The behavior in question is to ignore the entries entirely,
and not check their files at all.

CI: simplify chutney installation

See merge request tpo/core/arti!2600

This doesn't actually matter much when run in CI, but provides a better
example for running locally.

We no longer need a path to a checkout of the chutney repository; we
just need the chutney binary itself. We still allow this to be set
explicitly with CHUTNEY_BIN, but fall back to finding it on the PATH.

This takes advantage of a feature added in
tpo/core/chutney!31.

This is a step towards not needing to set CHUTNEY_PATH to point to a
chutney repository.

This will let us simplify chutney installation, which will be done in a
follow-up commit.

CI python3-checks: Don't install redundant python packages

See merge request tpo/core/arti!2618

tor-proto: Document RecvWindow::take more accurately.

See merge request tpo/core/arti!2611

arti-client: add comment to `TorClient::reconfigure`, and use `Reconfigure::cannot_change` for `DirMgrConfig::cache_trust`

See merge request tpo/core/arti!2620

`DirMgrConfig::update_from_config` uses the old value and not the new
value, so it should be reported using `Reconfigure::cannot_change`.

We install recent mypy and flake8 packages from pypy below; we don't
need to also install from apt.

tor-hscrypto: Abolish the HasKind impl of HsIdParseError.

Closes #1483

See merge request tpo/core/arti!2607

CI: track and expose timing information

See merge request tpo/core/arti!2605

cargo: update hashbrown version to 0.15.1

Closes #1725

See merge request tpo/core/arti!2613

tor-config-path: update `unix_literal` test to use `USER_HOME`

See merge request tpo/core/arti!2612

Version 0.15.0 was yanked, which makes cargo-audit unhappy.

I think `USER_HOME` was intended here.

The docs here seem wrong, so I am updating them.

tor-config-path: finish migration from tor-config

Closes #1717

See merge request tpo/core/arti!2593

This will make it a bit easier to look through the output in the gitlab
UI, particularly making it easier to see where time is being spent in
these long-running jobs.

Provide functions for starting and ending spans instead of the script
doing it itself. The current version of the functions are directly
from the gitlab documentation
<https://docs.gitlab.com/ee/ci/jobs/job_logs.html#custom-collapsible-sections>.

This is currently functionally similar to what was here before, but
using functions here will let us make this a bit more convenient, which
will be done in a follow-up commit.

This will let us more easily see where time is being spent.

https://docs.gitlab.com/ee/ci/jobs/job_logs.html#job-log-timestamps

The old code used a shortcut to access the `Figment` directly, but we
need to do the whole config builder thing to be able to parse the toml.

Also updated other packages to get `CfgPath` directly from
`tor-config-path' instead of 'tor-config'.

We only ever return `BadOnionAddress` is we are asked to connect to a
.onion address that doesn't parse.

Also changes the `ErrorKind` of `BadOnionAddress` to
`ForbbiddenStreamTarget`.

See context in
!2246 (comment 3045683)

Closes #1483

New CfgAddr type to represent SocketAddrs in configuration.

See merge request tpo/core/arti!2597