PDF is bad for the Web

I suspect I am not alone in having adopted Portable Document Format (pdf) as an easy and convenient way of publishing documents on the web, especially those that have been initially created in a word processor such as MS Word. The ease with which such documents can be translated to pdf and the resulting file uploded to (and downloaded from) a website are hard to beat. There is also the added advantage that the downloaded document can be almost guaranteed to print identically to the original.

Well, if distributing documents for printing is the aim, then pdf is ideal. But if making the content available to people to read on the Web, then it sucks.

Why? Several reasons come to mind:

1. PDF documents intended for printing will usually be in portrait orientation, whilst most computer screens are landscape
2. Printing font sizes are generally smaller than those commonly deployed on the web, so legibility may be a problem
3. it is unlikely that the 'look and feel' of the pdf document will match that of the website
4. PDF documents of more than a handful of pages will be very large & will take up bandwidth and load slowly
5. The server your site is located on may not be able to search pdf documents

Adobe was cleverly able to position Acrobat (pdf) as a nifty way of presenting documents on the web at a time when there were browser wars and before the Web Standards Project and CSS got underway. Now the challenge is to get word processor vendors to use XML creatively to describe document structure in an exportable fashion, allowing the serving site's style sheets to determine how that structure should be presented. Given that MSWord file structure is XML-based, this should not be technically difficult. Political will is another matter.

(Since writing this my attention has been drawn, via Lockergnome Bytes to an article by Jakob Nielsen entitled PDF: Unfit for human consumption which makes many of the same points.)

Posted by Chris Bunch at July 16, 2003 11:13 PM