Adam Wulkiewicz

@barendgehrels > I assume it is still possible to be more specific (for example: specify Haversine) or less specific (pass no strategy at all) Yes, I omitted the default and...

Thanks for reporting/asking about it. This looks like a bug or at least an inconsistency in how pointers to geometries are treated. ---- A workaround for this issue is to...

> I'm happy to run some additional tests if you'd like to see a different setup. No, that's fine. Thanks for sharing! I think it makes sense. In typical case...

> It appears to me that rtree.qbegin(bgi::nearest(pos, N)) now completely sorts up to N observations when the first pointer is dereferenced? The knn query iterator should access all values that...

Happy Thanksgiving! > I was never sure how to invoke the packing algorithm. Is it as simple as use the constructor taking an iterator? Yes, you can either pass a...

What is the goal? How did you discover that this may be an issue? Could you describe why do you think this is the part of the algorithm that is...

Hi, Thanks for the suggestion. This was something I was considering in the past. Ok so there are 2 issues here: your problem (1) and general predicates expression support in...

The first case looks like that:  red - poly1 green - poly2 blue - poly (result) With `BOOST_GEOMETRY_NO_ROBUSTNESS` the result is equal to the poly2 (with additional point). Without...

The second case:  As @linshu77 said it's the other way around WRT `BOOST_GEOMETRY_NO_ROBUSTNESS`. However this time the poly1 is the result. {171.60480351137778, 162.32096650635697} {172.21795926332567, 162.99736073096460} {172.21795926332567, 170.00000000000000} {171.60480351137778, 170.00000000000000}...

> I can exclude the start-turns for non-overlay operations, and I will. Ok but will then set operations and relational operations be consistent with each other? > Are you sure...