Comment 25 for bug 1632362

Although note that with the old persistent-cookiejar code, as soon as the wait gets to 100ms it does make only four further polls. A new PR is changing this machinery in a different way. It *won't* deal with stale lock files (if that's the underlying issue), but it will deal with lock contention:

https://github.com/juju/persistent-cookiejar/pull/21

We'll land this in develop today.