> Perhaps, but ... Isn't that what I wrote: > > it's probably not possible to realize a latency win here ? > On top of that, if you used some kind of pulse

> Perhaps, but ... Isn't that what I wrote: > > it's probably not possible to realize a latency win here ? > On top of that, if you used some kind of pulse scheme (i.e. morse code with neutrinos) it has to be slow "slow pulse" == long pulse. It will be "fast" in that it will go faster than the speed of light in fiberoptics and the path will be shorter, but it will be slower because error correction will demand a great deal of redundancy which, among other things, means long pulses. I think we're in agreement.