It does seem like quite a leap from the FTS triggering before reaching orbit to refueling, doesn't it? My hunch is that whatever happens, we'll see scenes of SpaceX people clapping and cheering and SpaceX fans calling the flight a success. From what they tell me, no SpaceX launch has ever failed.

No one's going to argue https://en.wikipedia.org/wiki/SpaceX_CRS-7 wasn't a failure. That's a production flight that went ka-boom with a lost payload.

Test flights are a very different beast. Early tests can be expected to fail; some are supposed to fail ("destructive testing"). Explosions and other failures become much more concerning later in the process; for example, Boeing's Starliner's latest test flight was supposed to demonstrate everything was working correctly. It failed to do so.

In the two Starship launches so far, SpaceX was very clear with expectations in advance. Each likely has dozens of objectives. The second flight very clearly accomplished more than the first; a successful test of the launchpad mitigations, a successful test of all first-stage engines staying lit, a successful test of separation, a successful test of the second-stage engines all lighting, etc.

how come people with sliding rulers and navigation computers 'knitted' by local grannies have lost 0 Saturn rockets.

But now an explosion is an expected outcome for your million dollar hardware.

That approach works in software dev.

Because Saturn rockets cost 1.3 billion dollars EACH.

A test starship is nowhere near that, though we don't have exact numbers. Probably closer to $100 million.

Also SpaceX is trying to solve a much harder problem. If you would have asked that Saturn team "Also, land all the stages of the rocket", the likely answer you would have got is "Screw off, you're insane" or "Give me 100 billion dollars".

Also getting Raptor engines to work. The second Starship test was a massive achievement because the SuperHeavy Raptors all worked. That's an engine design with numerous advantages which wasn't practical when the Russians tried it, but might now be practical (certainly looks to be from the test) because control systems technology is small and powerful enough (i.e. similar to how modern fighter jets fly in aerodynamically unstable regimes requiring active computer control - we have them now because we have the computers which can do it).

> how come people with sliding rulers and navigation computers 'knitted' by local grannies have lost 0 Saturn rockets.

The fuck? NASA lost people in production hardware multiple times.

https://en.wikipedia.org/wiki/Apollo_1

https://en.wikipedia.org/wiki/Apollo_6 "The damaged third-stage engine failed to restart for trans-lunar injection... Despite the engine failures, the flight provided NASA with enough confidence to use the Saturn V for crewed launches..."

https://en.wikipedia.org/wiki/Apollo_13

https://en.wikipedia.org/wiki/Space_Shuttle_Challenger_disas...

https://en.wikipedia.org/wiki/Space_Shuttle_Columbia_disaste...

Here's a fun one: They tested the Saturn crew escape system. The booster was supposed to blow up, and demonstrate the crew would survive... but the test went wrong, and it blew up early, when it shouldn't have. The crew abort worked, so test failed successfully.

https://www.youtube.com/watch?v=AqeJzItldSQ

Apollo 1 was very much not production hardware. "The Vintage Space" channel on YT has a good rundown, "Duplicate Apollo Spacecraft". They didn't lose any crew on production hardware, although 13 was close.

AS-204 was absolutely production hardware. The later hardware was different because three people died. They died in rehearsal for an actual mission.

> Apollo 1, initially designated AS-204, was planned to be the first crewed mission of the Apollo program.

https://en.m.wikipedia.org/wiki/Apollo_1

The second version was already starting production when three people died. There was never any intention to build any more of the first one, or to send it beyond low Earth orbit.

Changes were made based on the fire. The capsule was “production” in the sense that it was going to carry people. It doesn’t need to be mass produced to be “prod”. It’s prod like “this instance matters”.

Apollo 1 should have been production like it mattered. But instead the first prototype was pressed into service because the real one wasn't ready yet. The fire was proof.

“In service” and “production” are synonymous in this case.

Not as it turned out.

No Saturn V launch failed. The very first all-up test of the Saturn V, Apollo 4, was a complete success. Apollo 6 was mostly a success, aside from the failure of the S-IVB to restart, and no part of the vehicle was lost. Every crewed launch of the Saturn V was a success, the Apollo 13 failures were unrelated to the Saturn V. For the uncrewed launch of Skylab on SA-513 the booster's S-IC and S-II stages worked normally. The third stage was Skylab itself, a converted S-IVB. It reached orbit successfully despite sustaining severe damage resulting in the loss of the station's micrometeoroid shield/sun shade and one of its main solar panels.

The "accidental abort test" was launched on a Little Joe II, a simple single-stage solid rocket motor, not a Saturn of any variety. This booster was unrelated to any Saturn hardware.

So, if Starship killed a crew of three in a pre-launch rehearsal like Apollo 1 did, you’d be on here defending it as not a problem?

If the Starship second stage failed to restart causing it to fail to get to TLI, or tried to take out its manned crew like 13 did, you’d be claiming it’s still a sterling safety record?

You’d have considered this test a success if it reached orbit “despite sustaining severe damage” on the way up?

Come on.

> Every crewed launch of the Saturn V was a success

Fine. Same for Starship, so far.

“My hunch is that whatever happens, we'll see scenes of Apollo people clapping and cheering and Apollo fans calling the flight a success. From what they tell me, no Apollo launch has ever failed.”

It's ok to just say, "You're right, I was wrong". Try it some time, you might be surprised how much better you feel about yourself.

Does this approach work for you often?

If you're exploding million dollar hardware w/ your software development, maybe it's time to change careers.

Every Saturn rocket was lost.

We know right where each came down.

a) Those people with sliding rulers and navigation computers also had 2.5 % of American GDP annually at their disposal, which is several orders of magnitude more than anyone now has. (Be it SpaceX or Boeing or whoever...)

b) The Apollo program actually cost not just money, but human lives = a much worse outcome than just losing a test rocket. And Apollo 13 crew was very lucky to have returned to Earth alive.

c) Losses of test hardware can only really be somewhat prevented by being extremely careful and pedantic during development, which takes time, and guess what? More time spent obsessing about every single detail = more money spent on development. It is entirely possible that the "hardware is expendable" approach is much cheaper than what you would like to have.

If a test flight doesn't achieve the testing objectives, that's a failure. No one would argue that failing to find a bug before putting software in production was a testing success just because all the tests executed without errors.

There are multiple categories of objectives... Objectives that motivated the test flight. Objectives that are expected to succeed. Objectives that you threw on because "if the $xx million object is still intact it would be nice to get some useful data, and the FAA wants a complete flight plan anyways".

If that last category doesn't succeed that doesn't mean the test flight was a "failure" from the businesses perspective. They still got what they wanted, they just didn't get the icing on top.

The test flight that destroyed the pad is properly categorized as a failure. The flight termination system was expected to work. They expected not to cause that much damage to their infrastructure and surroundings, their relationship with the FAA, their ability to get licenses, etc. To argue that it wasn't a failure, you have to argue that the flight termination system was in the category of systems where a failure was considered acceptable. If that was the case, everyone involved in launching that flight ought to to go jail for recklessly endangering the public.

The most recent test flight though seems to have been a pretty resounding success. Everything that they said they really wanted to see happen, happened. Yes, both stages had failures in "we might not even get here" test objectives after that. That's ok.

> If a test flight doesn't achieve the testing objectives, that's a failure.

Demonstrably incorrect.

Examples:

https://www.youtube.com/watch?v=AqeJzItldSQ is a Saturn launch abort test that failed. It failed in a way that they were able to determine that intended test would have worked, though. 1:20 for the meat of things.

https://en.wikipedia.org/wiki/Apollo_6 also failed to achieve its objectives; "Despite the engine failures, the flight provided NASA with enough confidence to use the Saturn V for crewed launches; a potential third uncrewed flight was cancelled."

All that aside, SpaceX was quite clear that the latest Starship test had multiple objectives, and that "successfully splashdown near Hawaii" was deemed quite unlikely. It successfully demonstrated a number of items, including fairly critically the launchpad changes. It wasn't one test; it was a bunch of tests in one flight.

If I remember correctly there has been a couple of falcon failures, it should be noted that the starship project is is still well in the development phase.

The success criteria for integrated launch test 1 was to clear the launch pad (which it nearly destroyed) test 2 was get past the stage seperation, which it did.

Spacex have a very fast cadence for tests and seem not to make the same error twice. The best test environment is really trying to fly the ship.

You are getting downvoted for stating the obvious - elonheads dont like that.

They have not a stable platform, but will be testing advanced features?

This program is starting to take queues from its soviet ancestor.

How is that exactly?

Launch one really was a comedy of errors in the launch platform.

In launch 2 the booster did exactly what every other booster by everyone else (except the falcon 9) does. Got it's payload to the target altitude before separation.

The second stage reached space.

The N1 platform didn't do any of that.

You have no idea what a testing platform looks like, and are too used to NASA one shot rockets that cost 3 billion each just to make sure nothing goes wrong.

The booster exploded after separation, which is never a good sign. Normal boosters for modern rockets don't do that.

The second stage exploded as well, possibly because of FTS, possibly not (was it ever confirmed?).

Well it's not a normal booster. It's a test booster. Also normal boosters don't do a flip while getting blasted with rocket exhaust.

SpaceX clearly has a winning method, as they have a success rate of 99.3% and hold the record for most consecutive successful launches at 253. Second place is Soyuz-U, which had 112 consecutive successful launches and an overall success rate of 97.3%. SpaceX has more consecutive successful booster landings (173) than any space program has had successful launches.

> The booster exploded after separation, which is never a good sign. Normal boosters for modern rockets don't do that.

It exploded while attempting to test its landing maneuver. If it just turned off the engines and dropped itself into the ocean like a normal booster, it likely would not have exploded (for one thing, there wouldn't be any fuel left in it being saved for landing).

It exploded after flipping around 180 degrees.

A few people have slosh modeled the booster flip around duplicating observed speeds and accelerations and surmise the following. The flip around was too rapid causing slosh to one size of the booster. In addition there was negative acceleration that would have allowed the remaining fuel to experience zero gees pulling it away from the bottom of the rocket. When the engines fired back up it would have hit the bottom of the rocket with something between 20 and 100 tons of force. Based on SpaceX description, the FTS was not used.

The second stage at least appears to be oxygen leakage. The graphs shown on flight day started showing a more rapid depletion of oxygen after second stage lighting. SpaceXs language was something like 'flight was unable to reach orbital insertion and was terminated, pointing to the FTS being used.

SpaceX is taking a very different approach from those of other launch companies, taking a far more rapid prototyping style approach to development, knowing that the first few iterations are going to have problems that they can fix in the later versions.

This sounds wasteful on paper, but in practice it is much much faster and amazingly enough cheaper. With traditional rockets you never make enough to enjoy economies of scale, but with SpaceX's approach they've dropped the cost of the engine by about an order of magnitude over just a few years, and it was already a very inexpensive engine. For the cost of a single RS-25 you can buy 400 Raptors.

Exactly. With traditional development it is very expensive if you find a core problem in late development. In rapid development you expect to catch it early.

Seems to me it's more the inverse - Elon bashers trying to find a way to paint everything he's ever done as a dumb failure because they hate some unrelated thing that he did. Not saying the guy is perfect, but if he is a fuckup, you should be able to make the case without disingenuous arguments like misunderstanding the purpose of test flights.

Multiple iterations of test flights that are expected to fail is a much faster way to develop impressive tech than spending 3x longer doing on the ground analysis of every detail of everything because you can't bear the PR hit of something that can be painted as a failure.

I think you're showing a lack of understanding of Spacex's fast development processes but as an aside the word cue (rather than queue) is used when we are referring to a sign or signal given by something or someone.

I know nothing about rockets but bits of project management: it's common to break down projects into smaller parts and derisk designs/implementations by testing each piece.

The refueling team might be ready to test, even though the "don't explode" team are still working on it.

This is roughly equivalent to watching a developer work and insulting them when their code doesn’t compile. Nobody who writes code cares, that’s just what writing code is like. You could spend a ton more time writing code trying to get it to compile without errors ever, but that would be a whole lot more pointless work.

Writing software and building rockets are unlike each other in many respects; the cost of small errors being high among them.

> the cost of small errors being high among them

There are some great documentaries about the NASA team that does safety critical software: https://www.youtube.com/watch?v=iJNNJxC12UM

There are costs and then there are trade-off costs. In the long run, the trade-off costs of slow-moving perfectionism are FAR higher than the cost of the ‘fail fast and early’ culture at SpaceX.