Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

72% of smart TVs and 46% of game consoles hardcode DNS settings

labzilla.io

Post reply on HN
Print view View original

241–250 of 673 posts

Re: 72% of smart TVs and 46% of game consoles hardcode DNS settings

#241

Post by mstade »

gorbypark wrote: post #169
mstade wrote: post #157

Earlier quoted context omitted.

The IKEA Tradfri "smart" lighting gateway will stop responding to commands if it can't phone home to some IKEA server. I noticed this when I changed my router to use NextDNS, which blocked the IKEA lookups. I was ready to return the device as broken until I realized this. I've also had issues with Bang & Olufsen speakers in the past, and inclined to believe it's for the same reasons. I think it's insane that devices…

When I installed PiHole a few years back I blocked my tradfri gateway from connecting to Ikea's servers and everything kept working! I wonder if something has changed since then? Ikea devices are kind of nice because they don't actually rely on the internet at all and work completely locally (at least, they did a few years back).

All I can say is when I had NextDNS configured on my router it blocked requests to some IKEA domain, possibly smetrics.ikea.com from a cursory search through he logs, and my Tradfri gateway would just straight stop responding to anything at that point. I googled around for a while and found other people having issues with DHCP and QoS with Tradfri gateways, so I made sure it had a static IP set as well as all QoS “features” being disabled, but this didn’t help. It would work at first, for some period of time (30 min maybe?) and then stop responding. Once I saw the blocked DNS lookups I disabled NextDNS on the router and flushed any caches on the router, rebooted everything and it’s worked fine now for a good month or so.

I will admit I haven’t done any further investigation, but simply concluded that the gateway at some point started phoning home and if it didn’t receive a response went into some catatonic state. Maybe I’ll dig deeper at some point, time permitting.

Re: 72% of smart TVs and 46% of game consoles hardcode DNS settings

#242

Post by mstade »

benttoothpaste wrote: post #166
mstade wrote: post #157

Earlier quoted context omitted.

The IKEA Tradfri "smart" lighting gateway will stop responding to commands if it can't phone home to some IKEA server. I noticed this when I changed my router to use NextDNS, which blocked the IKEA lookups. I was ready to return the device as broken until I realized this. I've also had issues with Bang & Olufsen speakers in the past, and inclined to believe it's for the same reasons. I think it's insane that devices…

> I was ready to return the device as broken until I realized this. Actually you probably should return such devices as broken.

I’m inclined to agree, frankly.

Re: 72% of smart TVs and 46% of game consoles hardcode DNS settings

#243

Post by throwaway201103 »

t0astbread wrote: post #24
EvanAnderson wrote: post #19

Earlier quoted context omitted.

You're right to be suspicious. The DNS-over-HTTPS model favors those who run the servers (because they get exclusive access to monetizable end user name resolution data) and those who control the resolvers. You might control the resolver on your personal computer (for now). You probably don't control it on your phone. You most likely won't control it on your embedded devices.

The root evil here is that you can't change the root certificates in such devices. Even if you controlled its DNS, the device could still just be programmed to fail if it doesn't reach its analytics/ad/whatever server.

Then you return the device as defective and demand a refund. What you chose to pay for is up to you.

Re: 72% of smart TVs and 46% of game consoles hardcode DNS settings

#244

Post by ozaark »

amelius wrote: post #201

If you can spare the money, buy a Smart TV and send it to the FCC accompanied by a letter saying that you have no use for a TV that is invading your privacy and there are no alternatives on the market. If they get enough TVs, perhaps they will take action. (If you have less money to spare, buy the TV and return it to the manufacturer for a refund, make sure you state the reason.)

FCC doesn't seem to be on consumer side - at least until Pai and his ilk are out unfortunately. They showed that with NN repeal despite mass public disapproval.

Re: 72% of smart TVs and 46% of game consoles hardcode DNS settings

#245

Post by Ekaros »

sircastor wrote: post #182
mikro2nd wrote: post #138

Earlier quoted context omitted.

I live outside the cellphone coverage area. Your move, manufacturer.

... for now. Cell coverage isn’t shrinking.

I actually is. Or at least the individual cells are... At some point it will be fun to install cells in your own home...

Re: 72% of smart TVs and 46% of game consoles hardcode DNS settings

#246

Post by anonymousab »

grishka wrote: post #174
therealx wrote: post #117

Earlier quoted context omitted.

Sadly, some will look for other open wireless netorks and connect. In the future, I'm sure all will have a 5g connection.

At this point you could as well open it up and disconnect the antennas.

Well naturally they'll just not work if they don't have a connection or detect a hardware fault.

Re: 72% of smart TVs and 46% of game consoles hardcode DNS settings

#247

Post by ecf »

supertrope wrote: post #237

Hard coding DNS removes a troubleshooting variable and decreases customer support load and return rates. Some ISPs hijack DNS to serve up search ads (when it’s not down or slow). Some routers are part of botnets so DNS is not to be trusted.

Perhaps companies should hire more talented customer service representatives who have a technical background instead of continuing to make their products shittier.

Re: 72% of smart TVs and 46% of game consoles hardcode DNS settings

#248

Post by jon-wood »

zinekeller wrote: post #127
dingaling wrote: post #120

Earlier quoted context omitted.

So you'll only start working on the problem when IPv6-aware devices start failing? Why wait?

Not the OP, but you should also be reminded that many of these limitations are brought out by cost-cutting measures that we didn't have control of and could be used to fire us immediately, and even when IPv6 is implemented most IoT backend servers only operate over IPv4 anyway. Another one that you might be angry about is only including 2.4GHz-band WiFi even though 5GHz WiFi is already ubiquitous (although only inclu…

Exactly this, we’re a small team and have to pick our battles. In practice very few people have IPv6 only connections, and even fewer of the people who buy our products.

Re: 72% of smart TVs and 46% of game consoles hardcode DNS settings

#249

Post by zeta0134 »

ecf wrote: post #247
supertrope wrote: post #237

Hard coding DNS removes a troubleshooting variable and decreases customer support load and return rates. Some ISPs hijack DNS to serve up search ads (when it’s not down or slow). Some routers are part of botnets so DNS is not to be trusted.

Perhaps companies should hire more talented customer service representatives who have a technical background instead of continuing to make their products shittier.

Having worked in this kind of customer support, the issue isn't that technician doesn't understand DNS, but rather that the end user frequently hadn't even heard of it. It's a massive pain to troubleshoot something like that remotely when the caller is non-technical. Not impossible of course, but frustrating and requiring a great deal of patience on both ends.
Post reply on HN