Changes to make after tutorial

[jbednar]

• Merge columnar data into containers tutorial.
• Then the gridded data no longer needs to be omitted, as it's just a tiny amount of new material afterwards
• Add a section explaining what can be accepted as data for an Element
• Simplify the Parameters and Widgets page now that the concepts are introduced earlier, or at least define a better map. Indicate that this section will take less time
• Make initial taxi example match the one in the final tutorial (hours, not passengers)
• Fix WMTS string spec and reflect it back into the materials
• Adjust timings -- more on earlier, less on later
• More and more interesting exercises
• Improve or remove the annotations2 example
• Demos should include holomap/facets
• Annotating your data -- tell story using dataframes instead
• Custom Interactivity could use some more compelling examples
• Eliminate view parameters from deploying tutorial
• Example of writing your own element, not wrapping new JS but putting together existing glyphs, etc. into a new higher-level object.
• Fix outdated example that uses .csv

We suggested people file issues on hv about:

• Adding annotations in screen coordinates so that they don't disappear on zooming, either by actual implementation or by examples of working around it.

[jbednar]

Also fix the data files: they are cutting off LaGuardia, which makes them not useful for telling stories about the pickups and dropoffs.

[jbednar]

I've marked things as complete above if they are done either here or in the separate anacondaviz repository.