Known limitations for PDF output:

- when recompiling to PDF, a white background is always implied. Any
  background color configured in hypview.ini will be ignored.

- PDF output currently does not support @xref

- PDF output currently does not display any window title
  at the start of a page (they will occur in the contents view
  however)

- Popup nodes are written just like ordinary pages. Using
  interactive PDF features currently is not supported in most
  viewers (not even Acrobat Reader), and displaying them as
  footnotes would make the calculation of page sizes awfully
  complicated (especially  since they are not limited in size,
  and therefor may span more than a single page)

- Not all non-ascii characters may be displayed properly.
  This is a limitation of the PDF format, and solving it
  would require to embed the whole font into the file.
  This affects all the hebrew character from Atari codepage,
  as well as a few others. Characters that are part of
  of the windows codepage 1252 should work.

- Graphics (images, lines, boxes etc.) that span an output
  page boundary may be truncated.

- The table of contents is generated from an analysis of the
  @toc commands (same algorithm as used in the HypTree view).
  This may or may not work as expected, depending on how the
  input file was organized.

- Conversion of large files may need large amounts of memory,
  since the whole file (including all pages & images) has to be
  kept in memory before it can be written out. For this reason,
  "Save to PDF" is not an option in the TOS viewer. It is supported
  in the command line tool however.
