Trying to work with data stored on S3, I've come across the fact that when using the duckdb R package, I have to load the httpfs extension as a pre-requisite, which I don't have to do when using Python's duckdb module or the standalone CLI.
I am probably missing something, but it looks like this is linked with the fact that when I run FROM duckdb_settings() on the R package, I can see that the option autoload_known_extensions is set by default to false, while it is set to true in Python and on the CLI (which is actually consistent with the documentation).
As a consequence, I have to explicitly load the extension httpfs, should I want to query data stored on S3 for instance. It seems not completely consistent with the documentation. Do you agree with this, and if yes, would it be possible to consider turning the default of this option to true?
Trying to work with data stored on S3, I've come across the fact that when using the
duckdbR package, I have to load thehttpfsextension as a pre-requisite, which I don't have to do when using Python'sduckdbmodule or the standalone CLI.I am probably missing something, but it looks like this is linked with the fact that when I run
FROM duckdb_settings()on the R package, I can see that the optionautoload_known_extensionsis set by default tofalse, while it is set totruein Python and on the CLI (which is actually consistent with the documentation).As a consequence, I have to explicitly load the extension
httpfs, should I want to query data stored on S3 for instance. It seems not completely consistent with the documentation. Do you agree with this, and if yes, would it be possible to consider turning the default of this option totrue?