Hi Sascha, Quoting Sascha Kruse (2012-08-18 15:07:34) > I've written a patch for ticket #719 [1]. At this point this is more or > less just a prototype to see if this feature is usefull or not. So > please let me know if this is a sane approach for this and what needs to > be done to get accepted. Thanks for your patch. Here are a few observations:
1) s/a urgent/an urgent/ ;-) 2) nodes_head only contains tiling children. Since floating children can only be attached to a workspace and your changes don’t have any effect in this situation, this is actually correct. Nevertheless, it makes sense to think about whether to use nodes_head, floating_head or focus_head (which contains both, but in a different order) whenever you are using recursion. Therefore, please add a comment explaining the situation. 3) In src/x.c, you check if con_is_leaf and then call con_has_urgent_child. Since the latter also checks that, it seems superfluous. 4) Thinking about how you do it (setting urgent to true), I think it’s better to do this once (instead of on every draw_decoration call) when the urgency state changes, see src/handlers.c:handle_hints:819. It’d be worthwhile to have a testcase for this, too (or expand t/113-urgent.t). Please have a look at http://i3wm.org/docs/testsuite.html Thanks for working on this. Best regards, Michael
