PDF vs HTML

Original version e-mailed to Electronic Publication Implementation Committee (EPIC), Entomological Society of America (ESA), 22 July 1997.

Thomas J. Walker, 25 August 1997.

The question was posed as to whether HTML or PDF would be the more useful format for ESA's electronic reprints. The short answer is that it depends on whether the articles are to be viewed and used mainly on the computer screen or mainly as hardcopy.

Here is a brief comparison of the two:

PDF, Adobe Acrobat's Portable Document Format, retains the exact appearance of a document, no matter what platform (PC, Mac, Unix) is used to view or print it.

HTML is the language of the Web. HTML documents are thus designed for on-screen viewing and interactivity. They are intended to be accessed via computer whenever needed rather than to be printed to simulate a traditional reprint. Making HTML versions of traditionally printed articles is more expensive than making PDF versions and managing the files that are involved is more complex.

Here is a more detailed comparison:

• Cost to produce
  PDF: Very low, especially if the files used to produce plates for printing are available.
  HTML: Modest to much, in proportion to the attention given to interactivity and on-screen appearance. Lots of good hyperlinks and lots of thumbnail and enlarged images require lots of knowledge and labor.
• File structure
  PDF: Each article is a single binary file, usually between 100 KB and a few MB in size. Files are small compared to WinWord files of the same documents.
  HTML: The document and figures are separate files. A richly illustrated article with both thumb-nail and full screen versions of each figure could require 10-20 files.
• On-screen appearance
  PDF: Retains the exact appearance of the original article. Screen images, or selected portions, can be zoomed in or out from 12 to 800%.
  HTML: Appearance determined by mark-up tags, the browser, the monitor, and the settings of the browser. Thus an HTML document may look great on the state-of-the-art system used to create it but be debased by the software and/or hardware used to view it.
• Interactivity
  PDF: Some possible, but costs more.
  HTML: In addition to internal and external hyperlinks, HTML permits a variety of other enhancements over traditional formats. Among these are audio and video clips, and forms encouraging viewers to submit their corrections and critiques.
• Printed appearance
  PDF: Like a good photocopy of the original article.
  HTML: Like the document appears when viewed on screen, including thumbnails (if used) rather than full-size figures. (The full-sized figures can be printed in separate printing operations). If color is important in the on-screen image, a color printer is needed.
• Math symbols
  PDF: Renders math symbols seamlessly.
  HTML: Does not support math symbols, except as auxiliary bit-mapped files.
• Searchability
  PDF: While viewed in the Acrobat reader, a PDF-formatted article can be searched for any word or phrase, permitting users to find passages of interest. Search-service robots, such as those of AltaVista, do not retrieve and index PDF files. Commercial software can index and search all PDF files on a single server.
  HTML: Search-service robots retrieve and index Web-published HTML files when requested to do so. Once indexed, articles of interest can be retrieved by free boolean searches of the search service's index. Free software can index and search all HTML files on a single server..
• Chief advantages
  PDF: Low cost; produces images and hardcopy that researchers know and like.
  HTML: Interactivity (e.g., forms, internal and external hyperlinks); on-screen appearance; audio and video clips easily incorporated.
• Chief disadvantages
  PDF: Not easily made interactive.
  HTML: Hardcopy lacks interactivity (links and thumbnails are dead; format not designed for print). Hardcopy is less compact (more pages) and harder to read (no columns). Difficult for user to save the article in electronic form for later viewing (many files, sometimes requiring a particular directory structure).