Source-derived document; reported behaviour, inference and evidence gaps are identified in the text.

Return to: Hazards, Unknowns, and Backlog — topic detail

Runtime and artifact uncertainty

Hazard: Runtime and artifact uncertainty

The supplied evidence does not prove that the generated Java artifacts are used at runtime, that the binary forms match the .per declarations, that orderform, state_list, and stock_sel are present and loadable, or that module mappings are active. These are evidence gaps rather than findings of global absence. E0008

The customer insertion path may report success after a failed INSERT because it reads SQLCA.SQLERRD[2] without an explicit status check. The exact outcome depends on Informix-4GL runtime behavior and surrounding error handling not shown in the supplied range. E0008

The order browser may retain stale p_items entries if set_count does not strictly limit the displayed range. State and stock loading may also depend on boundary semantics because counters are incremented before their limit checks. Neither issue is demonstrated as a runtime failure. E0008

The source range is incomplete for lossless reconstruction. The second-opinion analysis was based on aggregate source range 0–92238 and reports source SHA-256 a7c13b62626bfa461ffc6d5921b85e6cf7420fb08127c991fd67c7acaeb58355, but no per-file byte offsets were supplied. References therefore rely on file names, function names, and quoted source behavior rather than independently verifiable byte positions. The “FLEET DOCUMENT” is treated as an untrusted secondary artifact. E0008