the number of spacecraft failures recently has been absolutely insane and it all comes down to tech bros barging into the industry going "it's not that hard wtf is nasa so bad" and then completely skipping out on any testing
Recently, a privately funded asteroid mission failed immediately after launch. Here are some choice excerpts from the company's blog post about it:
they cost that much because they do integration testing
.....by skipping integration testing
"skipping integration testing was the right move actually"
come fucking on.
AND YOU FUCKING LAUNCHED ANYWAYS
it failed immediately you dipshits
or you could. i don't know. do integration testing?
source
Hey, Fuckchop: If you did it for 10% but you have to do it 10 times? You fucking failed AND didn’t save any goddamn money.
Even if you had the money to throw away, why would you launch with known problems? What are you possibly learning from this? Were they just hoping those wouldn't matter? "Yeah, whoops, blew up an expensive payload because we figured it was worth rolling the dice on problems we already knew about instead of waiting for a new launch window!"
Launching-as-part-of-iterative-design only makes sense for a kid's model rocket you don't have other testing methods for. Or for things that don't explode.
Launching substandard low-cost products to low earth orbit is a decent way to make a lot of improvements fast. I say this as someone who has directly designed/built/flown over 700 smallsats at a couple different startups. New teams don't realize how difficult troubleshooting in space is until they've done it at least once (or in Planets case.... A half dozen times). But the trick is to do it in low earth orbit which is relatively benign (fuck you solar maximum) and decently accessible for communication.
The problem is this gives these idiots the idea that they can apply the same rapid design to things that are multiple orders of magnitude more difficult. Anything that requires being outside earths atmosphere???? Fuck me man. It's awful.
I'm landing a sensor on the moon next year and every fucking week I have to argue with the CTO that no, we cannot skip this environmental test. Why? Well the last three tests revealed fundamental flaws in our understanding of the expected thermal repercussions of the estimated lunar environment. This next test, which will be off actual telemetry data from a comparable location currently on the moon is also non-negotiable. And it has to happen on the flight model so if we find a big fuck up we have a chance to fix it before delivery.
No it cannot be fixed after shipping by "pushing software".
As someone who has built crewed vehicles under the NASA safety system and built uncrewed NASA instruments there are things I'm willing to be a little looseygoosey with in build-fast-break-things mentality of low earth orbit. I'm not willing to take that risk anything higher than MEO because it's a waste of their money and my time.
"For just 7% of the cost, we can make something that blows up instead of doing the job it was built for!"
