The Internet, 15 June 2026

I'm happy to announce pyacid v0.5.0a1 — about 10 user-facing changes over 44 commits since v0.4.0a1; highlights below, see the website for the rest.

WARNING: acid is alpha software, developing rapidly — the API will change between releases. Pin a version (pip install pyacid==0.5.0a1) for anything you rely on, and watch these emails for what changed.

Headline new features:

Last release gave you acid search — what's out there to download? This one adds its local twin, acid list — what do I already have, ready to open? It shows every catalog acid open / acid query can resolve by name, with the margin caches available for each.

acid list              # everything on disk, ready to open
acid list gaia         # filter by name

import acid
{c.name for c in acid.list_catalogs()}   # the same view from Python

No more ls-ing through your datasets directory or guessing the exact name a query wants — one command lists what acid can actually open, across local, ssh:// and http(s):// roots, with shadowed duplicates flagged.

Upgrade: pip install -U pyacid · Full changelog →

Also new:

• Query results now read like ordinary catalogs — the internal HEALPix index column (_healpix_29) is hidden from SELECT * and .to_astropy() / .to_polars() output (still written to disk by save).
• acid search against a remote archive fails fast with an actionable hint instead of hanging when a host isn't set up in your ~/.ssh/config.
• Saving a projected catalog (.select(...).save(...)) and re-opening it no longer crashes.

Heads-up: acid.list_catalogs() now returns CatalogInfo rows, not plain strings — use {c.name for c in acid.list_catalogs()}. And Result.save() is gone; write HATS output with Catalog.save(path, name=...). See the changelog.

Thanks for your time — as always, see the website and documentation for the details. If you hit any problems or have feature ideas, open an issue at github.com/mjuric/acid-issues.

— Mario & his AI buddies