Diagrams in Appendix 1

[taradinoc]

Appendix 1, "A Look Inside", contains many ASCII-art diagrams showing the internal representation of MDL objects. How should we transcribe these?

The diagrams mostly consist of table-like shapes for objects, with vertically stacked rows representing machine words, and each row optionally divided into two cells (the halves of each word). The divisions between rows are sometimes solid and sometimes dashed; the side borders are also sometimes embellished to show omitted rows.

Besides the objects, the diagrams also include arrows representing pointers. Sometimes the arrows link two or more objects through their fields (an arrow begins inside one object's cell and points to the exterior of another object, which has its own cell with an an arrow pointing to another object); sometimes the arrows link one row in an object to a previous row (representing a chain of frames on the stack).

Transcribing the ASCII art directly would be historically accurate, but wouldn't be very accessible.

Is this something that could be handled by a semantic diagramming language like Graphviz?

[OrangeToucan]

I had always imagined they'd be transcribed as-is using ASCII art. I've counted the width of some of them and they fit within the 70 character width. I suppose it could be done in Graphviz; I'd never thought about it. What accessibility issues exist and what makes graphs from Graphviz more accessible?

[taradinoc]

I'm guessing the ASCII art diagrams aren't very useful for blind readers. Graphviz should be better in theory, since it has semantic structure that a renderer can represent non-visually, although I don't know how much better it is in practice. Probably ought to consult a blind reader about that.

[OrangeToucan]

Asking may be best but I don't know any blind people myself.