Because opening a profile already had a progress dialog open, the import
process during a forced rebuild wasn't appearing, so it would look like
OSCAR was hanging for as long as the import took.
Temporarily hiding the profile progress dialog resolves the issue.
to act differently from profiles after a subsequent application launch.
It was only getting set at the end of Profile::OpenMachines, which was only
called by the Profile constructor, and which wouldn't succeed on a newly created
profile, having no files to open. Therefore, the only way for the flag to be set
was to quit and re-launch the application after creating the profile.
The flag's only remaining use was to make sure that OpenMachines() wasn't
getting called twice and trampling an existing list of machines, so the check
there was changed from looking at a brittle flag to looking at the actual list
of machines.
A critical warning was also added to the check, since OpenMachines() is
only getting called from the Profile constructor and therefore can't
be invoked twice unless a new bug has been introduced.
duplicate sessions being added when making preference changes to a brand new
profile with existing data.
Preference changes trigger a reload, see PreferencesDialog::Save and
needs_reload, leading to MainWindow::reloadProfile -> MainWindow::CloseProfile
-> Profile::UnloadMachineData.
This is the root cause of the duplicate sessions, but the fact that it only
happened with newly created profiles, rather than on subsequent launches,
demonstrates an inconsistency in state of Profile. It should be identical when
initially created and when loaded via subsequent application launch.
Currently there is a very messy tangle of dependencies between
loaders, machines, sessions, and profiles. Right now the
simplest way to create a test loader instance is to create
a test profile, under which the machine and session instances
will exist.
