Release 0.0.17

[eldonm]

Summary

Release PR for 0.0.17, per RELEASING.md: bumps jvspatial/version.py (what the publish workflow reads) and dates the [Unreleased] block.

Ships the DatabaseConfig field-name fix from #37 — worth releasing promptly, since it means app.yaml-supplied DynamoDB settings have never taken effect and Postgres inherited the same problem in 0.0.16.

Version choice: patch

One ### Fixed entry, no new API — RELEASING.md §3 routes that to a patch. The behavior change is that values passed by field name now take effect, which restores the documented contract rather than adding to it.

Also in this PR: a changelog repair

main has been describing a released version as unreleased. My changelog edit in cb22d19 (merged via #37) anchored on the ## [0.0.16] heading and failed to put it back, so 0.0.16's entries silently collapsed into [Unreleased]. My mistake, caught while cutting this release.

Restored here. The [0.0.16] section is byte-identical to its state at 3bed558 (verified by diffing the extracted block) — only the heading was missing, no entries were lost or reworded.

Structure is now:

## [Unreleased]
## [0.0.17] - 2026-08-05     <- this release
## [0.0.16] - 2026-08-04     <- restored
## [0.0.15] - 2026-07-31

Worth noting the failure mode: nothing catches a missing release heading. A CI check asserting version.py has a matching dated ## [X.Y.Z] heading would have caught it at PR time — happy to add that separately if useful.

Release notes

[0.0.17] - 2026-08-05

Fixed

• DatabaseConfig ignored values passed by field name (jvspatial/api/config_groups.py).
  Its aliased fields (dynamodb_*, postgres_*) carry a validation_alias, and a
  pydantic v2 model accepts an aliased field only by its alias unless it opts into
  populate_by_name — which AuthConfig does and DatabaseConfig did not. So every
  such value passed by field name was silently discarded: both
  server_config_overrides_from_env(), which keys its database group by field name,
  and embedding hosts constructing a DatabaseConfig directly. Nothing raised,
  because the adapters then read the same settings from env themselves — the
  configuration object was simply never the source of truth it appeared to be.
  DynamoDB was affected the whole time; Postgres inherited it in 0.0.16. Coverage:
  tests/api/test_database_config_population.py.

---

Pre-merge checklist (RELEASING.md §2)

• pre-commit run --all-files — all 8 hooks pass.
• git log origin/main..HEAD reviewed — the only change since v0.0.16 is Accept DatabaseConfig values by field name, not only by env alias #37, represented above.
• pytest — the fix's own suite plus the config/configurator/env-adapter suites are green; Accept DatabaseConfig values by field name, not only by env alias #37 passed full CI on 3.10/3.11/3.12 with this exact code. CI re-runs it here.
• mypy jvspatial/ — unchanged 86 pre-existing errors, identical on main (see Release 0.0.16 #36 for detail; that checklist step cannot currently pass).

After merge

The publish workflow tags v0.0.17 and uploads to PyPI via Trusted Publishing; §8 (cut the GitHub release from the tag) is still manual.

Downstream, jvagent has a branch waiting on this to thread its Postgres settings through ServerConfig — that work is written and tested but inert until this ships.

[github-actions]

Benchmark comparison

Threshold: ±25% (informational, does not block merge)

benchmark	baseline (s)	current (s)	delta	status
tests/benchmarks/test_deferred_save_benchmarks.py::test_bench_deferred_save_batched_100	0.038324	0.046776	+22.1%	OK
tests/benchmarks/test_deferred_save_benchmarks.py::test_bench_immediate_save_100	0.037020	0.045930	+24.1%	OK
tests/benchmarks/test_jsondb_benchmarks.py::test_bench_jsondb_batched_saves_500	0.411395	0.449098	+9.2%	OK
tests/benchmarks/test_jsondb_benchmarks.py::test_bench_jsondb_count_empty_query	0.846883	0.972887	+14.9%	OK
tests/benchmarks/test_jsondb_benchmarks.py::test_bench_jsondb_count_filtered	1.012687	1.077511	+6.4%	OK
tests/benchmarks/test_jsondb_benchmarks.py::test_bench_jsondb_find_filtered	0.758830	0.831176	+9.5%	OK
tests/benchmarks/test_jsondb_benchmarks.py::test_bench_jsondb_save_throughput	0.001582	0.001611	+1.8%	OK
tests/benchmarks/test_sqlite_benchmarks.py::test_bench_sqlite_count_empty	0.246569	0.317243	+28.7%	REGRESSION (+28.7%)
tests/benchmarks/test_sqlite_benchmarks.py::test_bench_sqlite_count_fallback_via_regex	0.282146	0.359316	+27.4%	REGRESSION (+27.4%)
tests/benchmarks/test_sqlite_benchmarks.py::test_bench_sqlite_count_pushdown	0.287517	0.328074	+14.1%	OK
tests/benchmarks/test_sqlite_benchmarks.py::test_bench_sqlite_find_fallback_via_regex	0.282463	0.355943	+26.0%	REGRESSION (+26.0%)
tests/benchmarks/test_sqlite_benchmarks.py::test_bench_sqlite_find_pushdown	0.239367	0.335863	+40.3%	REGRESSION (+40.3%)
tests/benchmarks/test_sqlite_benchmarks.py::test_bench_sqlite_sort_limit_pushdown	0.281429	0.381406	+35.5%	REGRESSION (+35.5%)