Stability¶
This page defines the current stability expectations for the repository.
It exists to separate:
- stable targets the project is actively preparing for long-term support
- bounded support that exists but is not on the path to parity before
1.0.0 - explicitly unsupported areas
For the current 0.3.x line, stability work is anchored to the verified Java 21 baseline, the
verified GraalVM 25.x runtime lane, and the currently verified Spring Boot 4.0.4 starter/sample
line.
Stability Matrix¶
| Area | Status | Notes |
|---|---|---|
Python runtime binding via PyExecutor |
target for stabilization | Primary path for the 0.3.x to 1.0.0 journey |
| Spring Boot starter on the Python-first path | target for stabilization | Verified on the current Boot 4.0.4 line; includes fail-fast, preload, warmup, actuator, and metrics behavior |
| Python-oriented code generation | target for stabilization | Limited to the documented supported parser subset |
JavaScript runtime binding via JsExecutor |
experimental | Supported as a bounded runtime path, not a parity target before 1.0.0 |
| Public API in non-experimental documented Python-first paths | stability-disciplined | Changes should be narrow, reviewed, and reflected in docs and tests |
| JavaScript code generation | unsupported | Not implemented in the current release line |
| Hot reload / live source replacement | unsupported | Not part of the current runtime contract |
| Multi-version source isolation | unsupported | Not part of the current runtime contract |
| Throughput-oriented shared-context parallelism | unsupported | Current model is safe-by-serialization |
What Experimental Means Here¶
For this repository, experimental means:
- the capability exists and is tested at a bounded level
- bugs in the documented behavior should still be fixed
- the surface is not a parity target before
1.0.0 - the contract may still be narrowed, clarified, or explicitly limited before stabilization
- support claims for that surface should stay narrower than the Python-first path
1.0.0 Direction¶
The intended path to 1.0.0 is:
- stabilize the Python-first runtime contract
- stabilize the Spring Boot behavior around that contract
- stabilize the Python-oriented build-time/codegen workflow
- keep JavaScript explicitly bounded unless it later earns a broader support claim
The project should not expand scope aggressively while those steps remain unfinished.