You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Would there be interest in creating a "backend system" for GeoDataFrames, so that we can use native Julia readers wherever possible, and fall back to ArchGDAL when needed? This would probably help performance quite a bit as well.
The other proposal was that GeoDataFrames could re-export Extents.jl and GeoFormatTypes.jl, so that users don't need to know about those!
The text was updated successfully, but these errors were encountered:
That's certainly possible, and I think that still works without wrapping either the geometry, or Dataframes, but is that still the right way ahead? I like spatial to be not special, but that makes it also harder.
With regards to exporting, what happens if we re-export here and someone does using Extents?
Agreed that spatial should not be special. IMO we can sneak in quite a bit of stuff through metadata and GeoAcceleratedArrays.
On the using extents part of it - it's fine, there are no name clashes by definition since it's the same function. (I mean literally @reexport using Extents).
From https://discourse.julialang.org/t/screening-interest-in-a-unified-vector-raster-package-akin-to-r-terra/121613, it might make sense to make GeoDataFrames a single point of entry for vector datasets.
Would there be interest in creating a "backend system" for GeoDataFrames, so that we can use native Julia readers wherever possible, and fall back to ArchGDAL when needed? This would probably help performance quite a bit as well.
The other proposal was that GeoDataFrames could re-export Extents.jl and GeoFormatTypes.jl, so that users don't need to know about those!
The text was updated successfully, but these errors were encountered: