Hacker News Reader

>> But to require six shuttle launches a year, there would have to be 18 satellites. "Barring some extraordinary breakthrough in technology," says an informed communications industry source, "that's inconceivable."

>This prediction, though, didn't work out so well. (SpaceX alone has more than 5,500 satellites, and is launching thousands yearly.)

Easterbrook was writing specifically about the shuttle. As he explains in the rest of the article, the financials of the shuttle as designed just didn't work without a very high launch cadence. Had each shuttle been capable of landing on its own back at the launch site and be reused 18 times and counting over three years as SpaceX has done with Falcon 9, Easterbrook's conclusions would have been different.

Overall, the article is flabbergastingly predictive of every single thing that happened to the shuttle program over its 30-year operational history. In addition to what you mentioned, Easterbrook also predicted

* The lack of survivability of the crew if an SRB failed (Challenger)

* The fragility of the heat-shield tiles (Columbia, and STS-27)

* The danger to the entire US space program—military, commercial, scientific—if the shuttle, intended to be the sole national launch system, was grounded

In fairness, there were extraordinary breakthroughs in technology between when the article was written and when SpaceX started constellation building.

... like a reusable rocket that's affordable

Not only but also.

On the demand side:

There was no web in 1980 (that was 1989), the non-web internet was barely getting started and home use of it was science fiction (even home computers themselves were only few years old, 3 years if you take the 1977 release of the Apple 2 given the 1 didn't have a case). Mobile phones had been car phones and experimental devices until just the previous year, they wouldn't become commonplace for about two more decades, and even then wouldn't be significantly bandwidth hogs with video calls and streaming until smartphones three decades after the article.

On the satellite side:

Satellites are mostly powered by PV, which itself was expensive in 1980 (I'm not certain, but guessing from the size of the Starlink satellites and the exponential growth curve for PV production and total installation having been only 105 MW globally in 1992, Starlink might have exceeded worldwide 1980 PV production all by itself). Packet switching comes with computational costs, and the computers in the satellites themselves were expensive (both in dollars and watts) and low performance. Phased array antennas were really hard, computational beam forming moreso. Ion drives existed in various forms, but Hall effect thrusters wouldn't be introduced to the west for another 12 years (1992, after the fall of the Soviet Union). Laser communication was being researched, but the first successful laser intersatellite link was 2001.

Genuinely curious

Other than a bit more insight on the impacts on human healthy from long term spaceflight.. What are the major scientific insights or developments made thanks to the ISS?

With Apollo there are quite a few, but with the ISS I can't think of anything

This was recently in the Danish news since it was done by a Danish astronaut on the ISS: https://www.esa.int/Science_Exploration/Human_and_Robotic_Ex...

I understand why people might feel this way. Esp, if you think of the ISS as "a bunch of tired astronauts in a floating lab box"

But the ISS should be considered a manned instrument bus. Consider OCO-3, a gimballed observer on the exterior of the ISS. It has revolutionized our understanding of global carbon cycles, and local carbon sink/sources.

There are numerous such missions that benefit from a common, well supported platform to operate from.

The next one should be robotic though.

https://en.m.wikipedia.org/wiki/Orbiting_Carbon_Observatory_...

but did it need to be attached to a manned space station, or could the observer run from a sattelite?

The cost of attaching a satellite is more than attaching it to the ISS (which has large solar panels, processors, downlink equipment, etc).

We are essentially shipping up an instrument only, rather than the instrument + all the supporting equipment.

For a datacenter analogy, it's like adding a HD to a server, rather than buying a new server for more storage.

It could run from a sat, but then we'd have much fewer missions we could fly.