This reverts commit 1a8f7e466c.
Current Plasma Discover complains if fwupd is masked. I just hope
tests don't start failing on fwupd errors again now. Sigh.
FreeIPA team bumped the F41 NVR past the F42 one, so now upgrade
tests from F41 to F42 are broken. To workaround this we need
bumped rebuilds for F42, F43 and F44, because if only do F42 the
problem moves to F42->F43 upgrades, etc.
Signed-off-by: Adam Williamson <awilliam@redhat.com>
I did this to check the size of an image built with a lorax
scratch build I was testing, but then figured heck, why not also
do it permanently?
Signed-off-by: Adam Williamson <awilliam@redhat.com>
81d5d2f408 inadvertently dropped
this, then I carried that change over in the profile group
rewrite. I never intended to turn it off on KDE, though - it was
likely just a git rebase mishap or something. So let's turn it
on for KDE by using the "updates-desktops" profile group like we
do for all the others.
Signed-off-by: Adam Williamson <awilliam@redhat.com>
The top panel stays in the application and does not hide. I might be
caused by the mouse being in top area which normally invokes the
panel to appear. Let's move the mouse and click into the middle
of the screen to free the top area.
For some reason, the login name is filled in double time.
Let's fix it with adding more wait time and also an extra
delete key to wipe the field properly.
It's not safe to fast-type things at graphical consoles. This is
failing quite a lot on aarch64 currently due to heavy load on
the worker hosts. Slow typing should help.
Signed-off-by: Adam Williamson <awilliam@redhat.com>
still getting some failures at 1800, so let's bump to 3600 and
also bump MAX_JOB_TIME to accommodate.
Signed-off-by: Adam Williamson <awilliam@redhat.com>
These seem to be timing out a lot currently on aarch64, probably
due to the heavy load caused by the Python update flood.
Signed-off-by: Adam Williamson <awilliam@redhat.com>
In IoT F43 and F44 at present, greenboot is failing if networking
is not immediately available on boot. Well, on our tests that use
static networking, it isn't. So that's a problem. Let's have the
static networking setup test stop greenboot if it's running on
IoT, that should get around this.
Signed-off-by: Adam Williamson <awilliam@redhat.com>
Matching on the cursor isn't reliable as it flashes and animates
a bit. Let's just match on an area to the right - this is
sufficient because the box doesn't have the blue surround unless
it's active.
Signed-off-by: Adam Williamson <awilliam@redhat.com>
Calculator changed the way functions are organized within the
application which broke the current Calculator test.
This fixes the subroutine to test for Functions and makes the
test reliable again.
Fixes: https://pagure.io/fedora-qa/os-autoinst-distri-fedora/issue/428
Now we can slow type in assert_script_run, let's do it, and make
this more reliable (I hope). It gets mistyped a lot currently,
especially on aarch64.
Signed-off-by: Adam Williamson <awilliam@redhat.com>
See https://github.com/rpm-software-management/dnf5/issues/2435 .
Lately this is sometimes failing because something else holds a
lock on the RPM database. To work around that, let's try again a
few times if it fails, with a bit of a wait.
Signed-off-by: Adam Williamson <awilliam@redhat.com>
This is a bit ugly but I can't think of another way to pass it
only on aarch64. Could try and factor out the other various
instances of the string against this variable, I guess, but I
don't feel like it right now.
Signed-off-by: Adam Williamson <awilliam@redhat.com>
Similarly to desktop_login, now we have more (and more reliable)
aarch64 resources, we should run this test everywhere. We need
a variant apps_run_kfind needle for some reason.
Signed-off-by: Adam Williamson <awilliam@redhat.com>
This test isn't run on aarch64 (probably because it was too slow
or unreliable on the old hardware). Let's turn it on. We still
can't run it on Silverblue because the background hacks don't
work there.
Signed-off-by: Adam Williamson <awilliam@redhat.com>
When systemd is at the top of the list there's a bit of a grey
highlight that throws off the match.
Signed-off-by: Adam Williamson <awilliam@redhat.com>
The cursor in gnome-calculator somehow changed between the 1.1
and 1.2 Beta candidates. It now looks a bit weird and slanted
(it's an optical illusion caused by odd anti-aliasing).
Signed-off-by: Adam Williamson <awilliam@redhat.com>
In a virtual machine, it takes a couple of seconds for the
upper panel to go away and show the terminal prompt.
Let's make it wait 60 seconds to be absolutely sure.
We need to replace the existing needle with one with a match area
on the 2 button, as explained in the recent commit where we
changed the test behavior. This is for Silverblue, where the old
UI is still present for now, so we need the assert_and_click on
the first shortcut page to press the '2' button for things to
work.
The removed needles all have not matched for months or years, so
let's get rid of them.
Signed-off-by: Adam Williamson <awilliam@redhat.com>
There's no 'go to 2' button any more, we'll just click a random
section (Windows) and check that. This would also work fine for
the old style if we still need that - just add the '2' button to
gte_shortcuts_one and make sure it's the click area.
Signed-off-by: Adam Williamson <awilliam@redhat.com>
