2.0 release notes November 2026

types and constants go to 2.0.0. The other packages, temporal, equations, maps, tracking and mcp, get a patch release that moves their range to the new types, and nothing else.
These notes cover what changed between 1.x and 2.0. For the code changes, step by step, see the migration guide from 1.x to 2.0. Questions? See getting help.
Overview
2.0 is a correction. The galaxy dataset had values that were wrong by orders of magnitude, and one that described something that doesn't exist: a supermassive black hole in Triangulum, which observations rule out. Fixing that needed a type change, blackHole optional, and that's a breaking change, so it waited for a major. The major also removes MeasureInterface, which was deprecated before 1.0 and announced for removal in 2.0.
It matters to you if you use GALAXIES, or GalaxyInterface and BlackHoleInterface in your own data, or still import MeasureInterface. Nothing else changes.
What's new
Galaxies at their real distances
Andromeda was 2.5 trillion light-years away in 1.x, a million times too far, and Triangulum 3 trillion. Both are now at their distances from Cepheid variables, measured with the Hubble Space Telescope:
| Galaxy | 1.x | 2.0 | Source |
|---|---|---|---|
| Andromeda | 2 537 002 537 000 ly | 2 482 050 ly (761 kpc) | Li et al. 2021 |
| Triangulum | 3 000 003 000 000 ly | 2 739 714 ly (840 kpc) | Breuval et al. 2023 |
Triangulum's distance is still debated: other methods give up to about 880 kpc.
A black hole only where one is detected
GalaxyInterface's blackHole is now optional. Triangulum has no supermassive black hole: the stars at its centre limit one to about 1500 solar masses (Gebhardt et al. 2001). In 1.x it had one of 1.5 million.
The black holes that are there have their measured masses, and their radius is now their Schwarzschild radius,
| Black hole | Mass | Radius | Source |
|---|---|---|---|
| Sagittarius A* | 4.2996 × 10⁶ M☉ | 0.085 au | GRAVITY Collaboration 2024 |
| M31* | 1.4 × 10⁸ M☉ (1.1 to 2.3) | 2.76 au | Bender et al. 2005 |
Each black hole sits at its galaxy's centre, with x and y 0, as BlackHoleInterface describes. In 1.x, the black hole's y held its galaxy's distance in au. Real 3D positions for the galaxies are planned (#373).
Diameters of one kind
A galaxy has no sharp edge, so a diameter depends on where you draw it. The diameters are now all the D25 diameter, where the disk fades to 25 B-magnitudes per square arcsecond: from the RC3 catalogue for Andromeda (137 570 ly) and Triangulum (56 420 ly), and an estimate from a model of its disk for the Milky Way (87 410 ly). The stars reach further out than that. The types are now the de Vaucouleurs types: SBbc for the Milky Way, SA(s)b for Andromeda and SA(s)cd for Triangulum.
Every value has its source in the dataset's documentation and in NOTICE.md, and the tests check each one against it.
What's not changing
- Everything outside the galaxies. The planets, moons, stars, asteroids, comets, belts and systems keep their shape and values.
- The other packages' APIs.
temporal,equations,maps,trackingandmcponly move their range totypes2.0. - Galaxy ids,
milky-way,andromedaandtriangulum, and the fields other thanblackHole.
Deprecated
Nothing new.
Removed
| API | Use instead | Deprecated in |
|---|---|---|
MeasureInterface | DistanceInterface | 0.63 |
MeasureInterface was identical to DistanceInterface, so the change is a rename.
FAQ
Do I need to upgrade right away?
No. types and constants 1.x get fixes for high and critical vulnerabilities for 6 months after 2.0, as the security policy describes. But only 2.0 has the corrected galaxy data.
How big is the upgrade?
Small. If you don't use GALAXIES or MeasureInterface, upgrading the packages is all there is. If you do, check for blackHole before reading it, and don't place galaxies by their black hole's position. The migration guide has a checklist.
Upgrade all the packages together, so they share one copy of types.
Does this change how the MCP server is set up?
No. Its galaxies resource serves the corrected data, so an assistant sees Triangulum without a black hole.
Getting help
- Something doesn't work after upgrading? Check the migration guide, then open an issue with the package versions and a short example.
- A question? Open an issue too; questions are welcome.
- A security problem? Don't open an issue. Report it privately, as the security policy describes.
- Every change in detail: each package's
CHANGELOG.md.