Google Scholar PDF Reader

Original Article ↗ Hacker News Discussion ↗

Reactions to the new Google Scholar PDF Reader

  • Appreciation for:
    • Inline, clickable references and hover previews of citations.
    • Simple, clean layout and multiple light/dark/night modes.
    • Tight integration with Chrome and Google Scholar workflows.
  • Skepticism about:
    • Performance of a JavaScript-based reader, especially on weaker machines.
    • Replacing native/OS PDF readers that users prefer (e.g., zathura, Sioyek, Sumatra, Okular).
    • Possible monetization, tracking, and use of reading behavior as training data.
    • Closed, Chrome-extension–only ecosystem and impact on embedded PDFs in other sites.

Concerns about Google Scholar and Google Reliability

  • Several users report Scholar search quality degrading (papers with many citations not appearing for obvious queries; results feeling personalized).
  • Disagreement over whether “cited by” and related-links briefly disappeared from Scholar; some insist it happened (possibly A/B testing), others never saw it. Outcome: unclear.
  • Broader distrust that Google will maintain this feature long term, referencing other discontinued products and fear of future “rug pulls” on notes/annotations.

Alternative Tools and Ecosystem

  • Local readers: frequent mentions of SumatraPDF, Sioyek, Okular, Firefox’s PDF.js, Drawboard, Skim, Preview, PDF Expert, Goodnotes, Notability, Readwise Reader, Readdle Documents, GoodReader.
  • Reference managers / platforms: Zotero (including v7 beta, plugins, sync/storage pain points), Mendeley (mixed feelings after feature removals), Paperpile, Semantic Scholar, scite, OpenAlex, various specialized apps (KnowledgeGarden, Cahier, Scholars, Synthical).
  • Some report rich personal setups combining file systems, Zotero, SumatraPDF, scripting tools, and Emacs pdf-tools/org-ref.

Reading Experience, Formats, and AI Wishes

  • Many still prefer native readers or even printed paper for deep work; others advocate large monitors, tablets, or e-ink devices.
  • Debate over PDFs vs HTML: PDFs seen as stable for archiving; HTML lauded for machine-readability and richer, semantic, interactive documents, but current tooling is imperfect.
  • Desired “next step” features include AI-based summarization, contextualization within a field, critique of methods, structured outlines, and plain-language rephrasing, going beyond mere rendering improvements.