Ryan Hiebert

Makes sense to me, I was thinking the same thing when I was refactoring, but I wasn't sure it'd be a common use-case. My proposed implementation would be to make...

Thanks! I've been wondering how that project was going.

That's an interesting use case. I think the first thing I'd need to see is how we'd determine that the current Python matches the system Python (does it actually need...

Interesting enough, I do not neccesarily expect that _any_ Python version will be the Travis version. I'd rather expect that all of the Python versions you might choose from would...

Thank you for opening this. I've been thinking about it the past day or so, and I've been questioning exactly what we do and don't want to do in this...

The strict mode is what this is about, and why the deprecation notice even exists. It was my intention to enforce this, but this isn't gonna be a good idea....

> see vstinner/faulthandler@957915c , for adding testing of an environment that is declared by tox. py26 is a part of the standard tox environment list. This is an interesting assessment,...

I've often wished that Travis had this feature natively, to report to GitHub statuses on each Job, with enough context to make this easy. I'd be interested in a feature...

The implication in this issue seems to be that we would _not_ want to fail the Travis build for an env failure. I think that would be a mistake. I...

The hooks that we'll want to use for this are: * `tox_runtest_pre(venv)` * `tox_runtest_post(venv)`