Antonio Cuni

If the goal is to run psycopg3 fast in PyPy, there are at least three paths: 1. use a cffi backend: this is likely to be the fastest one 2....

oh, I realize only now that you are calling the Python C API *from Cython*. In the discussion so far I thought/assumed you were writing C code directly. This use...

> I'm not sure it does need any cooperation with Cython. The entry point of an HPy module is different than the one of a C API module; for HPy,...

I think you nailed it, although your explanation is partly incorrect. > Both of them execute a fourth step 'Initializing scripts...' as the last initialization step. On the other hand,...

Sorry, I overlooked this issue and seeing it only now. > The PyScript examples tests when I last looked had to do a timing dance to look at the console...

That said, some of the test in `test_zz_examples.py` seems to be flaky: some are marked as `xfail` because of that, and I also saw `test_d3` failing here and there. And...

> FWIW, I went in a different direction than using the console for sending flags. The idea: set a `data-ps-status` attribute in a well-known location (e.g. ``). This lets any...

Sorry, chiming in late in the conversation. I agree with Fabio that there are many different use cases and workflows and we should find a way to accommodate all of...

FWIW, in case it's needed in the future, another alternative which is better than the current logic but more lightweight than full AST parsing is to use the `tokenize` module,...

Sorry for chiming in late in the conversation, just back from vacation. I'm writing this after reading the above messages but I admit I probably need more time to really...