Martijn Visser

Would be nice to have this. Tests pass, also checked locally. The potentially breaking changes are documented in https://github.com/JuliaLang/Compat.jl/issues/775, but none seem to apply for this package.

I believe in that case I was reading the HydroATLAS / BasinATLAS data. This is a direct link to the download: https://figshare.com/articles/dataset/HydroATLAS_version_1_0/9890531?file=20087237 One nice property of this dataset is that...

Great that you were able to improve the performance so much! I don't really have experience with generated / RuntimeGeneratedFunctions / GeneralizedGenerated, so it's hard for me to comment on...

Really sorry to hear that @yeesian. Of course take all the time you need, I wish you strength during a difficult time. I can probably help out a bit here...

@mathieu17g I'm a little bit lost. It would be of great help to me if you could write a high level summary of (1) the issue that this PR is...

Thanks a lot @mathieu17g for taking the time to write down all your findings! I'll get back to it when I have some more time.

@mathieu17g thanks for the bump, and sorry I did not yet sit down to get the proper look that this work deserves. I hope to get to this soon. I...

Awesome, I tried your inlining change from https://github.com/JuliaData/Tables.jl/issues/273#issuecomment-1037054582 as well. On level 4 BasinATLAS I see similar results, around 100x speedup of compilation, and no real effect on runtime, only...

Great, thanks! Though perhaps better to split it into two separate PR's since the two changes are unrelated. The inlining change is straightforward. For the dicts I guess the main...

Here is a slightly reduced example: ```julia julia> import ArchGDAL as AG julia> ds = AG.read("Test.gpkg") GDAL Dataset (Driver: GPKG/GeoPackage) File(s): Test.gpkg Number of feature layers: 3 Layer 0: polygons...