If you mostly use chat-style AI tools and you're trying to keep up without turning into a full-time engineer, this is the kind of update that's easy to scroll past and easy to misread. You see datasette 1.0a33, assume it is another niche database release, and move on. That is the wrong read.
If you only look at the surface feature list, you can waste time, budget, and attention in the wrong direction. The hidden cost is worse: you keep thinking Datasette is mainly for showing tables, and miss the part that could change what you build next. An update is worth your time not because of how many features it lists, but because it changes your next decision.
My take is simple: Datasette promotes queries into managed objects. A stored query here is just a saved SQL result you can return to like a page, not a one-off search. In 1.0a33, stored queries got three meaningful controls: edit, delete, and rules for who can change them [S001][S002].
That is why this release matters more than it first looks. Once a query can be edited, removed, and assigned change rights, you are not only publishing a database anymore. You are managing the query itself. The official docs also say stored queries live in Datasette's internal database, can be managed through the web or update/delete routes, and private queries belong to their creator by default [S002].
That does not mean Datasette suddenly became a full CMS. The claim is narrower than that, and more useful: 1.0a33 makes SQL feel like a lightweight content backend for query pages, not just a viewer for raw tables.
This read is based on the 1.0a33 release notes and official docs, not a live team deployment. If you are deciding what to test next, share this with the person who still files Datasette under 'SQLite viewer.' The upgrade to watch is not more table display. It is query management.