Instead of waiting a long time, let's just check that networking truly
is disabled, in the same way we do for the MAC spoofing tests -- after
all the code for the "Diable networking" feature actually originates
from the MAC spoofing feature. This way we do not have to introduce a
static wait for 120 seconds, and we do not need the anti-test.

They all rely on the `wait_until_tor_is_working` helper which has lately
proven to be still buggy. That's tracked by ticket #10497.

Will-fix: #10706

Will-fix: #10430

In Jessie, org.gnome.gnome-screenshot.auto-save-directory is empty
since the default is sensible, ~/Pictures. To save the removed
scenario we'd have to set it explicitly, which seems stupid. Besides,
the 'A screenshot is taken when the PRINTSCREEN key is pressed'
scenario is already testing that GNOME Screenshot saves into this
directory, which is enough.

This reverts commit 05fe0a1b.

Essentially I did:

    sed -i 's/Tails has booted/I have started Tails/' -- \
        features/*.feature features/step_definitions/snapshots.rb

(although there was a false positive that I had to restore)

Will-fix: #9402

Will-fix: #9143

Screenshots are now handled by GNOME Shell.

Make sure the condition we're looking for is not verified before the action that's supposed to make it true.

Sikuli isn't reliably detecting the notification, either because it
disappears too quickly given Wheezy's GNOME3's hardcoded timeout of
five seconds, or because it fails to show at all. Any way, this often
introduces failure of the 'Tor is ready' step, so this should improve
the robustness of the test suite.

The 'fuse' group was dropped in fuse 2.9.3-5 due to Debian bug

Factorize common pattern into a single "I start Tails from DVD with network unplugged and I login" step.

Sometimes Sikuli fails to find the shutdown icon in the shutdown
applet's menu (possibly because it's covered by a notification) so
that method isn't completely reliable at the moment. To contain this
problem we test the shutdown applet separately, and rely on the more
dependable remote shell for halting in the general case.

While we're at it we also make sure that the shutdown steps are
formulated so what they do is more explicit.

This reverts commit 293dac53.

This may easily lead us to forget about these tests, so we'll have to
invent a more involved mechanism for tests expected to fail.