total

6 A major objective of the effort is to improve experimental research capabilities and the computer environment at 10-15 institutions (total) over the next 8 years. Computer Science Research Network (CSNET) - Dr. C. William Kern

Dr. Kern presented the status of the CSNET effort.

1 The National Science board approved the plan to have NSF manage the effort for a maximum of two years with the proviso that NSF terminates their management at that time.

2 Salient features of CSNET

Access: Phone Net, Public Nets,, ARPANET Length of Project: 5 years Open to all - 100 institutions or research groups expected Multiplicity of services - file transfer, message service, remote resource sharing, interprocess communication

3 A management plan listing dates and milestones has been developed.

4 A CSNET Project Redefinition has taken place which has caused the following:

(a) The CSNET Gateway development has been eliminated as DARPA will develop the VAN Gateway.

(b) VAX/UNIX is the primary target for the CSNET development.

(c) The VAX protocol development will be concentrated at Purdue.

5 The current CSNET architecture is shown in Figure 1.

4 FIGURE 1. a CSNET ARCHITECTURE

'////////, ARPANET CONNECTION FOR DOD AUTHORIZED TRAFFIC DEVELOPMENT USE CONNECTION MCS 81-2717 6-11-81 D. New Faculty Investigator's Program and Postdoctoral Program Dr. Bruce H. Barnes

Dr. Barnes presented the status of efforts in this area

1. One grant was made under the Postdoctoral Program. A total of three applications were received. Because the number of Postdoctoral candidates coming out are small and there is too little interest to justify the effort and expense of a separate program the Postdoctoral Program has been cancelled. The regular research grants should handle this program.

2. New Faculty Investigators Program

A total of 49 applications were received from 33 universities. Of these 11 received high evaluations and 4 were selected to receive grants.

3. The National Science Board granted approval to continue the program. Because the proposed awards were relatively large, the NSF Board questioned them but approved continuing the program in response to the special needs of computer science. The number of awards were small because the Computer Science Section wanted to assure the Board that they were supporting extraordinary scientists. E. Computer Engineering Programs - Dr. Bernard Chern Dr. Bernard Chern discussed the Computer Engineering Program within the Engineering Directorate. There is a close relationship between efforts in the Computer Science Section and the Computer Engineering Program. Some projects are jointly funded between the two activities.

A good cooperative spirit appears to exist between the two organizations, however, some closer coordination could be beneficial. F. Information System Science Program - Dr. Edward Weiss Dr. Weiss discussed the Information System Science Program. The activities here tend to complement the efforts in the Computer Science Section. Some grants are jointly funded between the two organizations.

G. Oversight Reviews

Because of funding problems within the section, the oversight reviews for Computer Science Research Equipment and Intelligent Systems were not conducted.

5 «

May 28, 1981 - 1:30 P.M. to 5:00 P.M. Because of the cancellation of the oversight reviews, it was decided that part of the business of the committee planned for the second day would be conducted.

A. Chairperson's Report

Dr. Jack Minker reported on activities that he had been involved in since the last meeting of the Subcommittee.

1. Letter to the NSF Board in Support of the NSF CSNET Management Plan.

Dr. Minker wrote a letter to the NSF Board in support of the NSF CSNET Management Plan. As reported in the minutes of the meeting during the A.M., above, the NSF Board approved the plan.

2. Meeting of Chairperson's of NSF Advisory Committees.

Dr. John B. Slaughter, Director of the NSF called a meeting of all NSF Chairperson's of Advisory Committees on Saturday, April 4, 1981.

The major item discussed was the budget. Dr. Slaughter noted that there was a cut in the NSF Budget. Two particular areas affected by the cut proposed by the Office of Manpower and Budget (0MB) were:

a. Science education, and b. Behavioral and Social Sciences.

Dr. Slaughter believes that as part of the Administration, the NSF could not argue for increasing the budget to that proposed by the previous administration. However, he believed that the NSF should, with consultation with the OMB and Congress determine how its budget should be allocated.

The NSF Board wrote to the Administration on this matter and the Administration approved the statement.

6 .

3. Ad Hoc Meeting with Dr. William Klemperer

A meeting was called by Dr. William Klemperer to discuss the future of the Mathematical and Computer Science Division within the Mathematical and Physical Science Directorate in view of the retirement of Dr. John R. Pasta. Attendees at the meeting of May 6, 1981 were:

Dr. William Klemperer, Assistant Director, MPS Dr. Peter Lax, Courant Institute and NSF Board Dr. Saunders MacLane, University of Chicago Dr. Jack Minker, University of Maryland, Chairman of Computer Science Advisory Committee Dr. Ronald Pyke, University of Washington, Chairman of Math Advisory Committee Dr. William Rosen, Mathematics Section Mr. Kent Computer Science Section

The pros and cons of separating the Mathematical and Computer Science Division were discussed. The advice given by Drs. Lax, Maclane, Minker and Pyke was that mathematics and computer science should be made into separate divisions. It was noted that it would be extremely difficult to obtain a replacement for Dr. Pasta who would be acceptable to both the mathematics and computer science communities. Names of individuals who might be appropriate were explored.

B Direction of Computer Science Within the NSF

The Subcommittee discussed various possible reorganizations:

(1) remain together with mathematics;

(2) become a separate division within the MPS directorate;

(3) merge all computer science related activities into one directorate (Computer Science Section, Computer Engineering, and Information Science and Technology;

(4) merge the Computer Science Section and the Computer Engineering Program into one activity within the Engineering Directorate.

7

Curtis, *

The following motion was passed unanimously to address the immediate problem presented by the retirement of Dr. John R. Pasta:

We recommend that Computer Science become a separate Division within the Directorate of the Mathematical and Physical Sciences . In addition some members of the Committee felt the Foundation should consider merging all programs that support computer activities. The Chairperson was charged with drafting a letter to be transmitted to Dr. Donald Langenberg, Associate Director of the National Science Foundation transmitting the message and the sense of the discussion.

Attached is a letter transmitted to Dr. Langenberg. The letter was approved by the Subcommittee at the May 29, 1981 continuation of the meeting. . May 29, 1981 - 9:00 A.M. to 12:45 P.M. The first joint meeting of the Advisory Subcommittee on Computer Science and the Advisory Subcommittee on Mathematics was convened. Dr. Jack Minker was asked to chair the joint meeting.

A Tribute to Dr. John Pasta

A motion was passed unanimously to send a tribute to Dr. John R. Pasta who retired during the month of May as Director of the Division of Mathematics and Computer Science. The tribute was received by him five days before he passed away. A copy of the tribute and an obituary is attached.

B Guttag Committee Draft Report on the Role of the NSF in Supporting Cryptological Research.

A three and one half-hour discussion was held on the Guttag Committee Draft Report. In these summary minutes we briefly review the outcome of the discussion. A supplement to these minutes will be issued providing the details.

8 * "

Dr. Minker reported that the Guttag Draft Report was disseminated to:

(1) NSF personnel

(2) Advisory Subcommiittees on Computer Science and Mathematics

(3) Department Heads of Ph.D. granting computer science departments and laboratories

(4) Commerce Department

(5) National Security Agency

Dr. John Cherniavsky summarized written comments received from: Dr. Peter Denning, Dr. John Hamblen, Dr. Gerald J. Popek, and Dr. Paul Purdom concerning the draft report. He also summarized a new policy concerning work in cryptology submitted to NSF by the Massachusetts Institute of Technology.

Visitors who had read the report were given an opportunity to make comments. No visitors availed themselves of this opportunity.

The report was discussed as a whole, followed by a vote on each specific recommendation. Modifications were made to tighten the recommendations. The report, as modified, was approved unanimously by the joint committee. The sense of the report is that of the draft report.

The following two motions were passed unanimously

(1) Dr. John Guttag will prepare a final version of the report, to be approved by Mr. Kent Curtis, Dr. Jack Minker, Dr. Ronald Pyke, and Dr. William Rosen. Dr. Minker is responsible for transmitting the report to Dr. Langenberg.

( 2) The Computer Science and Mathematical Subcommittees express their appreciation to Dr. John Guttag and his Committee for preparing the Draft Report.

A supplement to this report will contain the votes on each motion, the entire report as approved, and the letter of transmittal to Dr. Langenberg. Dr. Minker and Dr. Pyke are responsible for seeing that the report receives as wide a dissemination as the ACE report.

9 May 29, 1981 1:30 - 4:00 P.M. The meeting of the Subcommittee continued

A. Discussion of Impact of Coordinated Experimental Research Program and CSNET on Funding Conventional proposals

Dr. James Ortega, Chairman of the Department of Computer Science at the University of Virginia requested a review of the new initiatives in the CER and CSNET as it impacts conventional grants. The following funding expenditures are relevant :

(% change) (% change) (% change)

CER GRANTS $820K 3805 (364%) 6900 (81%)

CSNET $145K 1200 (728%) 1200 (0%)

Young Investigator 0 335 960 (187%) and Postdoctoral Fellowships

Science Research $17490K (4.1%) 16250 (-7.1%) 17940 (10.4%) Project Support

CSS Total $18455K (9.9%) 21590 (17.0%) 27000 (24%)

MATH 25000 (9.6%) 27200 (8.8%) 33700 (23.9%)

PHYSICS 63400 (2.8%) 69700 (9.9%) 82000 (17.6%)

CHEM 51500 (13.9%) 55900 (8.5%) 65900 (17.9%)

MATERIALS 68700 (10.1%) 73900 (7.6%) 86800 (17.5%)

MPS TOTAL 227000 (8.7%) 248200 (9.3%) 295400 (19%)

10 The conventional funding has stayed essentially constant over the three year period FY 80, FY 81 and FY 82. However, based on inflation, the conventional funding has decreased. After extensive discussion the Subcommittee voted unanimously that:

The Subcommittee recognizes that the CER and CSNET initiatives have impacted the conventional programs. Nevertheless, the Subcommittee believes that the CER and CSNET are essential to the furtherance of computer science research and that it is too early to modify the direction being taken.

B. Science Education and Behavioral and Social Sciences

The following motion was adopted:

We recommend that the National Science Foundation be allowed greater participation in the determination of how funds be allocated within the NSF. In particular, we object to any drastic budget reallocations as exemplified by the recent cuts in the social sciences and science education programs, that were made without significant imput from the scientific community.

The vote on this motion was For - 4, Against - 1, Abstain - 2 This motion will be forwarded to Dr. Slaughter, Director, NSF and Dr. Branscomb, Chairman, National Science Board.

C. Election of a chairman for FY 82

Dr. Jack Minker was elected unanimously to serve as Chairman of the Subcommittee for FY 82.

D. Frequency of Advisory Subcommittee Meetings

It was suggested that if funding were a problem that instead of the Advisory Subcommittee meeting once a year that the Subcommittee be decreased to eight individuals and held twice a year. Oversight reviews could use primarily Advisory Committee members to conserve travel funds and continue to hold the reviews the day preceeding the Subcommittee meeting and perhaps hold less frequent reviews.

11 I

December 2-4 inclusive has been set for the next meeting, The meeting adjourned at 4:00 P.M., Friday, May 29, 1981.

The minutes are an accurate and complete summary of the matters discussed and the conclusions reached at the meeting held on May 28 and 29, 1981. Certified by:

for Computer Science National Science Foundation Department of Computer Science University of Maryland College Park, Maryland 20742

Attachments

12 * > UNIVERSITY OF MARYLAND College Park. Maryland

DEPARTMENT OF

Dr. Donald N. Langenberg Deputy Director National Science Foundation 1800 G Street, N.W. Washington, D.C. 20550

Dear Dr. Langenberg: The Advisory Subcommittee to the Division of Mathematical and Computer Sciences has asked me to transmit the following motion to you:

"We recommend that Computer Science become a separate Division within the Directorate of Mathematics and Physical Sciences." The motion has passed unanimously at the May 28, 1981 meeting of the Sub- committee.

The Subcommittee believes that computer science has become a mature discipline and that elevating computer science to division status is long overdue. The retirement of Dr. John R. Pasta as Director of the Division of Mathematical and Computer Sciences makes this a propitious time to -take such an action.

From discussions I have had with my colleagues in the computer science community there is wide spread support for such an action. There is also universal agreement that Mr. Kent Curtis has performed an outstanding job.. His appointment as Director of such a division would be widely acclaimed. Some members of the Subcommittee believe that the Foundation would do well to consider whether or not it is advisable to merge all groups that currently support computer activities Current coordination between the activities appears to be excellent. However, it is inevitable that computer activities will grow and consolidating all activities now could be benefi- cial from an organizational view.

Should you like to discuss these matters further, I would be pleased to do so.

Respectfully yours,

X.

Jack Minker t Professor, Computer Science JM:meg cc : Dr. William Klemperer Dr. Kent Curtis ©

COMPUTER SCIENCE /

/.;?. i

John R. Pasta, Pioneer

-.. f.^.^.^iM^i^''^ W35; 1 In Computer Sciences

' stituted a broad-based graduate pro- wS^r.TsuHlZr gram that attracted students from John Robert Pasta, 62, who retired around the world. last month as director of the division Dr. Pasta joined the National Sci- of mathematical and computer scie nce Foundation in 1970. He headed ences at the National Science Foun- the office of computing activities be- dation, died of cancer Friday at Sibley fore becoming director of the division Memorial Hospital. He lived in Wash-i of computer research in 1974. The ington. following year, the division was ex- Dr. Pasta was internationally rec- panded to include mathematics, ognized for his contributions to the In 1979, he was awarded the foun- design, development and application dation's Distinguished Service Award, of computers. A native of New York City and a A pioneer in developing techniques former New York City policeman, he for solving physics problems by com- attended a special three-year high puter, he worked in this field, now school there and graduated from the JOHN R. PASTA known as experimental mathematics, City College of New York 4 with the Nobel Laureate physicist En- During World War 11, he served as He was a longtime member of the rico Fermi and Stanislaw Ulam, a a cryptologist and radar specialist for AEC's mathematics and computer mathematician, at the Los Alamos the Army in Europe, for which he sciences research advisory committee Scientific Laboratory in New Mexico, received the Bronze Star. He was dis- and served as committee chairman and with John yon Neumann at the charged in 1946 as a captain. during 1965-67. Institute for Advanced Study at Prin- Dr. Pasta was a research fellow at He also was a US. delegate to the ."*~~t\.- r^ut_. * ceton University. the Brookhaven National Laboratory International Federation of Informa- Dr. Pasta, who earned a doctorate for atomic research in Mississippi dur- tion Processing. in physics from New York University, ing 1948-51. He later served as a con- Dr. Pasta was the author of books was a member of the staff at Los Ala- sultant to the laboratory's applied and articles in his field and a member mos during 1951-56, when he joined mathematics department and also was of a number of professional associa- the Atomic Energy Commission and a consultant to industry, tions. He had served as an associate moved to the Washington area. He A proponent of a mathematics in- editor of the Journal- of Computation- headed the AEC's mathematics and stitute, he recently saw the National al Physics. computer division until 1961. Science Foundation approve two such Survivors include his wife, Betty, of From 1961 to 1970, he was profes- institutes at Minnesota and at Berke- Washington; a daughter, Diane, of Se- sor of physics and head of the com- ley, Calif. He served as an adjunct attle; a son, David, of Palo Alto, Calif.; puter department of the University of professor of engineering and applied a brother, William, of El Reno, Okk, Illinois. As head of the university's sciences at George Washington Uni- and a sister, Eleanor Kropilak of Digifal Computer Laboratory, he in- versity. Brooklyn, Mich.

' ■ ■*-■■».-*_"■*

TRIBUTE TO DR. JOHN P.. PASTA

On behalf of the entire mathematical and computer science community, the NSF Advisory Committee for Mathematical and Comput- ing Sciences wishes to express its appreciation to Vi. Jokn R. Fait who has retired as Director of the Division of Mathematical and Computer Science of the NSF, for his outstanding and dedicated servi to the nation. Vx. Paita nurtured a new and growing field, compute science, provided support and encouragement to researchers, obtained increased funding, and helped the field grow to maturity. At the same time he shepherded continuing developments in the mathematical sc iences . Under his leadership new and important initiatives were under taken which will have major impact on the future direction of Loth disciplines. Two such initiatives within computer science are: coordinated experimental research activity, ar.l a network :o in connect computer research organizations throughout the natioi recognition of the needs for alternate methods of support in the matnematical sciences led to a detailed study whose forthcoming report will have far reaching and beneficial impact en the directi and support of mathematics. Throughout his tenure at NSF he focused on major nations comput r policy issues. His leadership during the formative years of science was crucial Whereas man's days on earth are numbered, one's contributions society may be everlasting. Dr. Pasta s contributions to the growth and maturity of computer science and the maintenance and furtherance of mathematics will be everlasting. We thank him for his devotion anf foresight to science in general and to mathematics and computer science in particular.

Tkib tribute, tcdi unanimously pa&std at a mzztinq o{_ the NSF AdvlioKy Committzz (_oi Ua.tke.ma tical and Computing Science, c, May 29, 19. J .

*__

(__