I built an online PDF management platform using open-source software
Platform & Features
- Online hub for PDF management: conversions (PDF→Word, CSV, PDF/A), OCR, translation, web-to-PDF, markdown-to-PDF, etc.
- Users generally find it useful and bookmark-worthy, though design is seen as somewhat unpolished.
- Some users report issues: mobile Safari UI overlap, timeouts, and at least one bad PDF→markdown result.
Implementation & Open-Source Stack
- Front end: Next.js + TypeScript; landing page in Astro.
- Back end: primarily Python + Flask, with JavaScript for web/markdown-to-PDF.
- Core tools mentioned: LibreOffice, Ghostscript, Tesseract, PyPDF2, Puppeteer.
- Multiple requests for clearer attribution and a dedicated page listing all open-source components and licenses.
- Concern raised about Ghostscript’s AGPL for any future commercial use.
- Creator plans to open-source the backend, but not all code.
Privacy, Security & Trust
- Many commenters will not upload sensitive or corporate documents to any cloud service, regardless of deletion promises.
- Creator states files are stored as temporary files and deleted immediately after processing via a Flask hook.
- Others note this promise is not clearly reflected in the privacy policy and worry about future monetization of document content and third‑party integrations.
- Suggestions to move more processing to WebAssembly in-browser and/or provide offline/self-hostable options.
Alternatives & Self-Hosted Options
- Strong interest in offline and self-hosted tools; Stirling‑PDF is repeatedly recommended as the “gold standard” with Docker support.
- Other tools mentioned for local or browser-only work: pandoc, ocrmypdf, LibreOffice, pdftk, PyPDF2, pdftool.org, SimplePDF.eu, marker (PDF→Markdown), Paperless NGX, sist2, WeasyPrint.
- Some distrust codebases initially bootstrapped with ChatGPT, countered by arguments that open source allows auditing.
Missing Capabilities & Broader Reflections
- Users miss full PDF editing (adding/editing text, inserting images/signatures) comparable to Acrobat Pro.
- Commenters note PDF’s complexity and the lack of a comprehensive open-source desktop alternative; most tools cover only subsets.
- Some argue a new, simpler document format is overdue; others want semantic/AI search over large personal PDF collections.