{
  "id": 3119804,
  "title": "What a semantic patch can honestly prove about WebAssembly output",
  "url": "https://urgent.news/2026/08/24/what-a-semantic-patch-can-honestly-prove-about-webassembly-output",
  "topic": "tech",
  "section": "Tech",
  "published": "2026-08-24T21:19:22.000Z",
  "source": {
    "name": "Dev.to",
    "slug": "dev-to",
    "url": "https://dev.to/wavect/what-a-semantic-patch-can-honestly-prove-about-webassembly-output-pbi"
  },
  "original_language": "en",
  "account": "SEMAPRAX is an Apache-2.0 agent-native systems programming language developed at Wavect GmbH. It aims to provide deterministic compiler output for semantic changes in systems programs. One of its key features is the ability to verify source snapshots and apply semantic patches independently, producing several deterministic compiler-owned projections such as semantic Graph JSON, an explicit capability manifest, Native C11 source, and a structurally validated WebAssembly Core module. The report records domain-separated digests and byte lengths for each projection, classifying them as changed or unchanged. This approach ensures that documentation-level changes, capability changes, and runtime behavior changes are not conflated. Deterministic output is crucial for reproducibility, as it allows for independent verification of the relationship between semantic changes and machine-facing artifacts. The target report uses domain separation to differentiate digests of semantic graphs from digests of Wasm modules. SEMAPRAX currently supports a narrow WebAssembly lane using the profile semaprax.wasm-core.v1 and structurally checks the emitted module using wasmparser 0.256.0. The tool is not production-ready and has limited WebAssembly support, focusing on bounded scalar exports and generated JavaScript/TypeScript bindings. To ensure evidence becomes more than just static facts, SEMAPRAX proposes a mechanism where a target report can be wrapped in an additive Evidence v2 capsule tied to the semantic patch. This evidence-gated apply route requires independent replay, reconstruction of the target report, and checking the binding before committing the change. This approach separates meaning, evidence, and authority, providing a replayable statement of the semantic change that is narrower, more precise, and easier to falsify.",
  "summary": "When a coding agent changes a systems program, a source diff is only the beginning of the question. The more useful question is: what exact machine-facing artifacts would this semantic change produce, and can another process independently verify that relationship? That is one of the research problems we are exploring in SEMAPRAX , an Apache-2.0 agent-native systems programming language built at…",
  "key_points": [],
  "editors_take": null,
  "illustration": null,
  "coverage": {
    "outlets": 1,
    "also_reported_by": []
  },
  "ai_generated": true,
  "disclaimer": "Summaries, key points and the editor’s take are written by software from other outlets’ reporting and may contain errors — always check the linked original."
}