# Revelei changelog — Alegauss

Every entry in Revelei's roadkeep ledger, dated by the commit that wrote it.

- 2026-10-23 · PB69: Dragging a crop corner or edge mostly does nothing: the press lands beside the handle and moves or pans instead
- 2026-10-23 · PB67: A photo cannot be cropped, so the subject cannot be framed before or after an upscale
- 2026-10-23 · PB66: A photo cannot be mirrored horizontally or vertically
- 2026-10-23 · PB65: A photo cannot be rotated: a sideways scan or a wrong EXIF orientation stays sideways
- 2026-10-23 · PB64: Once a photo is open there is no way to open another: the Open button and the drop zone go with the empty screen
- 2026-10-23 · PB63: The editor has no toolbar: opening, saving, undo and every basic edit live in the side panel or nowhere
- 2026-10-21 · PB62: Remove background says only "Finding the subject…" through a minute of edge matting on a large photo
- 2026-10-21 · PB61: After an undo the run's card still says Done and the status bar still times a result no longer shown
- 2026-10-20 · PB60: A 4x of an 8192 px photo closes the app, and a run's progress sits below a panel taller than the window
- 2026-10-20 · PB59: No UI case drives a first-use download to the end: progress, pause, a failure and the auto-run are unobserved
- 2026-10-20 · PB58: The installer is 628 MB, 84% of it the models of two operations most photos never use
- 2026-10-20 · PB57: The UI cases still work around five winwright gaps that 1.1.4 closed, and the drop zone and Save dialog have no case
- 2026-10-20 · PB56: A tiled operation throws on a GPU whose DXGI driver answers no video memory budget, as VMware's does
- 2026-10-20 · PB55: On --provider winml, 2x upscale, denoise and refined background removal fail, and engines build inside the first run
- 2026-10-20 · PB54: Face restore on DirectML pastes a flat brown square over each face with the default FP16 GFPGAN
- 2026-10-20 · PB53: The installer accepts Windows 10 1809, where Windows ML and so every operation cannot run
- 2026-10-20 · PB52: TensorRT-RTX builds its engine for 10 to 17 s at every session creation, so auto cannot prefer it
- 2026-10-20 · PB51: The installer carries no models, so a fresh install cannot run any operation
- 2026-10-20 · PB50: PhotoBoost's site declares japode.github.io as its origin, which now only redirects to japode.com
- 2026-10-20 · PB49: PhotoBoost's site does not say it is a Japode product or link to the brand's site
- 2026-10-20 · PB48: Background removal's edge stays soft where subject and background have the same brightness
- 2026-10-20 · PB47: Background removal leaves faint specks and a soft upscaled edge, with no refinement at the photo's resolution
- 2026-10-20 · PB46: App has no About page listing each model's license and the third-party notices
- 2026-10-20 · PB45: Exports drop the source photo's camera, date, copyright and GPS metadata, with no choice to keep it
- 2026-10-20 · PB44: Decoding a wide-gamut image to sRGB is untested, so a wrong colour transform would pass the suite
- 2026-10-20 · PB43: App cannot use the TensorRT-RTX or OpenVINO providers that Windows ML installs
- 2026-10-20 · PB42: Cases have no tiny model fixture, so an inference case would need the real model downloads
- 2026-10-20 · PB41: Initial tile size ignores the free video memory of the GPU
- 2026-10-20 · PB40: Unverified how to show a 50 MP image on the GPU inside a WPF window with smooth zoom and pan
- 2026-10-20 · PB39: No command-line flags to open a file, force a provider or point at a models folder
- 2026-10-20 · PB38: No winwright Cases project, winwright.json or first case that opens the main window
- 2026-10-20 · PB36: No denoising operation
- 2026-10-20 · PB35: Release build produces no self-contained Windows installer
- 2026-10-20 · PB34: No table comparing the FP16 and FP32 sets' size, inference time and quality on GPU and CPU
- 2026-10-20 · PB33: No README in Portuguese on running, generating models, Python and hardware requirements
- 2026-10-20 · PB32: No background removal pipeline producing a transparent image
- 2026-10-20 · PB31: No GFPGAN face restoration pipeline
- 2026-10-20 · PB30: No face detection and alignment to feed GFPGAN crops and paste results back
- 2026-10-20 · PB29: No UI choice between the bundled FP16 models and the downloaded FP32 High quality set
- 2026-10-20 · PB28: No operation history with undo and redo
- 2026-10-20 · PB27: No before and after slider comparison
- 2026-10-20 · PB26: Open, upscale and export flow has no winwright case proving it end to end
- 2026-10-20 · PB25: Results cannot be exported as PNG, JPG or WebP
- 2026-10-20 · PB24: No operations panel to run an upscale and pick 2x or 4x
- 2026-10-20 · PB23: No screen to open an image by drag and drop or file picker
- 2026-10-20 · PB13: The app has no button to download the optional FP32 models with progress and hash verification
- 2026-10-20 · PB12: The optional FP32 models have no public location the app can download them from
- 2026-10-19 · PB37: No design system: Brand, Theme, Appearance and the Fluent page styles from claude-tray are missing
- 2026-10-19 · PB22: Inference time and the provider used are not logged per operation
- 2026-10-19 · PB21: An out-of-memory failure aborts the inference instead of retrying with smaller tiles
- 2026-10-19 · PB20: UI receives no tile X of Y progress during inference
- 2026-10-19 · PB19: No Real-ESRGAN upscale pipeline for 2x and 4x
- 2026-10-19 · PB18: Large images cannot be processed: no tiling with overlap
- 2026-10-19 · PB17: No preprocessing to and from model tensors, and no unit tests for it
- 2026-10-19 · PB16: ONNX sessions are not loaded on demand nor released after idle minutes
- 2026-10-19 · PB15: No InferenceService queue that runs one inference at a time
- 2026-10-19 · PB14: Execution provider is neither detected at runtime nor shown in the UI
- 2026-10-19 · PB11: No models/<name>/{fp32.onnx, int8.onnx, meta.json} layout with SHA256 hashes
- 2026-10-19 · PB10: INT8 models are not validated against FP32 on a test image
- 2026-10-19 · PB9: Exported models are not quantized to INT8
- 2026-10-19 · PB8: No scripts/export-and-quantize.py to export each model to ONNX with dynamic axes
- 2026-10-19 · PB7: No scripts/download_models.py to fetch the original weights
- 2026-10-19 · PB6: No verified source URLs for Real-ESRGAN x4plus, GFPGAN v1.4 and RMBG-1.4 weights
- 2026-10-19 · PB5: Unverified which ONNX Runtime package (DirectML, Windows ML, CUDA) runs in a self-contained WPF build
- 2026-10-19 · PB2: No xunit.v3 test projects, analyzer gate or architecture tests in the build
- 2026-10-19 · PB1: No WPF .NET 10 solution with the Fluent ThemeMode that builds and opens a window