Dave

> why there is no easy way to save an sf object to disk as geojson file? `sf::st_write()` is the easy way to save an `sf` object to disk as...

Cheers @kylebarron. From reading all the discussions I think the conversion from Arrow to `sf` will be pretty straight forward, so I'm trying to align these "interleaving" functions with what...

> Anyways, I think there's still some good discussion to be had, so I wouldn't spend too much effort implementing any one proposal yet. yeah absolutely. I'm not "productionising" anything...

> How could you encode structure in this interleave form what do you mean by "structure" exactly? > There's no mapping between $data and $start_indices The index of the `[start_indices]`...

oh I see. Yeah I've not fully thought about it yet. Was going to come back to this once `geometries` and `sfheaders` are updated. There's a solution out there somewhere...

> we have to have triangles I should mention, my v0.0.1 plan of `{interleave}` is to create (only) POINTs, LINEs and Triangles (via ear-cutting) from any geometry (hence `{geometries}`), as...

Thanks. I see this is one step towards what I had in mind for https://github.com/dcooley/sfheaders/issues/25 I'm going to leave this as a PR until I finalise how the C++ API...

> Oh it's googlepolylines yeah actually, no, not any more. That _was_ my original way of reducing data size as it's sent to the browser, specifically for `googleway`, but, deck.gl's...

So I'm not fixed to `sf` structures at all, I only need columns of data. But I've focused on `sf` purely because it's widely used at the moment.

Thanks for the use-case. It's useful to see how this would be used. I haven't used it myself, but are you aware of the [bigrquery](https://github.com/r-dbi/bigrquery) library?