Why I'm running ethernet in my home


64 bytes from icmp_seq=30 ttl=51 time=86.784 ms
64 bytes from icmp_seq=31 ttl=51 time=115.341 ms
64 bytes from icmp_seq=32 ttl=51 time=104.645 ms
64 bytes from icmp_seq=33 ttl=51 time=81.736 ms
64 bytes from icmp_seq=34 ttl=51 time=90.551 ms
64 bytes from icmp_seq=35 ttl=51 time=71.943 ms
64 bytes from icmp_seq=36 ttl=51 time=32.487 ms
64 bytes from icmp_seq=37 ttl=51 time=76.986 ms


64 bytes from icmp_seq=146 ttl=51 time=18.077 ms
64 bytes from icmp_seq=147 ttl=51 time=21.553 ms
64 bytes from icmp_seq=148 ttl=51 time=18.745 ms
64 bytes from icmp_seq=149 ttl=51 time=18.092 ms
64 bytes from icmp_seq=150 ttl=51 time=18.556 ms
64 bytes from icmp_seq=151 ttl=51 time=18.921 ms
64 bytes from icmp_seq=152 ttl=51 time=18.135 ms
64 bytes from icmp_seq=153 ttl=51 time=18.109 ms

(ok, there are other reasons; gigabit wirespeed instead of the 50Mbps common on wireless networks (I didn't actually measure mine). Finally, my printer has ethernet but no wireless; by running cable to it and the WAP, I can print from anywhere in the house! :) Really, though, the latency and consistency are what I appreciate.)


germee said...

I would never discourage you from running cable. I have a full compliment of wiring tools and a box of CAT5e a few feet from me right now, but I think it's time to replace your wireless gear.

My ping times to the same target through a NETGEAR WNR3500L running dd-wrt and an ATT DSL line:

--- ping statistics ---
100 packets transmitted, 100 packets received, 0.0% packet loss
round-trip min/avg/max/stddev = 15.543/17.538/39.361/2.398 ms

--- ping statistics ---
100 packets transmitted, 100 packets received, 0.0% packet loss
round-trip min/avg/max/stddev = 14.697/15.474/17.810/0.376 ms

ben said...

Though I wouldn't be surprised if replacing my wireless helped, I'm putting the blame on the lathe-and-plaster construction in my home. Chicken wire in all the walls does a marvelous job of making wifi suck.

ben said...

To follow up, I did replace my shitty wifi router with a Ubiquiti Unifi AP AC Lite. It made a big difference in many parts of the house, but the walls still block too much signal. It falls off really quickly in the bedrooms and on the back porch. I look forward to getting a few more Unifis to mesh, which will make it better, but I still support cables for permanent fixtures (like a desk). It's even easier with thunderbolt, since the ethernet cable can connect to the monitor and not be yet another cable to plug into your laptop when you settle in to the station.