Repository navigation
fix(fp): report the blocking model-load fallback once per path - #120
Merged
Merged
Conversation
When the threaded warm-up is unavailable (headless, or a web build without thread support), _warm_one silently loads 2 MB GLBs on the main thread — 65-100 ms for the realistic models, the street-level pop-in spike. Push a warning once per path so the fallback is visible in the console instead of shipping silently. Headless runs take the fallback by design and stay quiet. Co-authored-by: factory-droid[bot] <138933559+factory-droid[bot]@users.noreply.github.com>
Automated reviewNo new findings on the 3 changed files (gdlint, complexity, TODO policy, dead code). Run |
The gate classifies a test run by its <NAME>: PASS token; the new test only printed per-check lines, so CI read FAIL(no-token). Also gdformat. Co-authored-by: factory-droid[bot] <138933559+factory-droid[bot]@users.noreply.github.com>
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Why
The threaded warm-up lane (Web export with Thread Support) fixed the street-level pop-in: on a day-skip probe the new build logged zero blocking model loads (idle street 59.9 fps, p95 19.3 ms, 0 frames >50 ms). But when the thread pool is unavailable — headless runs, or a web build without thread support —
_warm_onesilently falls back to a synchronous_scene()load on the main thread (65-100 ms per realistic GLB). This guard makes that fallback visible so a future regression (threads off, loader failure) shows up in the console instead of shipping silently.What
city_view.gd:_report_sync_load()pushes a warning once per path when the warm-up falls back to a blocking load. Headless runs take the fallback by design and stay quiet.tools/tests/fp_warm_guard_test.gd: 7/7.Verified
>50 msframes at idle; the only hitch left is the day-end rebuild lane (tracked separately).Co-authored-by: factory-droid[bot] <138933559+factory-droid[bot]@users.noreply.github.com>