Better support for reveal.js functionality

[mattilyra]

I really like creating slide presentation using Jupyter Notebooks, but the workflow is currently fairly cumbersome. I often find myself wanting to control HTML <section> tag parameters like

• slide transitions (per individual slide)
• individual slide backgrounds
• background transitions
• different fragment types (http://slides.com/hakim/reveal-js/embed?transition=none#/transitions)
  • grow
  • shrink
  • fade in/out
  • highlight colour
• footer ...

It would be great if controls for these could be added to the slide toolbar so that one does not need to edit the HTML file manually each time after running nbconvert --to slides.

[damianavila]

This is great feedback, but can you re-open this issue on the nbconvert repo wheres the static slides are implemented? Thanks!

[mattilyra]

Moved to nbconvert https://github.com/jupyter/nbconvert/issues/612

[mpacer]

Ha! I was just about to suggest that this belonged here because it was about changing the slide toolbar which I thought was implemented as part of the notebook interface. So I actually think this should be three issues, only one of which needs to be on both repos

1. In the notebook for creating the ui to incorporate these kinds of features.
2. coming up with a convention for storing this info in the notebook json.
3. Getting nbconvert to respect this convention & the metadata values to have them represented correctly in the converted output.

[mattilyra]

Riiight, I'm happy to help with this but I'm a little out of my depth as this is the first PR to Jupyter I've done. What did you want to file where exactly?

[mattilyra]

Also would it help if a describe what my workflow is and where the pain points are?

[bouzidanas]

Ha! I was just about to suggest that this belonged here because it was about changing the slide toolbar which I thought was implemented as part of the notebook interface. So I actually think this should be three issues, only one of which needs to be on both repos

1. In the notebook for creating the ui to incorporate these kinds of features.
2. coming up with a convention for storing this info in the notebook json.
3. Getting nbconvert to respect this convention & the metadata values to have them represented correctly in the converted output.

Hi, Ive recently been creating webpages and slide presentations out of my jupyter notebooks and created a personal solution for points 2 & 3 and since point 3 relates more so to nbconvert and the issue is still open there as well, Ive submitted it to nbconvert (https://github.com/jupyter/nbconvert/pull/1861#issue-1365317287). It uses very similar mechanics to how tags are handled.

For point 1, I suggest something similar to how tags are handled as well except perhaps limit the data attributes to what the installed version of reveal.js supports.