I was only supposed to have the stream on in the background yesterday morning. Coffee going, laptop open on the counter, half an eye on it while I answered some emails. That lasted about four minutes. By the time Super Heavy started its boostback burn I was standing there with a cold mug in my hand just staring at the screen like an idiot.
If you missed it: SpaceX flew Starship for the fifth time yesterday, and for the first time they brought the booster back to the launch site instead of dropping it in the Gulf. Not just back to the site — back into the arms of the launch tower itself. No legs on the booster. It just fell out of the sky, did a burn to slow down, and got caught by two mechanical arms (everyone's calling them chopsticks, and honestly the name is perfect) right below the grid fins. The whole descent took something like seven minutes from hot-stage separation to catch, and the launch itself went up around 7:25 in the morning Central time from Starbase.
I've been doing this blog since 2011 and I've live-blogged a fair number of Falcon 9 landings over the years, back when a rocket sticking a droneship landing was the wildest thing going. This was a different order of magnitude. Watching a booster the size of an actual building come down out of the sky, slow itself with its own engines, and get grabbed out of the air by a tower before it ever touches the ground. There's no version of that sentence that doesn't sound like science fiction, and yet there's a video of it that isn't CGI.
The engineering logic actually makes sense once you think about it for a second, which is more than I can say for a lot of "revolutionary" tech announcements I write about on here. Landing legs are dead weight you have to haul all the way to orbit and back. If the tower can do the catching, you delete the legs, the hydraulics, the whole structure, and you save a genuinely large amount of mass you can put toward payload instead. It's the kind of decision that looks completely insane until you realize it's actually the more conservative engineering choice, not the crazier one. I don't say that about SpaceX very often, for what it's worth, I think Musk's timelines are usually fantasy and the whole "humans on Mars by [year]" bit stopped meaning anything to me around 2018. But the catch itself, as a pure piece of controls engineering, is the real thing.
My one complaint about the whole morning: my stream kept buffering right at the worst possible moments, because apparently my ISP decided 9 AM on a Sunday was a great time to have a bad hour. I ended up watching the actual catch itself in 240p on a second tab while the good stream rebuffered, which is a genuinely infuriating way to witness a moment like that. I also tried explaining to my neighbor why I was this excited about "a rocket falling into some robot arms" and I don't think I fully got through to her, though she did admit the replay looked cool once I showed her.
Worth remembering too that the booster catch wasn't the whole flight. The upper stage, the actual Starship spacecraft, kept going on its own suborbital hop and came down for a controlled splashdown out in the Indian Ocean afterward, same basic plan as the last flight in June. It doesn't get any of the attention because nothing is on fire and nothing gets caught by a giant metal hand, but it's still most of the point of the vehicle long-term: the booster catch is what you need to reuse the thing fast, but the ship itself is what's actually supposed to go somewhere.
I've already got a tab open reading about what they're targeting for flight six, and I have no idea how they top yesterday visually. That's sort of the problem with a stunt this good — the next one has a lot to live up to, and "the booster didn't blow up this time" isn't going to cut it as a follow-up headline anymore.