At the time, typesetting and physically cut and pasting of images was still required to assemble many pages because the specialized composition, pagination, text formatting, and graphic design systems that produced the content could not operate together to produce integrated output.
PIL describes the layout, and allows the use of any combination of markup languages and image formats to encode the content.
It is useful for those who want to define a specific visual presentation rather than the sort of fluid layout that a web browser allows.
It defines a (theoretically infinite) hierarchy of canvases with coordinate systems, tags, frames, and content of any type.
The complete public domain distribution of PIL includes the language specification document (including a BNF specification, example files, a programmer's guide, and C-language source code for a parser and an output engine to produce PIL.