Hacker News Reader

Comment by 2OEH8eoCRo0 | Hacker News Reader

If I'm tasked with recording test data for x, y, and z and I only record x, is that a success? Now I need to launch another to try for the y, and z.

Edit: What would an unsuccessful test look like then?

View on HN

If you are planning on several y and z anyways, absolutely. On the most recent SmarterEveryDay video on encasing a Prince Rupert's drop in glass, sculptor Cal Breed talks about the moment when a process fails. He could stop there and restart, saving some time, but instead all the pressure is off, and he "makes as many mistakes as possible" for the rest of the build.

Quote is towards the end, but the whole vid is worth a watch.

https://youtu.be/C1KT8PS6Zs4

https://www.calbreed.com/

Adjacent to this discussion is the "All Up Testing" concept from the Saturn era. The conservative testing strategy was to test each component individually, then put them together and test them as a system. All up eschewed this conservative approach, testing everything that was ready. It's only marginally germane to this discussion, but a great historical note since other commenters are comparing the Starship with the Saturn V.

http://heroicrelics.org/info/all-up/all-up-flight-testing.ht...

Maybe if you were only planning on doing something once. They've already nearly completed the next rocket. The whole point of this launch was a test. There was nothing on it. The next iteration will also be a test, and likely the next few after that. They'll even fly their own satellites up on test vehicles until they've worked out more of the kinks. Then before you know it, it'll be as reliable as falcon 9 and launching 100 times a year.

Look at all the Martian rovers. That is a notoriously hard environment to operate in, so success has often been defined as "the rover functions for at least 30 days". And then some rovers ended up working for over a decade. That doesn't mean that every rover that didn't last an entire decade was a failure though! It just means that, regardless of whether you hit your main goal, if you still have something working left over at that point, of course you keep using it.

The main goal here was to clear the pad and get some atmospheric experience with the entire stack. Goal met. But of course they had contingency plans to get as much more experience with it as possible, if things continued to function nominally (as they have with some rovers, or indeed some previous first SpaceX test flights, like the Falcon Heavy).

As you point out, "success" depends entirely on your goals for the task.

SpaceX said their goal was to get off the launch pad going into this, which indeed may have been setting a low bar for success. However, there is no other alternative definition.

If you are tasked with recording X, and anything else beyond is bonus, doing X is success.

If the rocket had exploded on the pad, that would’ve been an unsuccessful test. The launch pad and tower are way more expensive than the rockets and take much more time to replace. They call it “stage 0.”

Btw this isn’t moving the goalposts. Clearing the launch tower was always the success criterion; the rest is gravy.

yeah I think many people saw the launch as the launch of the finished Starship+booster when what was really happening was a test of maybe v0.0.2 if not pre-alpha. It's not done yet, there's a long ways to go before the expectation becomes 100% success.

You are very brave or didn't read the room :)

I think this could have been more successful

‹Baby stands up and takes 2 steps before falling over›

sidibe: “I think this could have been more successful”

yes, it is.

Reply on news.ycombinator.com