2025 Problems - Value added typesetting
- Date: 14 September 2023
- Time: 6:30 to 7:30pm (UK Time)
Jonathan Fine Typesetting is about breaking things, to fit the page. A paragraph is broken into lines, each about the width of the page. The long sequence of lines (called a galley) is broken into pieces, each about the height of a page.
To a first approximation, typesetting a text file consists of inserting line-break symbols into the text. And some of the line-break symbols are immediately followed by a page-break symbol.
Looked at in this way, typesetting is a value-added process. It inserts the line and page breaks. That’s the added value. And if those breaks are removed, we get the original source file back.
I wrote an article in 1996 that described this process. At that time software and hardware to support wide use of such technology was rare (and I certainly didn’t have it). However, today such resources are plentiful, and such a process need no longer be a dream.
In the video I describe the motivation and process. For value-added-typesetting the output pages can be treated as a wrapper that contains the source document. Thus the output pages will support any rendering of the source document, including speech for the blind and partially sighted.
Nelson Beebe listened carefully to my presentation. Here are some potential problems and opportunities he identified, as I see them.
Comments in source documents can be very valuable, and such comments should be preserved.
TeX typesetting does not record word boundaries (starts and ends). This is something value-added typesetting must record. Such is vital for improved accessibility.
There are issues regarding the
\hyphencharin TeX. In particular, it’s just a character in a font, and the same character can appear in ordinary input.
Keeping the input in the output provides roundtripping. But do we get exactly the same document as the one we put in. This is related to the normalisation (or pretty-printing) of text files.
Martin Ruckert’s HINT system provides prior art for much of the technology needed for value-added typesetting. Cooperation with HINT would benefit both sides.
There are issues regarding macros, as in TeX and LaTeX. My response to that is to create just a typesetting program, and ensure that it supports roundtripping for its input.
It is the responsibility of the macro system to ensure that its processing supports roundtripping via the typesetting roundtripping.
For more information about the TeX Hour, including Zoom URL, see the About page.