Marlon James

`cancel()` was added in #2877. Deprecating `kill()` is to-be-decided.

`test_both_edge_triggers` should now pass with riviera 2023.04: https://github.com/cocotb/cocotb/blob/c34c2bc5cf2408c0c5c386d5d9c139ecf6453889/tests/test_cases/test_cocotb/test_edge_triggers.py#L232-L244

This closes #1522

The problem is that the scheduler only has the `_InternalEvent` trigger (because Waitables are not triggers at all), and that `_InternalEvent` is local to the coroutine `_wait()` that is running...

I think they should be triggers so that they can be `unprime`d and be able to properly clean up. Edit: Although, that's really just a different implementation of the same...

If we go with extensions, I think it makes sense for these to live in the cocotb GitHub organization.

Are you talking about a more general plugin system where cocotb provides hooks that the plugins can use?

1. Is supporting remote debugging on Python initialization desirable? 2. If so, can we settle on an approach to enable it? I like the simplicity of a single environment variable,...

An alternative approach is to support entry points, such as `cocotb.startup`, and run all of the provided entry point functions, allowing each one to determine it's own behavior through whatever...

@themperek I saw that, as well as some other examples using **entry_points**. Pros/Cons for the two approaches as I see them: | Approach | `COCOTB_STARTUP_ROUTINES` | entry_point | ---------- |...