If have a little doubt. Maybe this just my simple view of how this works and completely wrong. Maybe someone would care to comment?
If one would use for example the "wait_time = 400" to wait for cache to come in. This would increase the time before we start asking our readers for a CW with 400ms while we wait for a cached CW to come in or not. In case no cache hit then my local card will be asked, will also answer 400ms later than normal with a CW and then push this to my cache-ex peers.
Isnīt this a bad thing since my own generated cache will have an additional latency and then send to other cache-ex peers?
I mean, if everyone would set the wait_time to a higher value, then my own localīs generated cache will show up in others peoples cache later so they would have to increase their wait_time again. Itīs a bit like a dog chasing itīs own tail..