When set to 1, a test will fail at start if there are ipfs daemons running
that *seem* to have been started from within the test/sharness directory.

This helped weed out a number of failures to cleanup

Ideally should be enabled by default, but things are too unstable for this
to work :/

Without reflecting this in the tests we do not actually shut down
correctly during cleanup

Can not explain why this happens, but the fix seems to work. Change does
not seem to affect what the test is actually doing anyway ( the wait makes
the entire block synchrronous )

Some did actually prevent proper test shutdown

Set this to '1' if you are doing local dev and have redirects in go.mod

Docker on a mac might be available, but may not be runnng. The
way the original test was written is linux-specific anyway, so
just hoist the check into the DOCKER declaration

54b1969c correctly switched from nc to socat, but did not account for it
not being as widely available. Adjust tests that depend on it with the
correct message.

There are also a number of comments pertaining to nc - they were not
adjusted.

IPNS over PubSub as an Independent Transport

docs: add information on how to enable experiments

Change Reporter to BandwidthCounter in IpfsNode

update go-datastore

This was failing in CI.

chore: update readme badges to test ci

testing out circleci build triggers with a PR that tidies up the README badges

- remove somewhat redundant `project: ipfs` badge so they fit on 1 line. The link to the website is the url field for the repo.
- Use shields.io "flat square" style for godoc badge to match the others
- Use shields.io for the circleci badge too.
- re-order to be more visually pleasing.

License: MIT
Signed-off-by: Oli Evans <oli@tableflip.io>

Let's get this in so the datastore changes don't block other PRs.

chore(dep): update libp2p

Otherwise, we have port conflicts and other issues.

go fmt: go-ipfs-as-a-library

feat: web ui 2.7.2

extract the pinner to go-ipfs-pinner and dagutils into go-merkledag