@wnd it's (rounded) multiples of https://h3geo.org/docs/core-library/restable/#edge-lengths

which are messy because they're hexagons and hexagons don't quite tile spherically, so they have to bodge it with a few pentagons (i think you know all this actually)

@joeldn hmm, A5 is sort of on my radar, but it only really matters if you forget to use the specific length/area for each cell. the downside of A5 is you lose the traversal

which in my defence, looking now, i have only done in two places (the distance from centre calculation and the cumulative population). i will fix those :)

if you go into the settings you can flip the scale. in my head it's like a flame, red is cold, yellow lukewarm, blue hot. but literally everyone disagrees with me

@wnd

@bovine3dom @joeldn there will be a variation in cell area whatever you do. There are 12 pentagons in the H3 projection, which is 10% at the coarsest resolution and becomes vanishing small at higher resolution:

https://h3geo.org/docs/core-library/restable/

For me, it is then how much you care. The Earth is an oblate spheroid with an approximate set of spherical coordinates based on EPSG:4326. Which is about 2m accuracy at the equator, less so elsewhere. YMMV

@wnd @joeldn

h3 vary much more than that, we're talking like max ~40% difference in edge length. the a5 site has a nice visualisation of it https://a5geo.org/examples/area

within a single country you don't need to care all that much but once you look across an entire continent it starts to bite

@bovine3dom @joeldn maybe I need to reimplement my Full Automated European Luxury High Speed Railway using pentagons.

As I'm lazy and haven't read the manual, do you know whether these tiles have any rotational or translational symmetry?