Google, Microsoft, Comcast Say Verizon’s New Cellular Tech Could Wreck Wi-Fi • Hacker News Board

Re: Google, Microsoft, Comcast Say Verizon’s New Cellular Tech Could Wreck Wi-Fi

#181

Post by indlebe » Tue, Dec 08, 2015, 7:06 PM UTC

gst wrote: post #163

wtallis wrote: post #155

Earlier quoted context omitted.

> " My current Comcast line is so fast that I don't experience any issues due to congestion - so how would I benefit from QoS? " Are you sure you don't have a few hundred milliseconds of bufferbloat in your cable modem, as is the case for almost all modems out there?

No - I'm not. But my connection works well enough and I don't notice any delays on either SSH or VoIP connections. So why should I invest time/energy into changing something that doesn't affect me?

> So why should I invest time/energy into changing something that doesn't affect me?

This argument has a place here, but it has a cringe effect each time I see it. I find it a dangerous one if applied to other circumstances.

Re: Google, Microsoft, Comcast Say Verizon’s New Cellular Tech Could Wreck Wi-Fi

#182

Post by jrockway » Wed, Dec 09, 2015, 8:25 PM UTC

simoncion wrote: post #179

jrockway wrote: post #175

Earlier quoted context omitted.

Regarding footnote [2], this is where 5GHz really helps. With 80MHz channels, you'll be using 1/4 the airtime to send the same amount of data. 2.4GHz is to be avoided if at all possible. Also, measuring SNR is interesting, but you should really run a speed test of some sort. A team I work with wrote this to let access points speedtest clients without any code on the client: https://gfiber.googlesource.com/vendor/goog…

> Also, measuring SNR is interesting, but you should really run a speed test of some sort. No doubt. Next time I find myself in something similar to my apartment, but in the middle of nowhere so's the three or four-dozen 2.4GHz APs that surround me don't confound the results, I'll do just that. :) FWIW, the SNR measurements were taken with a UBNT SR-71E (802.11agn ath9k-driven) client that Linux reports as an AR928X.…

I think the open source ath10k driver is in a pretty good state. We're using it for the Google Fiber routers, and it's working pretty well. OnHub is also using ath10k.

That said, now that the drivers are working, it's old technology. 4x4ac is all the rage now, and ath10k is 3x3.

Re: Google, Microsoft, Comcast Say Verizon’s New Cellular Tech Could Wreck Wi-Fi

#183

Post by manigandham » Thu, Dec 10, 2015, 1:56 AM UTC

wtallis wrote: post #156

jlarocco wrote: post #97

Earlier quoted context omitted.

I was going to say this, but you beat me to it. I think there are a lot of smart people on HN, but that doesn't mean they're all interested in all the same things. Tweaking wifi settings just seems like a waste of time to me, and boring to top it off. My interaction with my AP stopped at setting the ssid and passphrase.

Tweaking QoS settings is largely a waste of time nowadays, but only if your router is running an OS new enough to include modern self-tuning QoS algorithms (and preferably a community-maintained project like OpenWRT, because the commercial vendors screw up their deployment of said algorithms). Even if you aren't going to be hacking on your router much, it still definitely pays to ensure you're using hackable hardware…

> Even if you aren't going to be hacking on your router much, it still definitely pays to ensure you're using hackable hardware.

This seems to contradict itself - how does it pay if you're not going to be hacking?

Re: Google, Microsoft, Comcast Say Verizon’s New Cellular Tech Could Wreck Wi-Fi

#184

Post by wtallis » Thu, Dec 10, 2015, 2:21 AM UTC

manigandham wrote: post #183

wtallis wrote: post #156

Earlier quoted context omitted.

Tweaking QoS settings is largely a waste of time nowadays, but only if your router is running an OS new enough to include modern self-tuning QoS algorithms (and preferably a community-maintained project like OpenWRT, because the commercial vendors screw up their deployment of said algorithms). Even if you aren't going to be hacking on your router much, it still definitely pays to ensure you're using hackable hardware…

> Even if you aren't going to be hacking on your router much, it still definitely pays to ensure you're using hackable hardware. This seems to contradict itself - how does it pay if you're not going to be hacking?

Because you get the benefits of those who do hack, simply by installing a recent stable version of OpenWRT. It's no harder than upgrading and configuring the vendor's firmware. If you tie yourself to vendor firmware, you're tying yourself to 5+ year old kernels and all the associated security and performance and stability problems and a much more restricted feature set.

The commercial vendors do an absolutely horrible job of supporting or maintaining their products and they tend to get a lot wrong with the initial software release. The OpenWRT community does a great job of putting out a solid product that works and has sensible defaults, but you can only get the full benefit of their work if you buy hardware that is open to their hacking.