You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Complete the first Wright-owned OSTW -> Workshop pipeline for the semantic surface established by #118 and validate it against explicit-root pinned OSTW v3.4.0 reference evidence established by #122, without requiring output-text identity.
Context
#117/#118 establish Wright's native OSTW project/syntax/semantic frontend and frontend-neutral HIR boundary.
The original version of this issue assumed #115 proved the protect-ban entry project was accepted and produced a stable 79-element Workshop artifact. #118's pinned P1 probe disproved that interpretation: the upstream language server compiles the last-opened document plus its transitive import closure, and the historical 79-element result belongs to utils/ServerLoad.del, not the ds.toml.entry_point graph. The entry-root graph currently rejects on three missing ../OSTWUtils/... imports.
#122 therefore owns the corrected explicit-root oracle baseline and must establish the accepted differential target(s) used here. protect-ban remains valuable real-world syntax/semantic evidence, but its incomplete entry graph is not itself an accepted compile-success oracle until its licensed dependency closure exists.
Add wright compile support for the declared OSTW source/project boundary through the same driver/session contract as OPY and raw Workshop.
Preserve observable semantics covered by [M13] Resolve and lower the accepted protect-ban OSTW semantic slice into HIR #118 and its pinned probes, including variable allocation, user enums, functions/defaults, rule priority/order, control flow, expressions, named/default Workshop calls, and player-variable access where supported.
Produce valid en-US Workshop through the existing Wright Workshop emitter; do not create an OSTW-specific Workshop backend.
Consume canonical Wright Workshop catalog/signature metadata; do not reintroduce an OSTW-owned duplicate Workshop game-content table.
The resulting Workshop is semantically equivalent under Wright's declared compatibility contract to the corresponding pinned v3.4.0 explicit-root reference result.
Parent: #90
Depends on: #118, #122
Goal
Complete the first Wright-owned OSTW -> Workshop pipeline for the semantic surface established by #118 and validate it against explicit-root pinned OSTW v3.4.0 reference evidence established by #122, without requiring output-text identity.
Context
#117/#118 establish Wright's native OSTW project/syntax/semantic frontend and frontend-neutral HIR boundary.
The original version of this issue assumed #115 proved the
protect-banentry project was accepted and produced a stable 79-element Workshop artifact. #118's pinned P1 probe disproved that interpretation: the upstream language server compiles the last-opened document plus its transitive import closure, and the historical 79-element result belongs toutils/ServerLoad.del, not theds.toml.entry_pointgraph. The entry-root graph currently rejects on three missing../OSTWUtils/...imports.#122 therefore owns the corrected explicit-root oracle baseline and must establish the accepted differential target(s) used here.
protect-banremains valuable real-world syntax/semantic evidence, but its incomplete entry graph is not itself an accepted compile-success oracle until its licensed dependency closure exists.Scope
wright compilesupport for the declared OSTW source/project boundary through the same driver/session contract as OPY and raw Workshop.Non-goals
Acceptance criteria
wright compilecompiles the explicit-root accepted OSTW target(s) established by [M13 correctness] Rebaseline OSTW corpus oracle around explicit compile roots #122 using only Wright production binaries/libraries.Relationships