Computing Research Association (CRA) has a new Web page featuring case studies to justify basic research funding in CS and computer engineering. Visit , or see the "What's New" entry on CRA's home page, . [Rick Weingarten , CRA Bulletin, 5/24/95.] (NSF should collect such information, but program officers have no time. Here's one example of a private organization doing a better job.)

Senator Pressler (R-SD), Chair of the Commerce, Science and Transportation Committee, says basic research should be privatized since "most major discoveries come from industry." On the other hand, he strongly favors NSF's EPSCOR support to states [such as his own?] with little research infrastructure. [Robert L. Park, WHAT'S NEW, 5/26/95.]

(NSF control was mostly in the hands of scientists until the agency recently grew large enough to attract budget committee notice. Congress has funded basic research "open loop": pumping money in without examining what comes out to guide further investment. NSF program directors don't have time for analysis, and don't report directly to Congress anyway. They can compile histories of research results -- or at least counts of papers published and students graduated -- if Congress ever asks, but Congress never does. Overview committees verify that the process is fair and reasonable -- an American fascination -- but individual grant results are not checked unless fraud is suspected. (Final reports are mandatory, but typically get only a quick scan for acceptable content.) Congress is tired of buying research by the pound, but is making cuts the same way: by opinionated guess. It's even planning to eliminate the Congressional Office of Technology Assessment, one of the few agencies capable of critiquing or guiding research funding. Will this open the door to lobbying, cronyism, and pork-barrel funding? You betcha. But perhaps the amounts will be small, if basic research funding is the responsibility of industry, private foundations, venture capitalists, entrepreneurs, universities, state development agencies, military sponsors, and a few federal labs and mission agencies. Somehow it worked pretty well before we had NSF, NASA, DOE, and Big Science. Multisource funding is complex and confusing, but creates entrepreneurial companies doing market-driven and capability- driven R&D in diverse niches. Grand Challenge funding -- the space race, atom smashing, aerospace, HPCC, NII -- supports proposal mills: large "me too" companies competing for any pot of money offered, without the corporate vision to do independent research or to exploit the technologies they and others develop. As long as their big competitors played the same game -- as they did in the US, until the Japanese became competitive -- few independent researchers were needed.)

"Whenever you have an efficient government, you have a dictatorship." -- Harry S. Truman. [EFF Quotes 6.0, 5/18/95.]

Harvey Robbins and Michael Finley say that companies tend to be growth-oriented (focused on expansion, core competencies, new products, new markets, and profit by doing) or "bottom line" (cost containment, downsizing, flattening, delayering, dehiring, and profit on paper). "Numerator" companies create something terrific and new; "denominator" companies compete in mature markets that can't be expanded. Numerator companies tend to be better at exploiting team dynamics. "A team is a surprising, perplexing, up-and-down, tragicomic, value-creating human thing that needs a ton of attention. That has to be pampered, fed, stroked, and have its pen hosed out from time to time." Teams need tools, vision, rewards, attention, or simply clarity. "Why Teams Don't Work, $20.95 from Peterson's/Pacesetter Books, 1-800-338-3282. [, alt.books.technical, 4/29/95.]

There's a new trend afoot. Companies are turning to professional design firms for complex development formerly done in-house: user research, industrial design, mechanical engineering, rapid prototyping, ergonomic studies, GUI and software design, graphics, and packaging. The design firms have high-priced talent and also workstations/software for 3D surface design ($90K), high-res color printing ($100K), physical simulation ($70K), mold-flow analysis, CAMAX rapid prototyping ($50K), and milling machines ($200K). Client companies need handle only production, marketing, and distribution. (Apple and NeXT, for instance, both used design companies to develop their innovative products.) IDEO Product Development (Palo Alto) and Fitch Inc. (Columbus) are two full-service, global design companies with about 200 employees each (growing at nearly 10%/year). IDEO even has an IDEO University to train clients' industrial designers, to pass along the new culture as well as new products. Mid-sized shops with 30-50 people specialize in design for certain industries. [Bruce Nussbaum, BW, 6/5/95, p. 88.] (Other firms around Palo Alto include GVO, Lunar Design, Palo Alto Design Inc., Venture Design, David Hodge Design, Vent Design Associates, Ginko Design Inc., SNAP! Design, Studio Red, frogdesign (Sunnyvale), BBiB (Los Gatos), and Bridge Design Inc. (San Francisco). See Business Week for ads from 63 design firms, many with email addresses. For the Industrial Designers Society of America (Great Falls, VA), contact , (703) 759-7679 Fax.)