WebKitWebView has an action signal "select-all" that when called selects everything in the current page. The signature is this: void select-all (WebKitWebView* web_view, gpointer user_data) By comparison, GtkTextView has the same signal, but with a different signature: void select-all (GtkTextView* text_view, gboolean select, gpointer user_data) If the boolean "select" is TRUE, everything is selected, which is what WebView currently does. But if it is FALSE, everything is unselected. It is actually possible to call "select-all" without a boolean on a TextView but the behaviour is the reverse of what WebView does. I think we should really fix the signature of WebView to match TextView. I think it's not probable that it breaks anything because I don't think that one usually connects a callback to that signal.
Created attachment 25445 [details] Adjust select-all as suggested This adjusts select-all as suggested. Unless anyone actually connected to the signal everything will work fine, at least in the C API. I don't really know about bindings, maybe someone more experienced in that area can confirm that.
Comment on attachment 25445 [details] Adjust select-all as suggested Alp is our gtk reviewer, no?
Comment on attachment 25445 [details] Adjust select-all as suggested (In reply to comment #2) > (From update of attachment 25445 [details] [review]) > Alp is our gtk reviewer, no? Alp isn't the sole Gtk reviewer, granted he tends to assume that position for the most part. Either way, let's take it off the queue for the moment, a discussion on IRC made clear that it isn't undisputed.