Results 963 comments of Ned Batchelder

Also, `insert_marker` should be `start_marker`, especially since the corresponding setting for the end is already called `end_marker`.

Available as of [setuptools>=62.6.0](https://setuptools.pypa.io/en/stable/history.html#v62-6-0).

This seems to have been fixed somewhere? It doesn't happen any more I don't think.

I don't have a solution to the problem, but I'm curious what your custom plugin does for Python files?

> It tries to address what is also reported as https://github.com/nedbat/coveragepy/issues/856: Find calls to functions that were compiled by TensorFlow's autograph functionality, and report the original function as covered. Interesting!...

Thanks! There was also some mention some place of moving the fuzzing code itself into this repo so that we could extend it. How do we make that happen?

I'm not fond of saying that ".jinja" means Python. As #1488 shows, the Python code made by Jinja will be named whatever the original template file was called. So it...

@etaylork Can you provide us with a way to reproduce the issue?

Ideally, you would have a repo with the exact code that demonstrates the failure.

This is possible by querying the SQLite database directly: ``` % sqlite3 .coverage "select distinct f.path from file f, arc a, context c where f.id = a.file_id and a.context_id =...