Nathan Goldbaum
Nathan Goldbaum
See https://pyo3.rs/v0.23.4/migration.html#from-021-to-022 for more details. This passes the tests using the rust backend on my local Mac setup except for one test that is also failing on CI. I also...
### Describe your issue. If I run the full SciPy tests on my M3 Macbook Pro running MacOs 15.5 with the default open file descriptor limit of 256: ``` goldbaum...
## Description Starting jupyterlab on the free-threaded build on MacOS immediately segfaults. See https://github.com/jupyter/notebook/issues/7512 for the same issue reported against `notebook`. ## Reproduce Install a free-threaded python on MacOS with...
### Describe your issue There are a couple of symbols in the NumPy iterator C API that are defined in a way that Cython doesn't like: https://github.com/numpy/numpy/blob/0dbee7ec68f1adcabfb0344c53d65b7f385ae6b5/numpy/_core/include/numpy/ndarraytypes.h#L1067-L1070 If I write...
Perhaps complicated by the fact that hatch isn't easily installable on the free-threaded build (#1801, but see https://github.com/pypa/hatch/issues/1801#issuecomment-2715720577), it doesn't look like it's possible to create a hatch environment with...
This ends up coming down to the dependency on `zstandard`: ``` ± pip install hatch Collecting hatch Downloading hatch-1.13.0-py3-none-any.whl.metadata (5.6 kB) Collecting click>=8.0.6 (from hatch) Downloading click-8.1.7-py3-none-any.whl.metadata (3.0 kB) Collecting...
This is a neat idea from tildes: https://tildes.net/~tildes.official/cjg/experimenting_with_a_way_of_avoiding_deeply_nested_comment_threads Basically if a comment thread is a back and forth between two accounts, don't visually indent the thread, just continue on the...
https://numexpr.readthedocs.io/en/latest/release_notes.html
Unfortunately actually doing this might end up being blocked on upstream support in CFFI for free-threaded Python, see https://github.com/python-cffi/cffi/issues/126 for more detail. We are working on fixing CFFI in a...
This uses thread-local storage to avoid a scaling bottleneck due to a global cache in CPython internals. Unfortunately it's pretty slow ~10% of the total runtime of my benchmark is...