jeremyrsmith

This is a tricky one... I think in the Spark case (which I'd estimate is the majority) ~~we'd want to use Hadoop APIs to access S3~~ in order to avoid...

@tmnd1991 absolutely I think a lot of things should be moved to `ZManaged`... I had a branch which moved in that direction, will be slowly trying to pick things from...

This looks good. But thinking about it a little more... I wonder if it makes sense for this to be a global preference rather than per-cell? Would people want to...

> does this really need to be a full persisted preference? Why not just a toggle button in the UI? If it's per-cell, but not persisted, you'd have to change...

@shangqing-li I was thinking of something a little different. Rather than configuring a single cell to use as a template, the configuration would point to an entire _notebook file_ to...

If the template notebook is configured as a filesystem path rather than a "notebook path", then it could live outside of the notebook root and thus be non-editable (at least...

We don't currently support date/time types (see #595). If you use milliseconds instead of seconds, it might work (vega claims to support that)

@AlexDvorak This would go in `SparkReprs`; you'd want to define an instance of `SparkReprsOf[Array[Row]]`. Thanks for taking it on; join us on gitter if you need help!

Welcome! Thanks for the suggestion! I agree that this would be great and it's on the roadmap.

@rabejens Are all these changes backward-compatible with prior Spark versions?