Corwin Joy
This fix sounds great! Can't wait to see it! My current drop-down workaround on the client side for DT is a real headache.
Interesting and thanks for the quick response! I will take a closer look at jaxtyping.
Thanks again for the suggestion. After looking at jaxtyping I decided that I didn't want to introduce a dependency on Jax since that adds quite a bit of complexity. So,...
Draft PR to illustrate query crash problem reported in https://github.com/duckdb/duckdb/issues/10699.
@soerenwolfers @m4rs-mt @Tishj It took me a while to come up to speed on duck_db, but here is a proposed PR to address the referenced issue. Testing the behavior within...
@Tishj 1. That is a good catch about the call to get(). I didn't realize that the exception was not propogated without it. I will add this. 2. I believe...
@Tishj For the interrupt test, I agree that it's similar to what is already there. I would say the only added feature is to benchmark the interrupt response time and...
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 Jupyter may kill threads...
@Tishj OK. I did some more investigation and here is what I found: 1. Running simple tests with pybind it seems I was mistaken, long running jobs are not killed...
Here is a bit of the backtrace / core dump from Jupyter. As above, it's basically a race condition during query shutdown. ``` (gdb) bt #0 __pthread_kill_implementation (no_tid=0, signo=6, threadid=139806566774336)...