If you mostly follow AI tools through headlines, this is the part you can miss: Datasette 1.0a33 is not just a SQLite viewer update. It changes what you manage. The important shift is that saved queries start behaving like something you can operate, not just something you run. [C002]
You see a database tool update, think "probably for engineers," and keep scrolling. Fair. But this is exactly how people lose time: they track feature lists instead of the one change that should alter their next decision.
My read is simple: Datasette 1.0a33 matters because it promotes saved SQL queries into first-class managed objects, not just output on a page. [C001] [C002]
The proof is deliberately unglamorous. Stored queries can now be edited or deleted from the web UI, and those actions have their own permission checks like update-query and delete-query. That means a query is now something you can manage, not only execute.
That mental model shift is bigger than it looks. If a saved query can live inside the app, stay private by default, and be updated through dedicated endpoints, you are no longer only publishing tables. You are publishing reusable questions.
One update is worth sharing when it changes the question you should ask next, not when it adds the most bullets to a release note. My boundary: this is only about the 1.0a33 release notes and docs, not a claim that Datasette is now a full CMS or a production benchmark.