Created attachment 356600 [details] [review]
page: fix logic to propagate small-screen property from driver

Since the driver is only set later to the page, and the initial value is
FALSE, the actual value was never propagated down when the driver was
set.

https://phabricator.endlessm.com/T17889

Any comments on this patch?

Review of attachment 356600 [details] [review]:

::: gnome-initial-setup/gis-page.c
@@ +119,3 @@
+      g_signal_connect_swapped (page->driver, "notify::small-screen",
+                                G_CALLBACK (small_screen_changed), page);
+      small_screen_changed (page);

this should cause no harm at least, but I fail to see where this may cause any issues since by the time a page instance is constructed the driver property will be set and the "real" small-screen value should be available.

Review of attachment 356600 [details] [review]:

::: gnome-initial-setup/gis-page.c
@@ +119,3 @@
+      g_signal_connect_swapped (page->driver, "notify::small-screen",
+                                G_CALLBACK (small_screen_changed), page);
+      small_screen_changed (page);

It will be set later, but things that depend on the small screen property (such as icons that get that value updated through a GBinding) won't get the property updated, hardcoding it to FALSE.

Review of attachment 356600 [details] [review]:

::: gnome-initial-setup/gis-page.c
@@ +119,3 @@
+      g_signal_connect_swapped (page->driver, "notify::small-screen",
+                                G_CALLBACK (small_screen_changed), page);
+      small_screen_changed (page);

oh I see, because they connect through a GBinding that's created at _init() time via gtk_widget_init_template() and properties are only available afterwards. ok, can you add this bit to the commit message? thanks

Attachment 356600 [details] pushed as b3223e6 - page: fix logic to propagate small-screen property from driver

Thanks, pushed to master.