Comment by onion2k 12 days ago
Would be unpleasant if there was crew.
19 people have died in the 391 crewed space missions humans have done so far. The risk of dying is very high. Starship is unlikely to change that, although the commoditization of space flight could have reduce the risk simply by making problems easier to spot because there's more flights.
My #1 rule for all engineering: simplicity is harder than complexity.
I truly wish more software engineers thought this way. I see a lot of mentality in software where people are even impressed by complexity, like "wow what a complex system!" like it's a good thing. It's not. It's a sign that no effort has been put into understanding the problem domain conceptually, or that no discipline has been followed around reducing the number of systems or restraint over adding new ones.
I've seen incredibly good software engineers join teams and have net negative lines of code contributed for some time.
If we ever encountered, say, an alien race millions of years ahead of us on this kind of technology curve, I think one of the things that would strike us would be the simplicity of their technology. It would be like everything is a direct response and fit to the laws of physics with nothing extraneous. Their software -- assuming they still use computers as we understand them -- would be functional bliss that directly represented the problem domain, with every state a pure function of previous state.
We might get to this kind of software eventually. This is still a young field. Simplicity, being harder than complexity, often takes time and iteration to achieve. Often there's a complexity bloat followed by a shake out, then repeat, over many cycles.
> Their software -- assuming they still use computers as we understand them -- would be functional bliss that directly represented the problem domain, with every state a pure function of previous state.
I love that this is also a model of reality. Everything is made of differential equations.
I've written something good when others look at it and say: "pshaw, anyone could have written that!"
> SpaceX also has a simplification streak: the Raptor engines being the canonical example.
Not necessarily. Your engine which used to have 200 sensors perhaps now only has 8. But you now don't know when temperatures were close to melting point in a specific part of the engine. When something goes wrong, you are less likely to identify the precise cause because you have less data.
Many of those sensors are not to enable the rocket to fly at all, but merely for later data analysis to know if anything was close to failure.
In yesterdays launch, if the engines had more sensors musk probably wouldn't have said "an oxygen/fuel leak", but would have been able to say "Engine #7 had an oxygen leak at the inlet pipe, as shown by the loud whistling noise detected by engine #7's microphone array"
Modern space ships are very likely to change that, as designs mature and improve.
Early aviation was extremely dangerous. Now a plane is among the safest places to be.
Modern space ships are full of sensors measuring everything interesting.
That data is streamed to Earth, so you usually have pretty good indication of what happened, even if your ship is now a puff of stratospheric smoke .
The higher frequency of launches seems likely to have a big impact on reliability. It's no different than deploying once per day vs once per month. The more you do it, the more edge cases you hit and the more reliable you can make it.
SpaceX also has a simplification streak: the Raptor engines being the canonical example. Lower complexity generally means less unexpected failure modes.