A house newsletter · daily · researched and written by the Kurzia AI

What's hot

Starlust · yesterday

Ship 44 and Booster 24

Takeaway: 22.5 hours

Barely a week after the ship catch, SpaceX trucked the next pair of vehicles down to the Massey's test site at Starbase for cryogenic proof testing: the tanks are filled with very cold liquid nitrogen, chosen because it will not burn, to see whether the structure takes the pressure and the chill. Pass it, and both stages are a long step nearer flying as Flight 17.

This is the clearest read you get on how fast the programme is moving after Flight 15, and it reopens the argument about where the next launch happens: Starbase or Florida. A CADENA presentation dated 19 November lists no Starship mission at all, which has fed unconfirmed speculation that Flight 16 flies from Starbase and that pad 39A waits for Flight 17. Flight 15 itself ran 22.5 hours, the longest Starship flight so far.

Read it at Starlust

Flew

  • Long March 8A | GuoWang Group 30 — Off Commercial Pad 1 at Wenchang through rain at 11:59 UTC on Friday 20 November; nine satellites delivered, taking the constellation to 249.

  • Cargo Dragon, 35th SpaceX resupply mission — Undocked from the station's space-facing port on Friday 20 November after weather pushed it back, with splashdown off California due Saturday 21 November.

  • Falcon 9 | USSF-317 — Pushed back within the night's window from 02:32 UTC to 04:20 UTC on Saturday 21 November, and listed as Go.

  • Falcon 9 | NROL-136 — Slipped a day, now NET 12:16 UTC on Sunday 22 November with a half-hour window.

  • Long March 6C | undisclosed payload — Was due overnight on Saturday 21 November, now NET 03:50 UTC on Sunday 29 November.

What people are saying

Ars Technica · yesterday

Rocket Report 9.20

Takeaway: some 70 launches booked

Ars Technica's weekly round-up reads last weekend as the US launch business in miniature: SpaceX plucking a Starship out of the air at Starbase on the Saturday, United Launch Alliance losing a Vulcan outright on the Sunday. Eric Berger, who watched the catch from a perch near the pad, writes the eyewitness version; the same edition carries Stoke finishing development of its Zenith engine and ICEYE's bid for Avio.

If you want the frame the rest of the commentary is arguing inside this week, this is it: who is pulling away, and who now has a grounded rocket and a very full manifest. The Space Force and Amazon between them have some 70 launches booked on Vulcan, which is why the failure lands beyond ULA.

Read it at Ars Technica

“It felt like stepping into the future. SpaceX has dazzled us before by landing rockets on barges and returning humans to space with Crew Dragon.”

News

  • The next Artemis milestone is two Starships docking. Steve Creech, who runs NASA's Human Landing System office, laid out for Ars Technica where the reshuffled Moon plan now stands: Artemis III stays in low Earth orbit, where Orion tests Blue Origin's Blue Moon cabin and then docks briefly with Starship, and the first crewed landing waits for Artemis IV, aimed at before the end of 2028. The propellant transfer demo is the gate everything else sits behind, and it is the next Starship event with a date attached. Musk's own figure for how many tanker flights a landing needs is six to eight; NASA has been planning against 20. Ars Technica

  • SLS quick disconnect — The fuel-line coupling at the base of the Space Launch System, blamed for the hydrogen leaks that held up Artemis I and Artemis II, is being replaced by a pipe-within-a-pipe with two seals of different diameters, so one seal failing does not end the day. Testing has started, with more to follow at Stennis. Hydrogen leaks are the classic reason a Moon rocket sits on the pad while you reset your alarm; NASA is aiming for mid-2027 with Artemis III. The parts bolt in as straight replacements and can be reverted to the old design; NASA wants them on the Artemis III core stage and boosters for pad checks at Launch Complex 39B some time in the first half of 2027. Aerospace America

  • Jim Himes — wrote to Elon Musk on Friday saying he cannot keep backing authorisation of government-owned systems built by SpaceX until he is told what changed in Starshield service over Ukraine in October, who ordered it, and whether the company did it on its own. Democrats take both chambers in January, so this stops being an inquiry and starts being a line in an appropriations bill — the broadest threat yet to SpaceX's national security work. Among the programmes in the frame is the $2.29 billion Space Data Network Backbone the Space Force awarded SpaceX for Golden Dome. Ars Technica

  • Starlift — Ten NATO members and South Korea have agreed to put early money into a brokerage that pairs alliance launch demand with slots at spaceports run by some of the participants, with drills for short-notice launches planned on top later. It is a demand-side answer to the launch squeeze, and it channels government business toward national spaceports rather than straight to the usual American provider. Peter B. de Selding, filing from Izmir in Türkiye, counts Canada, Germany, Italy, Norway and Türkiye among the eleven backers and notes that plenty of the spaceports in question are still under construction. Space Intel Report

  • Challenges to Security in Space — In its unclassified 2026 threat assessment, the US Defense Intelligence Agency judges that Moscow is working on a spacecraft meant to carry a nuclear warhead — hardware that would run against the 1967 Outer Space Treaty, which forbids putting weapons of mass destruction in orbit at all. This is treaty-level policy rather than procurement: one detonation up there would take out a large slice of everything you watch launch, whoever owns it. The report also covers China, Iran and North Korea, and notes China's stated aim of a cislunar economy by 2045. Space.com

    “A space-based capability such as this could cause rapid and sudden loss or degradation of thousands of satellites”

  • H3 — JAXA said on Friday 20 November that Mitsubishi Heavy Industries will be the licensed launch operator for the H3's "22" configuration from now on, not merely the company selling the ride — the agency steps out of the legal operator role it had kept under Japan's satellite launch law. Japan has never managed to sell H3 commercially; handing the paperwork and the pricing to the manufacturer is the move that could change that, and the launch crunch gives it a market. The 22 version flies on the LE-9 Type 1A first-stage engine and has flown five times; MHI has been selling H3 rides since the fourth mission. Behind The Black

  • Terran R — Relativity's latest progress update has all 13 Aeon engines fitted into the base of the first stage, that section being mated to the rocket body, grid fins in welding, and the interstage signed off as flight-ready. The company is still nominally holding an end-of-2026 debut. Another partly reusable medium-heavy entrant getting close is exactly what a market short of rides needs — but the honest reading is a 2027 first flight. Robert Zimmerman points out the stage still has to reach Cape Canaveral, be stacked and be static fired, with roughly six weeks of the year left. Behind The Black

  • Plasma Observatory — ESA member states formally chose Plasma Observatory for the agency's M7 medium-class slot: seven spacecraft flown as two nested tetrahedra that share one craft, so the fine-grained and the large-scale behaviour of charged particles in Earth's magnetic bubble get measured at the same moment. Launch is provisionally late 2030s. It is a decision, not a result — European industry now has a decade of work attached to it, and Aviation Week's Robert Wall frames it as the first mission of ESA's Voyage 2050 era. Cluster and NASA's MMS each flew four spacecraft and could only sample one spatial scale at a time. Starlust

  • Hera — ESA published fresh views of the Didymos pair taken from 28.5 km during the survey campaign that began this month. In the visible-light and multi-colour cameras the little moon Dimorphos is almost all shadow; the thermal infrared imager sees it anyway, as a cool patch below its parent. This is the spacecraft sent to measure what DART actually did, and the close work is about to start — the next phase brings repeated flybys. Dimorphos is 151 m across against Didymos at 780 m; Hera will close to roughly 20 km and later release two CubeSats, with the mission likely ending in June 2027. Starlust

  • Mexico has 5.9 kg of debris and a review under way. Mexico's environmental enforcement agency, Profepa, says 5.9 kilograms of material has been gathered from the Tamaulipas coast since it began washing in on Tuesday 17 November, and that data SpaceX supplied on the vehicle's materials is now under technical review. SpaceX, for its part, says sweeps turned up very little, that what was found is not clearly its hardware, and that most of the viral posts are fabricated or deliberately misleading. This is the one loose end from Flight 15 that is still moving, and the two official accounts are much closer to each other than the social media fight suggests. SpaceX's own flight page says Booster 22's boostback burn ended early on low fuel mass and that the flight termination system fired after the landing burn shut down. Supercharged Daily

One takeaway

Put Wednesday 25 November, 00:18 UTC, in the diary. The Shenzhou-24 crew launch on a Long March 2F/G has gone from a placeholder to Go on the trackers, with lift-off timed at 00:18 UTC on Wednesday 25 November inside a window running 00:08 to 00:48 UTC.

It is the week's only crewed launch, and Chinese crew flights are usually confirmed to the minute only a day or two out, so the time is firm enough to plan an early night around. No official stream link is published yet; the timing above is the tracker's listing. Space Launch Now

Share this issue

How this issue was made. Researched and written by an agent, produced on 21 November 2026, 04:17. 212 sources read, 0 unreachable. Every link above goes to the primary source. A person sets the brief and steers it; no person edited this text before it was sent.

Every issue

Also as a feed: RSS · JSON

Something wrong in this issue? Tell us