Ideally, this will get rid of some of our spurious test failures that occur from relying on httpbin.org and will make the test suite faster as we won't be going over the network, we'll just be using localhost (which is still slow but less slow than actually going over the network).

The README says

All you need to do is [...] wrap your axios code in a VCR.use call.

and then the example code didn't use any code like that, VCR.use?

I assume this is due to API changes.

Is it bounding by mountCassette and ejectCassette instead these days?

I wonder what is the intended logic behind the 'new_episodes' option ? I couldn't find it in the README, and I had some "strange" behavior with it...

I ran multiple requests to the same endpoint, in one vcr with --record-mode=new_episodes.
But only the first attempt of each request was registered.

So something like :

import time

import pytest
import aiohttp


@pytest.mark.

Packages that use vcr for testing have the need for similar guidelines regarding testing (at least as long as the use of cassettes isn't widespread).

• Could a template live somewhere in this package?
• It might include a part that says "fill this in with workflow regarding setup/teardown" cf #137
• It might give a link to the book
• Could it be mentioned by use_vcr(), "add the following

While the team building aries-vcr has been pretty good at adding logging in a useful way, there is some inconsistency in log entry handling. This task is to formalize the logging approach across the board, and then to do a pass through the code ensuring that the logging statements match the intention of the approach.

The first part (formalizing) is likely just creating something like a "logging