Making a PDF that's larger than Germany

Original Article ↗ Hacker News Discussion ↗

A playful experiment in creating a PDF page “larger than Germany” prompts a deep dive into how the PDF format encodes physical page dimensions, where its real limits come from, and how those differ from constraints in viewers like Acrobat or Preview. Commenters branch into related technical details—such as hand-crafting minimal PDFs, cross-reference table quirks, security concerns, and tools for manipulating or sanitizing PDFs—while also comparing PDF’s design tradeoffs with formats like HTML, PNG, and archival variants like PDF/A and PDF/X. The thread mixes math corrections and esoterica from the PDF spec with broader reflections on maps at 1:1 scale, the philosophy of representation, and the practical hazards of trying to print or exploit such extreme documents.

Overall reception

  • Many readers found the “giant PDF” experiment delightfully nerdy and entertaining.
  • Several linked related videos and personal experiments with unusual PDFs (e.g., embedded games, Game of Life in PostScript/PDF).

Maps, scale, and literary parallels

  • Multiple commenters connected the “page larger than Germany” idea to classic “map vs territory” jokes and literature: maps at 1:1 scale, blank maps, and thought experiments about mapping empires.
  • Some extended the joke to DIN paper sizes and imagined A-series formats large enough to cover countries or the universe.

PDF technical details and limits

  • Key clarification: the much-cited “381 km” limit is an Adobe Acrobat implementation limit, not an inherent PDF format limit.
  • Several commenters dug into the spec: a page “MediaBox” is just four numbers represented as strings, with no explicit size limit; practical limits come from readers’ numeric ranges (e.g., IEEE 754).
  • Discussion of UserUnit scaling and unit conversion mistakes; multiple math errors in the article were spotted and corrected (million vs billion inches, 1/72" conversions, Earth–Moon distances).
  • Minimal “Hello World” PDF examples and commentary on how text, fonts, and xref tables work, including how many readers tolerate malformed xref data.
  • Surprise by some that PDFs are largely text-based, with embedded binary streams and optional compression.

Security, malware, and tooling

  • Concern about PDFs as an attack surface: embedded active content vs crafted files exploiting font/rendering bugs.
  • Suggestions included specialized analysis tools, using cloud rendering, hardened readers, and strong sandboxing.
  • Debate over how “executable” PDFs are, especially with JavaScript and past exploit history.

Practical issues: printing and viewers

  • Warnings and jokes about “print bombs,” with several stories about print queues being nearly impossible to cancel and printers stubbornly finishing huge jobs.
  • Observations that different viewers disagree wildly on reported page size or rendering (e.g., Chrome vs Preview vs Firefox).

PDF as a format: pros, cons, and alternatives

  • Some praised PDFs (especially constrained profiles like PDF/A or PDF/X) as stable, user-controlled artifacts compared to SaaS documents.
  • Others highlighted the difficulty of extracting structured content, especially paragraphs and tables, arguing that PDF remains fundamentally a print/page description format, not a semantic document format.