@morganherlocker want to try this out?
So: is the output of cellid.totoken ordered - as in, rangeMin and rangeMax will correctly compare as strings?
This is important because we need an index encoding that easily supports range queries.
Yes, they appear to be. I went in a line and gathered up a few of the tokens, and they were ordered alphanumerically.
I went 4 cells west => east, but I believe it should follow the ordering of the hilburt curve across more cells. I will take a look to verify this. I think that the ordering should be like this:
Ok, they are definitely ordered, and everything looks good. The indexes follow a hilburt curve, and lower zoom cells have a truncated hash of the higher zoom cells, so ranges should work across zoom levels.
fyi: here is the code: https://github.com/morganherlocker/cover-extract