There are already an overwhelming number of documents like this that we
reference and want to reference in the future, including many bitmap
formats, PostScript, dvi, spreadsheets, formatted documents from various
word processors, etc. Support for PDF should be through the same mechanism
as handling PostScript documents or any other non-HTML document that will
be displayed or rendered by a different application or DLL (in the case of
Mac or Windows) rather than the browser itself. Translation of other
documents to PDF or encapsulation of some sort would consume too much time
and space. A browser should not have to understand PDF.
I discussed some of the methods that could be used to provide hypertext
references to non-HTML material in an earlier message (LinkToLiving). This
needs to be addressed for the existing large body of non-HTML documents
since those documents will always outnumber HTML documents by several
orders of magntitude. PDF is a minor issue in comparison.
ka