Should we ping DBMS libraries to encourage them to support Temporal?

[justingrant]

Many of our feedback issues mention getting date/time data from a database.

Would it be worth reaching out to maintainers of the top 1-3 JS client libraries each for Oracle, Postgres, MySQL, SQL Server, MongoDB, and AWS and Azure DB libraries? The goal would be to encourage them to think about I/O with Temporal instances so that developers could pass objects to/from a DB.

[kaizhu256]

don't forget sqlite3, the Most Widely Deployed and Used Database Engine

[1st1]

We're considering adding direct support to EdgeDB JS drivers. One of the questions we have is how close is the spec to stage 3? Any ETA?

[littledan]

I like this idea, but maybe the framing should be around asking them for a detailed review with their use cases in mind (once all the materials are ready).

[ptomato]

One of the questions we have is how close is the spec to stage 3? Any ETA?

We're planning to have the API documentation, polyfill, and full spec text mostly frozen and ready for TC39 delegates to review in mid-October. We expect that to take about 2 months, so we expect to be able to propose moving to Stage 3 in the January TC39 meeting. (That's an estimate, not a guarantee.)

[1st1]

We expect that to take about 2 months, so we expect to be able to propose moving to Stage 3 in the January TC39 meeting. (That's an estimate, not a guarantee.)

Thank you, makes sense. FWIW we'll start reflecting our DB datetime types to simple objects with readonly attributes you define in the current spec. The expectation is that when the proposal is implemented in Node, we'll just start returning native Temporal object. Users will have more APIs to play with, but the actual types will be compatible.