/tech/ - Technology

Why didn't le internet authorities just utilize a v3 onion addressing style solution to solve the address exhaustion apocalypse so they could keep using the IPv4 backbone?

>>3980
onion addresses are used like this https://tor.stackexchange.com/questions/672/how-do-onion-addresses-exactly-work . To use them for ip would require some sort of authority to keep records of the real ip, so it can't solve the address exhaustion problem.
ipv6 could have been just a longer ipv4 address or even just have two ipv4 address for each peer would have created so many more addresses. They have to add so many more features, including adding your mac to your ip by default (https://superuser.com/questions/243669/how-to-avoid-exposing-my-mac-address-when-using-ipv6).

4afffc31a40ba2830c4fa1d0f6ac9f6ba4342080743690d5bae47929d09bb828.PNG
[Hide] (728.8KB, 599x599)

nobody i know uses it, i have it disabled by default and all the networks i've used all still use IPv4, it's not going to come into effect for a while but you'll probably want to learn it before the end of the decade. it's needlessly complicated and instead of 4 numbers we now have to throw in hexadecimal and colons everywhere. it's going to be a pain in the dick.

>>3982
IPv6 is a foot-in-the-door for less privacy. While all the desktop OSs may have the privacy extensions, it's less likely all the dumbsumer IoT devices will, but they are all Trojan horses anyway, so bad actors will already have access to your network if you use them.

>ipv6 could have been just a longer ipv4 address
That would have been nice. When now hearing the word features, I cringe and hope whatever they are don't cause too much damage.

2b4984f0dad413d68f44557f52060296115c9091bff24ab23041aa4a15afca3b.gif
[Hide] (11.7KB, 300x220)

>>4022
You are more likely to see it at the ISP level and above. A lot of them have already switched and use translation. There is not any advantage to end users implementing it. It is very unlikely end users would exhaust usable network IPs, unless evil technology like IoT dust and other wasteful things need vast numbers of addresses. But why would anyone want those things unless they are retarded?

So give me some examples of how an end user could be fingerprinted and tracked using this. Does Tor mitigate things or does IPv6 introduce some extra risks and fuckery?

>>4034
Try testing with some ipv6 test sites with your friend's computer and with Tor, eg: https://ipv6-test.com, Tor should protect you if all of your connections are going though it.

>>4027
>IPv6 = less privacy
Is it true, though? The CIA niggers and/or owners of the server can track your location if you use IPv4 anyway?

>>15248
No and read the date.The author of the post definitely won't reply to you three years later.

>>15248
>owners of the server can track your location if you use IPv4 anyway?
Yes but you probably share that ipv4 address with other customers from the same internet service provider. And then the connection terminates at a router which you might share with other people in the building.

>>15319
Yes, nat provides a layer of separation.
>>15309
ipv6 doesn't need nat because there are enough addresses for each devi e to have one. There is an rfc for randomizing it https://datatracker.ietf.org/doc/html/rfc8981 .

>>3971 (OP) 
The biggest issue is that they didn't just extend the address space and drop the redundant bits (like optional source routing, which nobody used anyway), they tried to introduce a bunch of new mechanics which fuck things up.

For example, SLAAC: nice in theory, but garbage when your ISP is retarded and only gives you /64, preventing you from making subnets which also work with Android (thanks Google!). I've even heard of retardation like giving out /128 only!
But even if I'd be okay with not being able to use my phone on my own subnet I can't figure out how to make the router actually give out a subnet, so I simply don't use IPv6 and just sit behind a double-NAT IPv4, which was substantially easier to set up even if it's shit.

But what I consider to be an even bigger issue is that IPv6 can and still will suffer from (severe) routing table fragmentation, as you can get some arbitrary IPv6 range and use it pretty much anywhere (on the same continent, at least, not sure what the exact requirements are). I believe the only real solution to this is to map addresses to physical coordinates and enforce that an address maps relatively close to its corresponding coordinate.
Hell, everyone uses DNS anyway and IPv6 obviously isn't meant to be typed in directly, so why bother reserving specific ranges to specific entities instead of regions?