Known Limitations

Known Limitations

Area	Details
Windows only	Officially supported on Windows 10/11 only. Word/Excel/PowerPoint COM automation, registry-based dark mode detection, and PyInstaller packaging are Windows-first. macOS/Linux are not officially supported and are planned to come later as separate software.
Build size	~225 MB compiled, inherent to Python + PySide6 + Matplotlib + pdf2docx (which pulls OpenCV) and other libraries. Trims: SFX sounds shipped as OGG, Qt translations kept for the 10 most widely spoken languages only, orphaned Qt DLLs removed.
PDF → DOCX quality	Depends on the source PDF. Scanned PDFs with no embedded text layer are only partially recoverable without an OCR tool. The native fallback preserves text, tables, and embedded images; Word COM and LibreOffice are tried first when available.
PDF → DOCX page size	Word COM refuses PDFs whose pages exceed Word's maximum supported page size (~22 in). Such files automatically fall through to LibreOffice or the native fitz fallback.
PDF OCR	OCR is currently limited to image → PDF conversion. There is no PDF → Word OCR stage; scanned or image-only PDFs are reported by text-only mode instead of being OCR'd automatically.
Forced Quick Convert cancellation	Quick Convert uses cooperative cancellation first. If an engine remains stuck beyond the grace period, the context-menu process can force-exit, which may leave a temporary/staging file or an external Word/LibreOffice process.
Settings dialog scale-in	On Windows, the Settings dialog briefly scales in when opened. Qt geometry remains fixed at 650×550; a temporary DWM transition-disable test did not remove the effect and was reverted. This is a presentation issue only.
.ppt legacy format	Requires Microsoft Office or LibreOffice. python-pptx cannot open old binary .ppt files natively.
Manual distribution	Not on the Microsoft Store. Distributed via GitHub, Official WebSite file-converter and Itch.io.
Solo development	Progression depends on availability outside of studies.