Antonio Cuni

> I plan to write an improved version of `wait_for_console` which works even in case the output is already in the console ok, I did it in PR https://github.com/pyscript/pyscript/pull/1363 :)

> But now something weird is happening... tests are _crashing_ both in CI and when run locally, when running threaded. They seem to be the same collection of tests that...

> Tests do all successfully complete when run not-threaded, so I suspect it's some interaction with `pytest-xdist` or similar. I tried running with `--numprocesses 4` (instead of `auto`) to use...

> @antocuni that is not quite the situation I have been seeing, and also my commits this morning have changed things a bit: > > 1. Running tests locally and...

> @antocuni I refactored `test_pyscript_src_not_found` to remove the `with pytest.raises(PageErrors)` - if you look at the output of [these successful CI tests](https://github.com/pyscript/pyscript/actions/runs/4980976639/jobs/8914584534) vs [these failed CI tests](https://github.com/pyscript/pyscript/actions/runs/4981330397/jobs/8915441467) (between which nothing...

I'm fine with merging this PR as is but yes, the problems that it uncovered must be fixed. If our test machinery is not bullet-proof, this means that it's just...

I have started to investigate the issue. In theory the execution flows like this: 1. We are in the `main` thread: `PyScriptApp.installPackage` calls `interpreter._remote.installPackage`: `interpreter._remote` is a Synclink proxy, so...

@madhur-tandon I think there is a bit of confusion w.r.t the final goal. In theory, the current situation is the following: 1) PyScript uses `Synclink` which uses `SharedArrayBuffer`, which is...

> @antocuni Although I have added the error message, it is not in the form of a `UserError` / alert banner. @madhur-tandon see my previous message: https://github.com/pyscript/pyscript/pull/1374#issuecomment-1502885214 The source of...

+1 for fixing this soon