Closed falbrechtskirchinger closed 12 months ago
Any chance you can come up with a test especially for the minimized state restore? Should be enough to use a dummy window with a CGI ID matching an "enabled" window
Just want to be sure we'd catch e.g. the virtual call bug in the ctor you found.
Any chance you can come up with a test especially for the minimized state restore? Should be enough to use a dummy window with a CGI ID matching an "enabled" window
Done. Let me know if this is what you had in mind.
Just want to be sure we'd catch e.g. the virtual call bug in the ctor you found.
That's rather difficult to test. I couldn't call SetMinimized()
from the constructor, because that ultimately calls Resize()
, and clang-tidy
would rightfully warn, that this doesn't call the most derived Resize()
as one might expect/want.
Instead, I'm calling Window::Resize()
myself, which is already called from the constructor, so I wouldn't expect any unanticipated side effects.
@Flamefire Rebased and resolved the one-line merge conflict in IngameWindow.h
. You can re-enable auto-merge.
PersistentWindowSettings
constructors and move default initialization in class.PersistentWindowSettings
only once inIngameWindow
(assumes pointer stability). This will make more of a difference when more settings are saved in these PRs:1601
1606
Resize()
from the constructor.