Destroy PpPPDSelectionDialog in dispose method of PpNewPrinterDialog
as its transient-for property was unset by previous commit
and the dialog is not destroyed automatically now.
Unset transient-for for PpPPDSelectionDialog in its responce callback
since due some reason its parent PpNewPrinterDialog is destroyed first
and once the PpPPDSelectionDialog is being destroyed it tries to
unset the transient-for but unsuccesfully resulting in warnings like:
"instance with invalid (NULL) class pointer"
"g_signal_handlers_disconnect_matched: assertion 'G_TYPE_CHECK_INSTANCE (instance)' failed"
Fortunately for us GtkTreeView still operates pretty
much exactly like in GTK3. Other than that, it's a
dense junction of all that we've done so far to port
other panels.
pp_new_printer_dialog_get_new_print_device is replaced with
pp_new_printer_dialog_get_new_printer which returns a PpNewPrinter.
gtk_show_all is used instead of gtk_dialog_run for PpNewPrinterDialog
and PpPPDSelectionDialog.
This is in response to feedback:
https://gitlab.gnome.org/GNOME/gnome-control-center/-/merge_requests/930#note_1202114
This is accomplished by moving the calls to pp_printer_add_async
directly to CcPrintersPanel. pp_printer_delete_async calls are already
done directly in the CcPrintersPanel so there is consistency gained by
this implementation in addition to PpNewPrinterDialog actually being a
GtkDialog.
A pp_new_printer_dialog_get_new_print_device method has been added to
PpNewPrinterDialog to allow getting the PpPrintDevice selected by the
user to add. This can be called anytime after a response callback
with a GTK_RESPONSE_OK reponse_id.
PpNewPrinterDialog still does asynchronous operations to populate the
dialog, but the create dialog -> receive signal -> destroy dialog flow
can all be handled like a traditional GtkDialog without additional
callbacks or signalling.
The dialog is now displayed by calling gtk_dialog_run and
gtk_window_set_transient_for in the window displaying the dialog.
Cleanup is handled using gtk_widget_destroy.
Destructors in the windows using the dialog no longer cleanup the dialog
as it is assumed it will be either cleaned up after the response
callback, or handled by the destroy_with_parent property set on the
dialog itself.
G_IS_OBJECT checks fail when calling g_signal_connect_object because
PpPPDSelectionDialog is a struct, not a type descending from G_OBJECT.
This makes the changes required for PpPPDSelectionDialog to be a
GObject, and defines PpPPDSelectionDialog as a GObject. It also updates
consumers of PpPPDSelectionDialog to destroy created PpPPDSelectionDialogs as GObjects.
https://gitlab.gnome.org/GNOME/gnome-control-center/-/issues/1126
When the search field in the "Add Printer" dialog is edited, no
authentication with a remote SMB server should be attempted. The server
should just be added to the list and marked with "Server requires
authentication". The user can then authenticate the server by clicking
on it.
The behavior described above is probably the intended one when the code
was written, since there is no closure registered for the
"authentication-required" signal in the function this commit modifies.
This commit should simply restore the intended behavior.
Resolves: https://gitlab.gnome.org/GNOME/gnome-control-center/issues/755
Add a scheme to the address which we test for correctness
by g_network_address_parse_uri(). It does not work without it.
Use "none" scheme if user did not entered one.
Use port number 0 if user did not specify any.
Fixes#679
The "Add Printer" dialog should be smart enough to know whether
an item listed in the dialog is a samba server or just a printer.
If it is a samba server, it should go for the authentication page
instead of emitting a GTK_RESPONSE_*.
https://bugzilla.gnome.org/show_bug.cgi?id=778277
