In 1975, the USSR fired a cannon from an orbiting space station

Astronauts would have been aware of Apollo era automation problems. The most notable the Apollo 11 Rendezvous Radar Switch eating all available CPU and nearly aborting the mission.

There's also a cultural problem here. Many of the early astronauts were test pilots, who came from a "I fly this" culture. Taking away their autonomy was difficult and every screw up that engineers made caused further rifts between the two. Engineers vs astronauts was a common narrative back then.

> History proves you wrong again. A lot of environment built on fear produced rugged dependable products.

I think he meant political fear - in modern terms, the fear of your manager yelling at you, or firing you, or demoting you - and back then, of your manager sending you off to break rocks in the far east.

That's not the same thing as a fear of your hardware killing a man, and over-engineering as a result.

>The shuttle went the way of the dodo and we are using Soviet's 70s technology to get into orbit.

We're between systems, like the seven years between Apollo and the STS. Its a happy coincidence Russia still has a legacy system for lifting people into the largely STS built and ferried ISS. The SLS/Orion and SpaceX systems will be the next generation of spaceflight that will take us out of the LEO. Neither the Soyuz or the STS could leave LEO.

>was the first fully automated space shuttle, so again autocratic state wins?

The Buran sent as many men into space as Zimbabwe. Zero. That's not a "win" in anyone's book. It was a prototype and a blatant copy of the STS.

The STS had automatic capabilities but astronauts, feared rightly, that if the landing gear went down accidentally they would die. The Shuttle does NOT HAVE the capabilities to lift its own landing gear. That's done by a ground crew. If this went off accidentally in space then these people are stranded. If it goes off during re-entry they would die.

Having a removable and stowed cable for plugging the flight computer into the landing gear was a rational compromise. This allowed scenarios where an abandoned shuttle could fly itself home.

> we are using Soviet's 70s technology to get into orbit.

If you're digging that far in the past, we're using German's 40s technology.

> Roskosmos, meanwhile, launched six Saljut stations, Mir, and about as many planetary missions as the rest of the world combined, including the first extraplanetary rovers.

You've mistook Russia for USSR? Roskosmos definitely not a Soviet space agency.

It was to keep astronauts feeling important and in the loop, it was a political decision. If space shuttles could be completely automated imagine how you'd feel as an astronaut, who trained for all those years, waited, and then you just mostly sit and watch the flight computer do the work for you.

I'm not quite old enough, but it was just before my time and there were some stories in the press that it was for shooting their own satellites, at least officially. Remember the space shuttle was being pimped as being all things to all people, including the military, right around that time and satellite retrieval could happen to Russian as well as American satellites, after all abandoned non working wrecks and right of salvage at sea and all that. Now it would be a pity if the Russians shot at a decommissioned spy sat while the Americans shuttle on a classified .mil mission was trying to steal that satellite... also run the numbers on low earth orbit satellites, a decent head shot to a low orbit satellite deorbits the main pieces and the little parts have an awesome surface to volume ratio so they naturally deorbit and the precious delicate space shuttle isn't going to fly thru a cloud of satellite parts to steal a russian spy sat, now, is it?

The shuttle has enough delta-V that in the old days before advanced surveillance sats its not unrealistic it could have snatched a decommissioned low orbit russian spy sat while out of ground observation range. In fact, how exactly do you know what did or didn't happen during the many classified early missions...

There was also a some popular press scientific claims. If "we" think "we" can shoot actively dodging incoming ICBMs out of the sky in the 80s, shooting a "big" asteroid should be pretty easy in the 70s, right? So wait for a close approach, record the spectrograph of the impact, compare to the known composition of the shell, report back on the first direct chemical analysis of a meteor in space before re-entry. This is a little impractical, but ...

Any rational actor does not respect any treaties. That's the normal way of business in international relations. It's a grand chessboard. The point is to put your opponent in check and keep it that way. In order to do that you can utilize any means necessary to do so while appearing as if you are following the rules.

It's the same way all government organizations in the West operate.

"Astronauts insisted that there had to be something only humans could do"

Its been a common rationalization meme for many years because it feels so good.

However the actual reason is during the entire mission everything is either unrecoverable due to equipment failure (see actual disasters) or happens slowly and/or reversibly enough that some combination of human and automatic controls can return to nominal.

Except the landing gear. That deploys and you're done. Not like normal aviation gear where you can lift gear in flight. I would have to look at the L/D ratios and the heading alignment cones but it "feels likely" that an early deployment at any stage of the mission other than 15 seconds before landing means you're not landing on the runway.

Everything else you can compensate for or have an abort mode or so "something". You can only lift the gear on the ground with support crew not wearing space suits removing a spring loaded lock and even then the shuttle doesn't have the plumbing to lift the gear. Simplicate and add lightness.

http://science.ksc.nasa.gov/shuttle/technology/sts-newsref/s...

This is something that never really made it into sci fi books or fan fic or thrillers or whatever. If you want to trap a shuttle in orbit and make it un-deorbitable you don't need to take a fire axe to equipment racks or fire rifles or all kinds of hollywood idiocy, just push the gear down button and its all over, that's one ship that won't be landing without extensive spacewalk work.

So... you don't need autoland capability unless a suicidal crewman or russian agent or whatever popped the gear, because you don't currently have a way to un-pop the gear... Pre ISS there were no failure modes that would result in the crew not being able to ride the orbiter down even if they had to be in suits (O2 failure or something). Post ISS if the O2 system failed rather than trying a landing in full pressure suits they'd probably have tried an auto-land... and guess about when they developed the autoland cable...

In practice if it happened in orbit, perhaps attached to the ISS, someone would figure out and ship up some kind of weird levers for shuttle astronauts to use to re-lock the gear, then do an auto-landing and hope it works. Also I think you have to boot up the APU(s) first before pressing the button would do much of anything.

Something to think about is the shuttle was originally sold to be all things to all people and the .mil probably didn't like the idea of the Russians being able to remotely steal a shuttle or remotely mission kill it.

As a meta question why would you launch a manned mission if an unmanned mission would have worked just as well?

And linked to above, if all you need is a "simple" 25 foot cable to autoland, why will the meme of "can't autoland" never die? Its like talking about mysql not having transactions, despite transactions having been added back in '95 or so.

Also, Digital Apollo by David Mindell.

Early on--being the late 50s and into the early 70s--the desire for pilot control was entirely the rational thing. Automation was very primitive and pretty much limited to doing the right thing only when the situation was nominal. Pilots are easily overwhelmed by tedium and a "pilot in the loop" design methodology prevailed at NASA. Basically that means if a human can do it, they should. The Soviet design took a different tack and invested more in automation, dooming some cosmonauts but being probably for the best long-term.

Nowadays we know how to do some things very, very well--rendezvous--so heavy automation makes sense. In other areas that are more research projects pilot in the loop still makes sense.

It's a tradeoff of efficiency for flexibility and fault tolerance.

What happens when you shoot a piece of cardboard with a bullet in a vacuum though?

(I'm much less confident than I was FWIW)