Re: [LAD] wfs streaming project report

nescivi wrote: very rarely we did indeed have packet losses across the laser link, but since they were so few and far between (even in bad weather), i don't have reliable data. one possible weakness in the whole scheme is that the UDP redundancy methods of both jacktrip and netjack will send redundant packets right next to each other, so that if you have a burst failure (which is common), you are screwed.

for me, the morale is: lasers can be made reliable enough if you can tolerate the occasional single or short burst packet loss (loss rates of about 0.0001%), the general internet cannot, unless you get end-to-end QoS, but you can sneak past that if you have lots more bandwidth than you are going to need. but nothing in the world short of http streaming will protect your ass against crappy border gateways and switches that barf on udp stream traffic. as to bird losses, the class 3 lasers operated at 8mW, so the chance of a bird being vaporised is, ahem, slim. the main issue with respect to laser safety was eye damage. the minimum safe distance to look directly into the laser was about 50m. but since IR lasers do not trigger a lid-closing reflex (you only see a dim red shimmer), this minimum safe distance is determined for a duration of 100s of continuously staring into the lens (for lasers in the visible range, this duration is below 1s, iirc). so you would actually have to hover in front of the laser for well over a minute before your retina takes serious damage. therefore, we can safely rule out eye damage to birds as well, unless they are very skilled flyers, very bored and very very stupid.