DARPA is soliciting innovative/revolutionary research in distributed robotics, possibly in extremely small robots, reconfigurable robots, systems of robots, biologically inspired designs, innovative methods of robot control (including innovative interfaces), and methods of implementing pooled capabilities and/or layered intelligence. 24Oct97; BAA 97-41. , E.C. Urban, DARPA/ETO, (703) 696-2206 Fax; and R. Dugan, DARPA/DSO, (703) 696-3999 Fax, . . [CBD, 14Aug97.]

The Naval Facilities Engineering Service Center (NFESC; Port Hueneme, CA) has issued a broad agency announcement (SOL N47408-97-R-1879) for basic and applied research in logistics information systems. "DARPA and USMC tasks are working on data management structures, server constructs, and network architectures." There may be set-asides for minority institutions. . Rosalie Hammonds, (805) 982-5097. [CBD, 29Sep97.] (This was filed with keywords CS, AI, expert systems, data analysis, IS, and IR.)

NSF/CISE deadlines coming up include: CISE Research Infrastructure Program, 20Oct97; CISE Postdoctoral Research Associates in Computational Science and Engineering and in Experimental Computer Science, 07Nov97; Computer Systems Software, 10Nov97 (and 13Jan97 for compiling techniques); Software Engineering and Languages, 18Nov97; Numeric, Symbolic, and Geometric Computation, 20Nov97. . [NSF Bulletin, Oct97.]

WebLab and PBS ONLINE have a new Web Development Fund to underwrite independent projects that "promise to seek fresh perspectives on complex issues" or "make imaginative use of the Web and its capabilities as an interactive, participatory medium to explore both personal and public issues." Audio and video streaming, threaded discussions, and a variety of other licensed technologies will be available. Support will average $25K or less, but may reach $50K in cash and services (including any needed guidance, technical support, and design services). 02Nov97 proposal deadline. , or see for an article about the project. [Suzanne A. Seggerman , net-hap, 29Sep97.]

Princeton I is a successor to the annual Paderborn computer Othello tournaments. Entrants must register by 15Oct97, for the 18-19Oct97 competition (via telnet). [Michael Buro , comp.ai.games, 03Oct97. David Joslin.]

Congress appears ready to fund NSF research at 5% above last year, which is more than the President requested. DOD may get no increase for basic research (after five years of cuts), but a 9% increase for applied research. Physics and nuclear science under DOE are doing fairly well, especially if they escape last year's earmarks and recisions. [Robert L. Park, WHAT'S NEW, 03Oct97.]

The US government has over 700 research facilities, plus research projects at universities. It's tough to keep track of all that's going on, and the government has had trouble moving developed technology into industry -- especially when the work cuts across several agencies. Now there's a new model, a "virtual national lab." The prototype is the Extreme Ultra Violet Limited Liability Corp., created with $250M in private money, mostly from Intel. It's a consortium of Intel, Motorola, and AMD, dealing with Sandia, Lawrence Livermore, and Lawrence Berkeley labs in the area of extreme ultraviolet (EUV) lithography. The virtual lab supports 100 government scientists, who remain government employees. Their research is available to any US company that buys in -- for $5M per share -- or to licensees of the consortium. The government gives up patent rights except the right to use developed technologies as it chooses. This virtual lab organization eliminates the inefficiency of each company dealing with each lab directly, and frees up results for commercial use. Federal administrators are considering more such public-private consortia. [Tom Quinlan, SJM, 29Sep97, 1E.]

The research is paid for by private companies, but it's still a way for wealthy companies like Intel to benefit from tax-supported infrastructure. (In return, they create jobs and pay taxes -- but they wouldn't be doing this if it weren't a net win. Is there a corresponding loss by competing companies and from other taxpayers, or is the pie just getting bigger?)

Alternatives to government R&D include in-house labs, proprietary labs (such as Interval Research), industry consortia, and contract R&D companies such as Battelle, MITRE, and SRI International. Plus university research, of course. China is privatizing nearly all of its state-owned businesses, because state control just doesn't work; Japan's MITI has been learning much the same lesson about research management. We've followed the opposite course with R&D, to develop military technologies and as a side effect of easily available grant money (and pork). Under current policies, industry labs and private foundations have withered; nonprofit labs and consortia have failed to prosper; and our new PhDs are seldom able to continue their research.

NSF, NIH, and the government labs have done their jobs well, but too much is expected of them. I favor a reduced role for government science. Let research be paid for by those who see the need, or who can be persuaded to contribute or to invest for profit. Encourage private foundations and grass-roots R&D funding (e.g., university endowments, the March of Dimes, the Loebner "Turing test" competition, the SETI project, or James Randi's $1M prize for any paranormal demonstration). Let NASA sell commemorative T-shirts and webpage banners. For basic research, support university R&D and entrepreneurship via tax policy rather than direct grants. With exceptions, as Congress sees fit.

(Just my opinion, of course.)

A Year 2000 problem recently deleted 90K items from a Defense Logistics Agency (DLA) inventory. Correcting the problem took 400 hours. DLA supports over 1,400 weapons systems, and has 86 automated information systems containing 39M lines of code. The GAO is pushing them to "take fundamental steps" and set priorities for fixing Y2K problems. [Science News, 13Sep97. EduP.]

Ravinder Kachru et al., of SRI International, claim to be less than four years from marketing a holographic storage system for Internet servers. A memory unit the size of a sugar cube can access hundreds of thousands of holograms at 10GB/second. Varying the color of light used could theoretically increase capacity by 1M times. Reliability problems have pretty much been solved. The developers are spinning off a for-profit company, DenseNet, to work on making the technology cheaper. [Science. Eric Berger, UPI Science News, 02Oct97. Bill Park.]

Biometric units for identifying fingerprints have dropped to $300, from $1,200 last year. By next year we may be seeing them in keyboards or mice, reducing the need for passwords. [IBD, 01Oct97.]

MIT's Lincoln Laboratory is developing a tiny reconnaissance plane less then 6" long and weighing two ounces. They are working on small-scale sensors and micro-electromechanical systems, and hope to have a two-gram, one-cubic-centimeter camera for 1M-pixel images. The prototype plane, or "flying chip," will cruise at 100 meters, at 20-30 mph, for up to an hour. [Technology Review, Oct97. NewtNews.]

Rumors say that Apple is now planning for System 8 to be its desktop operating system, with Rhapsody for servers and eventually power users -- much like Windows NT.

There are still 25M Macs in use, but new sales have fallen from 4M to 2M/year. Very few Mac-only developers are left. Still, it's a good sign that MacSoft has finally released a Mac version of the Micropose Civilization II game. Sales have been "brisk, surpassing expectations." [Cox News. SJM, 28Sep97, 4F.]

Heidi Roizen, retired Apple VP of developer relations, has joined the board of Be, Inc. [SJM, 26Sep97, 1C.]

Gateway 2000 Inc., which bought the Amiga brand and patents in March, says it will introduce a new line of Amiga desktop machines within a year. It may support industry-standard graphics cards, universal serial bus ports, and several operating systems. Over 400 companies have asked for developers' licenses since 16Sep97. Amiga was a leader in desktop graphics, animation, 3D, and video editing (competing with SGI rendering engines and Avid digital video editing systems). It could also run most Mac programs. Clubs and newsgroups are still active, including alt.amiga. and comp.sys.amiga. The small, stable Amiga operating system might be good in consumer electronics -- palmtops, TVs, security systems, etc. [Karen Whitehouse, ZDNN, 26Sep97. Matteo Castelli , it.comp.emulatori.] (Commodore, now bankrupt, is widely faulted for neglect and poor marketing of an excellent computer and operating system.)

Dartmouth (Hanover, NH): postdocs in CS, robotics, or microelectromechanical systems (MEMS).

CMU (Pittsburgh): BS/MS programmers for an algebra math tutor.

Rutgers/DIMACS (Piscataway, NJ): postdocs in theoretical CS and discrete mathematics for optimization, DNA computing, molecular biology, etc.

New Jersey: MS+ speech scientist in DSP, pattern recognition for telecom.

UKentucky/CS (Lexington): profs, esp. in AI, graphics, computer vision, DB, HPC, OS, or scientific computing.

Vanderbilt U./CS (Nashville, TN): faculty in AI, experimental systems, SE, DB.

UAlabama (Tuscaloosa): department head in AI, HCI, DB, etc.

Rice U./CS (Houston, TX): faculty in AI, theory, graphics, modeling, systems, DBMS, languages, etc.

Texas - Pan American (Edinburg, TX): dean of the College of Science and Engineering.

UNew Mexico/CS (Albuquerque): chair of the department.

UNew Mexico: postdocs in ML, genetic algorithms, biological modeling.

Washington State U. (Pullman): CS prof in SE, visualization, intelligent systems, or scientific computation.

UCSD/CS (La Jolla): three sr. endowed chairs.

UPlymouth (UK): postdoc in VR, data mining.

Manchester Metro U. (UK): research lecturer in NLP, AI.

EPFL AI Lab (Lausanne, Switzerland): asst prof in constraint programming, new computing methodologies.

UNew England (Armidale, NSW, Australia): lecturer in AI, distributed computing.

"TOMORROW'S PROFESSOR: Preparing for Academic Careers in Science and Engineering" is a new IEEE book by Richard M. Reis of Stanford University, for graduate students and beyond. It describes the North American academic enterprise and offers strategies for preparation, job seeking (while maintaining government/industry options), negotiation, and lifestyle. Over 80 vignettes and personal stories. Softcover, 400 pp., May97; $39.95 (or $35 for members); IEEE Order No. PP5602-QCL, ; ISBN 0-7803-1136-1. . [, misc.jobs.mis, 29Sep97.]

Dr. Dobbs' recent special issue about software development careers is available online at . [Yonat Sharon , comp.software-eng, 30Sep97.]

Although 42% of North American businesses in one study now support telecommuting, only 7% of their employees were doing it -- and that figure has held steady for four years. 20% of telecommuting arrangements fail, due to unrealistic expectations by employees or fear of lost control. For many people, working at home is "just another stress producer." [Tampa Tribune, 15Sep97. NewtNews.]

The Garbo archives at UVassa (Finland) offers a select collection of shareware, freeware and public-domain programs for MS-DOS, plus smaller amounts for Windows, Unix/Linux, X11, Mac, Next, and Sinclair QL. Topics include archivers, virus scanners, utilities; math/stat programs; educational games; AI/NN files; Turbo Pascal and other programming; editors; terminal programs; linguistics; astronomy; and information files. and , plus about 30 mirror sites. For a history, get . [Timo Salmi , comp.archives.msdos.announce, 05Sep97.]

The SoftSEA Shareware/Freeware Programs Center is . [Network News, 02Aug97.]

Kris Kunze runs a shareware website called "Oddities, Curios, and Rarities for Macintosh," at . You can download a simulation of a stapler, or hypnotizing, spinning yellow wheels, or an animated stick figure, or a simulated lobster that squeaks appreciatively when you pet it. [, TidBITS, 15Sep97.] (Kunze says there's nothing like this in the Windows world. "The cold, grey corporate machine strides into the infinite night, filled with a frigid unswerving purpose, dragging its empty and corrupt creators into the dawn of an emotionless void." Or, as he says, "studies have shown that we Macintosh users often lack important enzymes required for the stable processing of thought.")

-- Ken