Hacker News Reader

Comment by Aloha | Hacker News Reader

I get that there is probably a desire to solely blame Boeing for this - but this seems like as much of a Project Management and integration failure on the NASA side, as it does Boeing putting defective hardware/software into orbit.

It shouldn't have been allowed to happen, period.

I get that what I said will not be popular, but this has been the consensus of every previous thread on this topic that I've seen - NASA is sorta playing the role of a systems integrator here, and assumes the liability for defective components from their subcontractors, and has the ultimate supervision authority to decide if something can or cannot fly.

View on HN

I agree it shouldn't have happened.

> Boeing removed the Starliner’s autonomous undocking feature from its software. The aerospace manufacturer wants to push a software update to the spacecraft in orbit, but NASA fears it could do more damage.

I'm very far away from NASA internals, but as I've seen them in the "older" days was that just a _finished_ product went into space and they would not allow to go into space with something they _knew_ required updates in-mission to complete the _original_ mission.

So it seems like a management issue on both sides.

I think its dumb they couldn't have both software profiles available and be able to choose manual or auto, but the original mission did not call for autonomous undocking.

NASA may be taking a different approach, but in civilian aerospace it is prohibited to have code that doesn't run or doesn't trace to a system requirement. I assume that was their rationale for removing it.

This same base capsule design is also supposed to operate autonomously. It previously attempted an autonomous docking. The code was written, it was approved for flight previously. It is not like "autonomous docking" was never a part of the specification for this model of craft; it just wasn't the expected process for this particular mission.

Boeing just didn't bother to make it so the mission could potentially be changed in-flight, they just decided they'll load a software profile on the ground and that missions will just never change.

No, it's really stricter than that in my experience. You verify your software against the requirements. There shall be no deactivated code, even code which as previously flown before is not safe. See the first Ariane 5 launch for a reference. I assume that is why they removed it.

Of course, this isn't to say that Boeing doesn't have much more serious problems going on, they clearly do.

Starliner was already suffering delays at every step. They probably thought they could shave off a couple weeks by only qualifying the parts of the software absolutely needed for the mission and removing everything else

> I get that there is probably a desire to solely blame Boeing for this

I'm sorry, where do you see this impulse? I'm just not following.

every previous thread this has come up, essentially.

They have to blame Boeing so no one can blame NASA and by association the National Space Council (whose Chairperson is now running for President).

I personally would prefer if NASA handled this internally rather than trotting out an individual to chastise for no benefit to anyone. Accountability on an org? Absolutely. Just don't scapegoat someone when organizational or process incompetence is actually to blame.

Meanwhile, boeing is a publicly traded company with what amounts to a monopoly on many, many, many contracts. What was the point of defeating communism if we just give handouts to non-competitive and highly unproductive organizations who can't even hold themselves accountable? And the people who come out of the woodwork to defend them are even more confusing. Don't you want to avoid burning tax money and calling it progress? If they wanted to be seen as a public good they would have voluntarily given up on profits decades ago instead of trying to come off as parasites.

> NASA is sorta playing the role of a systems integrator here

That used to be the case. For example with the Space Shuttles that is how it went.

With the Commercial Crew Program the idea is that they find commercial providers who take their astronauts where the astronauts has to be when the astronauts has to be there.

One is like building an experimental aircraft from a set of parts. Where they definietly are the system integration. The other is more like paying for airplane tickets.

To quote from the wikipedia page of the program: "The spacecraft are owned and operated by the vendor, and crew transportation is provided to NASA as a commercial service."

> as the ultimate supervision authority to decide if something can or cannot fly.

No. They have the authority to decide if their crew is safe to fly on it or not. But Boeing doesn't need NASA's permission if they just want to fly the Starliner on their own. (As long as they don't fly it with NASA personal, or close to NASA property.)

It is as if you bought a return ticket to fly to some destination and while en-route onwards you heard a death rattle from the engine. Then you have to decide if you trust the company flying you for the return trip, or you find some other way.

Reply on news.ycombinator.com