Rapid cursor not ideal for long skips

[jpreiss]

In the question-asking period after a presentation, we often need to traverse almost the whole slide deck. This can be a little slow using rapid cursor. For example, a 1-hour talk that uses Beamer transitions might have around 60 slides according to numbering but more like 100 pages in the PDF.

In PPT-like software, presenters sometimes exit to the editor view and scroll to the desired slide. This is also helpful because the numbering in the PPT editor corresponds to the "logical" numbering shown on the slide. For PDFs, we only know the rendered page index and not the logical slide number, so even if we implemented this view (lots of work) it wouldn't be as useful.

In a perfect world we could read logical slide numbers - maybe this is stored as metadata somewhere? If so, we could use the keyboard for instant skipping.

Possibly speeding up rapid cursor is good enough for us.

[jpreiss]

The command pdfinfo -dests gives page numbers for things like theorems and numbered equations. There is probably some way to make Beamer create such a destination for every slide. Then we could parse the output of this command (assuming pdf2image doesn't already have that functionality) to reconstruct the mapping from a logical slide number to a set of page numbers. Also sounds like a lot of work.

[jpreiss]

For now, just speed up rapid traverse and get rid of dissolves so it's easier to see when to stop. in 0d1f368