Thijs

``` ➜ duckdb git:(main) ✗ ./build/debug/test/unittest test/sql/attach/attach_transactionality.test Filters: test/sql/attach/attach_transactionality.test [1/1] (100%): test/sql/attach/attach_transactionality.test =============================================================================== All tests passed (32 assertions in 1 test case) ``` Thanks for the report, could you possibly...

We already have a test for this, I'm not convinced this change is needed https://github.com/duckdb/duckdb/blob/b3a53b3e6730ffd4b492ebfa732913b06a8a0059/tools/pythonpkg/tests/fast/api/test_query_interrupt.py What you're experiencing is likely related to the Jupyter specific logic present in the python...

Hmm I might see the benefit, the execution logic is separated from the interrupt check Essentially creating an extra temporary worker thread I have some questions regarding exception handling We...

To continue on that last part, the routine executed by the std::future can produce an exception, we need a get() on the future to make sure we propagate it, otherwise...

I see, I can not reliably get Jupyter notebook working, I have set it up in the past but haven't been able to get it working since - in a...

@corwinjoy regarding the existing interrupt test, many hours have already been sunk into that test, this is the most reliably consistent we have gotten this to behave I recently tested...

> Also: > > 1. Added suggested changes for future.get() and an abort signal to the thread. > 2. As mentioned in the issue thread, I give details on why...

@corwinjoy Thanks for your investigation on this! We're not super keen to add std::async and std::future to the project, as we've already seen that misstakes are easy to make, for...

`ParquetWriter::TypeIsSupported` is missing support for the ARRAY type

Ah actually, we are already aware of the issue, the linked PR explains it It will be addressed at some point