Don’t forget:
[ul]
[li]Cost of whole new software for the new computer[/li][li]Cost of testing new system so it’s 100% reliable[/li][li]Cost of training all traffic controllers to use new system[/li][li]Cost of training all maintenance personnel to work on new system[/li][li]Cost of lost lives, hardware and confidence by travellers if something goes wrong with new system[/li][/ul]
Speed is not everything. Would you fly if the air traffic controllers used software as buggy as those on your PC? In aerospace, reliability is everything. And reliability is expensive. The aerospace industry uses many old technology just because they have proven themselves to be reliable. And I’m glad they do. (Though of course, sometimes you can take it too far and end up with systems so old it’s hard to find replacement parts.)
Soyuz 11 spent 23 days in orbit and docked with the Salyut space station. During re-entry on June 29, 1971 a faulty valve allowed their cabin to lose air. The craft landed successfully. When the hatch was opened, the three cosmonauts Georgi Dobrovolksy, Vladislav Volkov, and Viktor Patsayev were found dead. So obviously the crew had nothing to do with landing it. It was all done by mission control.
You’re wrong. It takes a long time for the engineering on a new spacecraft to be completed, tested, approved, and pass through all the bureaucratic channels. They didn’t slap some off-the-shelf equipment together a couple of days before Columbia’s first launch in 1981. The first designs for a reusable space plane were begun in March 1970. On January 5, 1972, Nixon approved the plan. By then, they already knew what hardware and technologies would be used. The computer systems aboard the shuttle use Z80 processors with ferrite core memory. The Z80 was one of the first radiation-hardened CPUs available and it’s still used on interplanetary missions. I think this means the OS is CPM. The first orbiter, the Enterprise, was constructed shortly thereafter and delivered in September 1976. Although minimally equipped, most of the basic functionality was in place. Formal testing began in 1977. Time from construction of the first orbiter to the launch of Windows 1.0? a little more than 10 years.
Didn’t the computers on Atlantis all get an update when it was refitted a couple of years ago? I thought I remembered reading that.
Actually, I think they have all (save the Enterprise) been updated to some extent.
The reason that NASA would not use up to date technology is because new technology is not yet debugged. Using a chip that is older but more reliable is sometimes the more wise route.
Quoth Scott Dickerson:
Provided that the body’s ratio of cross-sectional area to mass and its reflectivity remained constant, absolutely nothing. It’ll be in a slightly different orbit than it would be were it not for the solar wind, but the orbit it’s in will still be a perfectly stable Keplerian orbit. It makes a difference for something in Earth orbit because the Earth and the satellite don’t have the same mass/area ratio, and it matters for a solar sail spacecraft because the craft’s area/mass ratio will be deliberately varied.
Okay, all that makes sense. And I forgot just how well the equipment would have to be tested, and how long that would take. I also made a mathematical error: 1981 - 1976 = 5, not 15. I was tired, and head math has never been my forte. I’m sure a radiation-hardened Z80 with ferrite-core memory would have improved my calculational accuracy.
So if they have a crew of N, do they carry N-1 body bags? Just curious…