Hi, I read the following ticket listed as major priority: http://bugs.i3wm.org/report/ticket/662
First it is quite old, but that's not the point. Second I doubt that any of these features would be really useful. The "time out" is clearly a bad idea. Letting the computer deciding itself to reset a mode which has been enabled by the user can introduce some sort of desynchronization between the user and the computer state. Depending on his mood one can perceive the time differently so a fixed time out value can be sometime confusing. It ends up that there is no good fixed value for such a time out. The "bindelse" is less "evil" but it does not address the issue, for instance, if I press the wrong key, my reflex is to immediately correct this move and press the good key. In the case of "bindelse" the first wrong key would reset the mode and the effect of the next good input is undefined. So I think that this post has one solution: http://infra.in.zekjur.net/archives/i3-discuss/2011-January/000180.html The reply is clear: "This appears to be the silver bullet!" This already working solution allows a fine-grained control of which inputs reset the mode. The next reply about sticky idea is very clever and reflect what I experienced when playing with binding modes. It happens that if one key should reset the mode then there is a high chance that the whole key binding group should do it too. For now I have a bunch of "mode default" and it's fine. -syl20bnr-
