News for hackers. Stuff that matters.
Re:Trump’s F.C.C. Pick Quickly Targets Net Neutrality Rules by Yizahi on Tuesday February 07, 2017 @05:23AM (#13588401)
Wireless is so so much harder and more expensive. I live in an average size apartment built from reinforced concrete panels and it is hell to setup wifi. Basically 2.4 GHz signal halves after each wall. 5 GHz is worse. I have two APs for 3 room 60 sq.m. apart and there are still dead zones everywhere.

Same for GSM - if you are too high, too low, too far or in shadow from tower it just doesn't work properly. Holiday overload on the cell - goodbye internet.

Wireless is barely functioning as it is and more customers will just make it worse.

PS: also when something in the last mile is not working it is insane to debug it. Phone disconnects for some reason and that's it. Router is working but devices can't connect or auth. Some APs work and some aren't. My devices from same vendor and same model lineup can't even properly restart by themselves after power outage. I need to go and reboot them in order one by one each time. I hate wireless... :)

Re:Trump’s F.C.C. Pick Quickly Targets Net Neutrality Rules by 0xcde4c3db on Tuesday February 07, 2017 @09:54AM (#13589712)
I think the fundamental problem is capacity. I haven't run a WISP, but my understanding is that a highly directional rooftop antenna solves most signal strength problems for individual subscribers. However, there's a practical requirement for the tower to have antennas with broader coverage (I think WISPs conventionally use sector antennas), because otherwise you'd just have too many antennas that need to continually be adjusted as subscribers come and go. Half-duplex medium and protocol overhead means that you're not going to get very many stations on a given combination of antenna and channel before performance starts taking a major hit. I've seen general guidelines that recommend no more than 20 stations per AP; I assume a WISP tower setup can optimize for more stations per channel, but I'd expect that to be on the order of a 2-5x multiplier, not 100x.