Background¶

Below are some observations that inspired the design of Pris.

Observations¶

While a declarative source with a theme (such as html with css, or LaTeX) works great for long documents, it is somewhat inflexible for slides. Good slides need to be designed, and this is much more a manual process than typesetting large blocks of text. If you want to be able to change themes with a single click, then you need to shoehorn slides into a pretty limited format. Separating markup from content is not a good idea for creative slides: markup is an important part of the content. Conclusion: Don’t try to separate semantics from markup. Specifying semantics is a non-goal. Offer full control over markup instead.

To design graphics (which is what good slides should be), a graphical editor can be much nicer than a declarative format. Compare Inkscape with LaTeX and TikZ. However, a text-based declarative format has several advantages: it is nicer to work with in source control, and it can easily be scripted. Compare animating a hand-drawn graph in Google Slides with animating a TikZ drawing in LaTeX and Beamer. Manual animation often involves copying the entire thing, after which changing parts becomes difficult. It is possible to build a graphical editor for a text-based format, but these often produce messy output. Conclusion: A declarative text-based input format that is intended to be written by humans is preferred over an opaque format with a graphical editor.

Supporting external graphics (svg logos, jpeg photographs, png graphics) is a must. For consistent graphics, having basic drawing capabilities embedded would be useful (like TikZ in LaTex). The possibility to leverage external tools is limited if consistency is a goal. (For instance, even if the same font is used in externally drawn graphics, the graphic must be scaled to make font sizes match.) Care must be taken to keep the scope narrow: TikZ is a huge project in itself. Conclusion: Allow placing external graphics and expose a minimal but sufficient set of embedded drawing operations.