Webb keeps finding more bright galaxies in the first half-billion years than the models allow, and the fixes on offer range from boring to universe-breaking. This one is pleasingly mundane: the galaxies were always there, they just cleared their own fog. Notice how much of cosmology turns out to be housekeeping about dust.

Everyone remembers the noise and the flame; fewer people think about the hole a heavy-lift rocket punches in the ionosphere on the way up. A new preprint maps the disturbance stage by stage, and the effect on satellite navigation signals is the part that makes engineers sit up. Space weather, it turns out, can be locally manufactured.

Rubin has barely cleared its throat and it is already turning up things the galaxy's census missed. Two new satellites from a sliver of preliminary sky coverage is a strong hint of what the full survey will do to the missing-satellites problem. Notice the dust-hidden star cluster: not every discovery is about being faint, some are just about being behind something.

Mars has been mailing us samples for free for billions of years; we just have to find them in the sand. This one is old enough to remember the planet's crust forming, which makes it more useful than its plain looks suggest. A tidy little Kottke pointer to a story that sample-return planners will be quietly jealous of.

Most previews of the Sun's retirement stop at the white dwarf; this one shows the furniture left over afterwards. Two rings of debris with a clean gap between them suggest something sizeable is still orbiting and sweeping the lane, which is the kind of detail that makes a dead system feel oddly lived in.

Hayabusa2 keeps paying out. Finding the full set of nucleobases in material that never touched Earth's biosphere takes the usual contamination objection off the table, which is what made the meteorite results arguable. None of this is life, of course; it's the parts bin, and the parts bin was apparently standard issue in the early Solar System.

Chang'e-5 brought back volcanic rock only about two billion years old, which is awkward for a Moon that should have gone cold long before. The proposed answer is that Earth itself did the stirring, flexing the Moon with gravity and keeping its interior warm. A pleasing case of a sample-return mission handing theorists a homework problem.

Missing your launch window used to mean waiting two years; ESCAPADE's answer is to fire the engines mid-flyby and make up the difference. Worth noticing how much a small, cheap mission can get out of orbital mechanics when it's willing to be clever. The science payoff is two spacecraft flying in formation around Mars, which is a lot more useful than one.

A 1977 spacecraft that still accepts patches is reason enough to read, and this one comes with a milestone attached: a round trip for a single command is now two full days. The JPL team keeps finding new ways to coax the old hardware along. Flight software people will appreciate what it takes to debug anything at that latency.

Built for a two-year mission, Swift spent nearly twenty-two catching gamma-ray bursts in the act and swinging round to look before the afterglow faded. It finally came down over Indonesia, which by Penang standards is practically next door. A fine example of hardware outliving its spec sheet by a factor of eleven.

Worth reading as an engineering plan rather than a flag-planting. The notable choice is two launches and a rendezvous around the Moon, because no single rocket has to carry the whole stack. Apollo veterans will recognise the shape of it, with the docking moved a long way from home.

The textbook line was that a planet-scale collision blasts the atmosphere off into space, which sounds reasonable if you've ever dropped a soufflé. Now the bookkeeping points the other way: the mass delivered by the impactor may matter more than the heat. A nice reminder that planetary formation is mostly an argument about accounting.

We get a sample from another star system by pure luck, point Webb at it, and the news is that it behaves familiarly. That's the exciting part: chemistry that doesn't care whose sun it formed around is a small, quiet argument about how ordinary we are. Borisov had no such telescope waiting for it.

NASA did the punching in 2022; Europe's job is to turn up afterwards with a clipboard. The bit to notice is that the orbit changed by more than anyone predicted, so the follow-up is less a victory lap than a proper loose end being tied off.

Betelgeuse has spent years behaving oddly enough to keep amateurs refreshing their light curves, and now the suspect has a name. Siwarha is a small star doing laps inside a supergiant's outer layers, which is a commute few of us would complain about after hearing it. Worth noticing that the evidence here is radio, not a dimming chart.

Hubble is sinking slowly enough that the deadline sits in the next decade, which by spacecraft standards counts as a crisis. NASA's request is pleasantly open-ended: lift it, upgrade it, or see it off safely. The nice thing to notice is that the agency that once sent shuttles to fix this telescope is now asking who else has a ride.

Four people just spent more than a year on Mars without leaving Houston-adjacent real estate, and the reward is a group photo in actual weather. The scare quotes in NASA's own caption around returns are doing lovely work. Nobody has ever been so pleased to stand on a planet they never left.

The old record was Opportunity's marathon, and Perseverance has squeaked past by a rounding error's worth of Martian gravel. The interesting part is the software: this distance comes from a rover increasingly trusted to plan its own route rather than wait out the light delay for instructions. Flight software people may take a small bow.

Nearly fifty years after launch, the probe gets a milestone with a pleasing unit attached. Spare a thought for the flight software people: every command now comes with a two-day wait for confirmation, on hardware rationing its last watts. That is remote debugging at a scale no on-call rota has ever contemplated.

Fainter than faint, and that's the point: the small leftovers are the ones that keep the receipts on how the galaxy was assembled. Rubin has barely started its survey, so treat this as the opening trickle of a very long catalogue.

Trojans are the asteroids that share a planet's orbit, herded along at its gravitational balance points, and Neptune's have been stubbornly hard to catalogue. These bodies are slow-moving evidence of how the outer solar system got arranged. It is pleasant to see students get their names near the find rather than a footnote.

A cubic kilometre of Antarctic ice makes an unlikely telescope, but it is about the only sort that can see something that passes through you without saying hello. The interesting bit is what these high-energy arrivals might say about the violent places that made them. ScienceDaily's headline does the usual ghost routine; the physics needs no costume.

Bring home a sample from an asteroid and it immediately starts arguing with your models. Ryugu's fine-grained clays vary far too much for one parent body to have had a single watery history, or so it seemed. A nice reminder that the real payoff of a sample return comes years after the capsule lands.

Astronomers have been arguing over this canvas for decades, and it is nice to see the painter treated as an observer rather than a mere visionary. Notice how specific the claim is: a date, a Moon phase and a named star cluster. Whether it fully settles things is another matter, but the method is the fun part.

The ultraviolet sky has gone without a dedicated survey since GALEX shut down in 2013, which is a long time to leave a whole band unwatched. A hundredfold gain over its predecessor is the sort of number that makes astronomers start drafting proposals now. It also fills in the other end of the spectrum from all those infrared surveys.