(Looks like this one got missed.)

Allow creating unbootstrapped `TorClient`s (and `DirMgr`s)

Closes #293

See merge request tpo/core/arti!298

This commit changes how the `TorClient` type works, enabling it to be
constructed synchronously without initiating the bootstrapping process.
Daemon tasks are still started on construction (although some of them
won't do anything if the client isn't bootstrapped).

The old bootstrap() methods are now reimplemented in terms of the new
create_unbootstrapped() and bootstrap_existing() methods.

This required refactoring how the `DirMgr` works to enable the same sort
of thing there.

closes #293

Write some examples for TorClient::connect()

See merge request tpo/core/arti!303

tor-socksproto: Implement HasKind

See merge request tpo/core/arti!299

These should explain better what you can pass as an address, and
what you should do if you _really need_ to provide a SocketAddr or
something.

Refactor the Error type to remove the yucky internal hidden Truncated
variant.  Instead, there's now an embedded tor_bytes::Error value.

If that tor_bytes::Error is Truncated, we bubble it up when we convert our
handshake result to the nested error struct.

Thus there is still (sadly) a variant of tor_socksproto::Error
that shouldn't be exposed to user code.  But refactoring every
inner method under handshake.rs seemed like a bad idea: once we're using
Result<Result<..>>, the ? operator no longer helps us much.

I'm not in love with this solution; the others just seem a bit ugly
too.

(This error isn't yet wrapped in TorError, but it will be eventually
when we implement socks proxy and PT support.)

Fix typos

See merge request tpo/core/arti!301

tor-bytes: Use InternalError.

See merge request tpo/core/arti!300

This crate's Error type is too low-level to have an ErrorKind, but
it does make sense to use InternalError for the internal errors
here.

tor-config: Add HasKind support.

See merge request tpo/core/arti!296

This required a few new ErrorKinds.

This edge-case was there even before the migration of
595fe1ab, but now it's more explicit and ought to be
revisited.

Remove the use of Mutex in channel unused_since timestamp

See merge request tpo/core/arti!293

Temporarily allow failures in the coverage CI.

See merge request tpo/core/arti!295

Right now it's failing on nightly through no fault of our own.

Another set of improvements to the error code

Closes #321

See merge request tpo/core/arti!294

This fixes a compilation error.

Instead of declaring a macro that takes vis as an argument, we now
conditionally declare a macro that applies an appropriate visibility.

There's a long comment explaining the rationale here, along with a
couple of other solutions that don't work.

This is closer to common usage.

(Not that we all agree with common usage, but it's closer to what
people expect.)

This is closer to what we described in Errors.md.

Also, remove the (sometimes private) Result alias: it was only used in
one or two places, and never exposed in public.

This change lets us make TorError's members unconditionally hidden,
and makes our API a little more consistent (since basically nothing
else is a public field).

Hopefully uncontroversial improvements to new Error code

See merge request tpo/core/arti!291

Expire channels that have been unused for too long

See merge request tpo/core/arti!273

These tests turned up a need for using the #[track_caller]
annotation in order to get accurate locations, which is fortunately
stable since Rust 1.46.0.

(By our convention, these errors should say what we were trying to
spawn when the error occurred.)

At least by default, we should have Error be private, and not expose
it as part of our APIs.

To keep functionality in `arti`, I had to add an `ExitTimeout` error
kind.

For interface consistency, I also re-exported ErrorKind and HasError
from `arti_client`.

I had planned to use assert_impl, but that can't check for
the 'static lifetime.

This patch makes only minimal changes in lower-level error types:
we have more refactoring to do.