Popper with overflow prevention(config
which is set by default), when run in Karma test environment with iFrames - results with certain elements positioned with offset equal to the test's frame's offset.
It happens only for elements which are due to be positioned at "y: 0".
Popper either recognises those elements as conflicting with (overflowing) the iFrame or in some other "weird" way includes he frame's offset in it's calculation.
It does it only for first / top most / elements overflowing the frame.
It's a glitch somewhere between Popper and Karma.
As running Karma tests without iframe is not feasible ATM this may need further investigation as it's a scenario which affects our layering in general.
testable with: inlineDialog2 as it's configuration does not enforce
Usage described in documentation
Demonstration of all states
Interactive state (screen readers)
Visual regression tests
Navigation (only keyboard / only mouse)
Component render legibly
Element status (At Rest/Pressed/Selected/Read Only/Disabled)