> What we do is translate throughputs - to 8 patterns if I remember correctly. > Only when the throughput cross threshold level NIC is disturbed with > changing blinking pattern. Huh ok. I suppose you want a throughput notification from mac80211 for that. And you want to define blinking within the LED framework, rather than just intensity. > Not for traffic but for example some laptops vendors wants slow blink > in unassociated state some wants led to be off. > > The problem is that led behavior in clean design would be defined by > trigger. So I would prefer to put flexibility of LED behavior in that > level. I'm not so sure. I think it'd make more sense to define state triggers (associated) and do whatever behaviour based on the state. Dunno. > Most of todays laptop has only one led for Wifi that's a problem. > I agree it has a lot of benefits to use LED framework, but it can be > pushed to driver or some module. Unfortunately I'm not sure myself Maybe an intermediate level in the LED framework could be written, something that itself has multiple LED "inputs" (LED devices) and takes multiple triggers, translating it into a new trigger that the single laptop wifi LED can be bound to. Sounds like too much complexity though... I don't really care much, but I definitely don't want to add new callbacks for LEDs to mac80211. If you need something special please do it within the LED framework. johannes
Attachment:
signature.asc
Description: This is a digitally signed message part
