WASHINGTON, D.C.: 'Direct flights' have never really been all that direct. Because planes have always been guided by a byzantine structure of ground-based guidance systems, pilots have traditionally followed routes straight out of Escher. A Los Angeles-Chicago run, for instance, takes a plane far north, through Nevada, Utah and Wyoming, of the direct route between the two cities. But if a new FAA guidance system pans out, airlines will be able to fly the shortest distance between two points. Beginning in 1999, the agency plans a two-year test of new navigational systems for 2,000 airliners based in Hawaii and Alaska. By choosing their own routes, FAA officials say, pilots will improve flight safety and speed air travel by steering clear of brewing storms and avoiding the long-established but heavily jammed FAA flight routes. The reduced airline operating costs that would result from the so-called "free flight" could also mean cheaper airfares for passengers. One sizeable barrier remains, however. It will cost a mean $7 billion to outfit the 200,000-plane U.S. commercial fleet with the direct satellite links, digital radios, automatic surveillance broadcast systems and other sophisticated components necessary to run the free flight setup. An electronic zone surrounding each carrier will allow air traffic controllers to avoid crashes and order route changes. If the experiment proves anything like the much-hyped Web launch of FAA airline safety and accident data (www.faa.gov), don't look for a speedy trip any time soon. After clicking through reams of bureaucratic mumbo-jumbo, users might find out exactly what the government thinks happened in last year's TWA accident, but more likely than not the agency's server is what crashes.