Zacharias Knudsen

If I recall correctly, the state of the branch (which hasn't been touched in a year :open_mouth:) is that we managed to get GiST and SP-GiST operator classes to compile...

That sounds great :smiley: You are correct that the bitwise logic is mostly unused. I have directly copied `h3index.h` from the h3 repository since I needed the `H3_GET_INDEX_DIGIT` macro. I...

I've split the old branch into two, so we can get GiST working and merged before moving on to SP-GiST.

Somewhat related #277 .

Hm, these are all functions that have been renamed at various points. For example (in `h3--3.4.1--3.5.0.sql`): ```sql ALTER FUNCTION h3_get_unidirectional_edge_boundary(h3index) RENAME TO h3_get_h3_unidirectional_edge_boundary; ``` But it is still pointing to...

As far as I can tell the functions are changing link symbol during updates, which we are testing using the `ci-install` tests (link_symbols are in the `prosrc` column). @Komzpa did...

There is still some logical error in the current implementation, as seen from this simple test: ```sql \set hexagon '\'831c02fffffffff\'::h3index' CREATE TABLE h3_test_gist (hex h3index); CREATE INDEX GIST_IDX ON h3_test_gist...

That's the problem. The regular operations work fine, but breaks when using the GiST index as it is currently implemented in this branch: ```sql CREATE TABLE t (hex h3index); INSERT...

Ah sorry for the confusion -- as far as I can tell, your `gist_cmp` works as intended :+1: I think the problem lies in the functions ```c // opclass_gist.c h3index_gist_consistent...

I haven't touched the branch in a year, but I don't wish to abandon it. We need to correctly define the [GIST operator class methods](https://www.postgresql.org/docs/9.0/gist-implementation.html) for the H3 data type....