cursorrules: align output-name policy with the ComfyUI frontend - #146
Conversation
The old rule forced English output names in all localized Outputs tables, but the ComfyUI frontend itself localizes semantic output names per locale (zh: 正面条件/负面条件, es: positivo/negativo) while keeping technical identifiers English (task_id, GLB, part_names, mask). New policy: mirror the frontend exactly, per output, per language, using docs-generation/data/node_translations.json (outputs.<index>.name) as the source of truth. Outputs without a frontend entry keep the English name. Motivated by PR Comfy-Org#145 review friction: the drift-repair retranslation had aligned docs to the frontend, and a reviewer applying the old rule reverted that alignment, de-syncing docs from what users see in the UI.
|
No actionable comments were generated in the recent review. 🎉 ℹ️ Recent review info⚙️ Run configurationConfiguration used: Organization UI Review profile: ASSERTIVE Plan: Advanced Run ID: 📒 Files selected for processing (1)
Included review availability: Your plan provides up to 1 included review per hour; 0 remain after this review. 📝 WalkthroughWalkthroughThe output naming rule now follows ComfyUI frontend translations by language. Semantic outputs can be localized, while technical identifiers remain in English. English remains the fallback when no translation exists. ChangesOutput localization guidance
Priority: ⬇️ Low Merge Risk: ⚪ Minimal · up to This guidance update aligns output naming with frontend translations and introduces no established production risk. 🚥 Pre-merge checks | ✅ 2✅ Passed checks (2 passed)
✨ Finishing Touches🧪 Generate unit tests (beta)
✨ Simplify code
Comment |
Summary
Replaces the blanket "output names stay English" rule with a frontend-first policy.
Why
The ComfyUI frontend itself localizes output names per locale. Evidence from the frontend translation export (
docs-generation/data/node_translations.json):正面条件/负面条件/Latentpositivo/negativo/latente模型 task_id/GLB/FBXtask_id del modelo/GLB/FBX分割 task_id/part_names/parts/masktask_id de segmentación/part_names/parts/maskSo "output names are programmatic identifiers" is only true for technical identifiers (
task_id,GLB,part_names,parts,mask,FBX). Semantic outputs (positive/negative-style) are localized in the UI, and docs should match what users actually see.Recent translation runs had aligned docs to the frontend; PR review applying the old English-only rule reverted that alignment, de-syncing docs from the UI.
New policy
Mirror the frontend exactly, per output, per language, using
node_translations.json(outputs.<index>.name) as the source of truth. Outputs without a frontend entry keep the English name. Data types stay English (unchanged).