APPEARANCE AND STRUCTURE

From: LLOYD RASMUSSEN (RASMUSSE@mail.loc.gov)
Date: Wed Apr 13 1994 - 02:03:09 PDT


            The whole PDF business is focusing our attention on the
          definition of accessibility. We need to work this out, and we
          will continue to work it out as neeeds and technologies change.
            If I try to do a braille translation of a document, it might
          come in three different forms: a print image file, a WordPerfect
          file that is properly coded so that as you scale its font size up
          or down, the basic layout does not change; or a tagged file, such
          as SGML or .MEG MegaDots file, where everything has a tag showing
          where it belongs in the document's structure or outline.
            Many aspects of proper braille formatting can be gleaned from
          the appearance of a document, as represented in PDF or in coded
          WordPerfect files. But too many people treat their word
          processor like a typewriter, and don't know how to code something
          so that it is scalable, or even see the need for scalability.
          Even in a properly coded WP file, there must be a combination of
          automated and manual methods to determine how best to restructure
          the document for Braille.
            Until we start working with style-based Braille translation
          systems like MegaDots and the new Duxtury, we have a hard time
          even explaining the advantages of putting structure into
          documents, and not just appearance.
            You need to hear Dr. Raman's system for reading equations in
          synthetic speech. It takes advantage of the structure of a TeX
          document to "display" the different levels of an equation with
          different degrees of volume, left-right placement, head size, and
          other characteristics so you can hear what is a superscript, the
          numerator and denominator of a complex fraction, nested
          parentheses, etc. It would have to be under the user's control,
          and not recorded on a tape, so the reader can ask the machine for
          the level of detail desired at a particular time.
          In summary, there are times when it is vital to know the
          appearance of a document. But there are going to be other times
          when the structure is more important than the appearance,
          especially if we want to do more than a passing job of
          translating a document into speech, braille or large print.
            Lloyd Rasmussen
          rasmusse@mail.loc.gov
            l.rasmussen2@genie.geis.com



This archive was generated by hypermail 2b29 : Sun Dec 02 2012 - 01:30:03 PST