1. Data from Jennie Williams' Oceans of Kinfolk dataset https://voyages3-staging.crc.rice.edu/past/database/oceans-of-kinfolk, translated into a Redis Graph database https://github.com/JohnMulligan/voyages_graph

1. will make my data calls like this: https://observablehq.com/@cengique/api-data-example and return labeled nodes and edges

1. the labels will have to be pre-flattened because enslavers and enslaved will have differently-structured biographical data, and we don't want to deal with that in javascript -- and we should assign enslavers and enslaved different colors on the graph (and maybe subdivide enslavers based on roles -- the problem here being that some have multiple roles, so that makes for a lot of individuated groups)

1. we could also pre-calculate the maximum size of the nodes, as we want that non-linear -- some of the buyers are so outsized that they literally block the graph

1. and inputs like this: https://observablehq.com/@observablehq/interactivity-in-observable#range

const scale = d3.scaleOrdinal(d3.schemeCategory10);

return d => scale(d.group);