Michael Sumner

Absolutely agree. Whatever is easiest, but we do need to have though about where things go - I'm a bit wary of requiring igraph (for example). I'm happy for it...

Still a pressing issue? We can easily leverage sf to do this, I might need help with the structure-composing on the other side, but the sf-decomposing is no problem, **and**...

> Could someone monitoring this issue create some test points (longitude, latitude) -> (X,Y) with Esri software, possibly using ESRI:54099 "WGS_1984_Spilhaus_Ocean_Map_in_Square", so we can check interoperability ? inadvertently, not seeing...

cool, thanks both - great to see this discussed, I would add a few thoughts (just OTOH I'll be following this in more detail), and some general concepts, and just...

It's really the arbitrary cell thing I think that arrow could help with, currently in R for example to do a really big polygon extraction over a long time series,...

> * Ability to read lower resolution without computational resampling (in my understanding cloud-optimized GeoTiff is resampled at write time). That's right, it makes a huge difference for arbitrary windowing...

oh indeed, thanks for the context-clarification - it's very much on brand for me to brain dump too fast when I'm excited. I'll keep thinking about the library functions part,...

fwiw I have R versions of the cell logic in vaster and similar for tiling in grout, both unpolished: https://github.com/hypertidy/vaster https://github.com/hypertidy/grout I think these makes sense on their own, grid...

It needs segmentizing (adding redundant vertices) along the degenerate parallel at -90. Personally very interested in the variety of available solutions to this, the ultimate of which ( in the...

Oops I see you already segmentizing, which suggests it needs to be offset from -90 by a tiny amount to stay in bounds for some projections. Keen to share what...