Abolishing DNS

that makes it sound even more based. fuck censorship and internet policing.

>>And yet you don't mention the most obvious problem
it's a big one if you know what all that means, but not the most obvious, in layman terms.

since OP is clearly an idiot, and said himself he doesn't know network, I didn't want to be too technical.

If domain names suddenly became IP addresses, how do you distinguish one server from another that are represented by the same name?

the average zoomer really thinks there is one big computer for the whole world answering to facebook.com

Routing tables wouldn't be impossible, but they would be ridiculously massive.
No more structure to how IP space is broken down means that you have to have every single IP in use in your routing table and anyone can claim any address to be their own.

So if you claim tinycock.net as your address, and some other girthless individual across the world does the same, how does the internet know which one to route to?

> pedos
oh those nasty pedos!
> it would take ages to find
and you claim you're a network engineer? you're just a spastic fraud that couldn't network together a router and a laptop. take your gay larping back to facebook.

first.learn how ipv4 works, because you dont understand things like routing and lookup tables

also look into things like namecoin and i2p, but realizethe failings of conflating dpmains with ip addresses

I believe that the reason IPv6 has taken forever to catch on is because some autistic faggot decided that we needed to jump all the way to a 128 bit number, represented in a disgusting hex string.
a 48 bit number would give 281.4 TRILLION different IP addresses, and would look either like the MAC addresses we use today, or would just have 2 more octets on a IPv4 address.
What's easier to use?:
>69.42.192.168.1.1
or
>2001:0db8:85a3:0000:0000:8a2e:0370:7334

This is already a thing. See: en.wikipedia.org/wiki/Content_centric_networking

The problem would still be name assignment.

You are replacing one regulatory with another for desirable names. That being said - It's not the worst idea I've ever seen this being said - that might be IPv6

And the part you are missing is the human element. IPv6 is a committee disaster, every SUPERDUPER SPACE BRAIN involved put in their 2 cents and by their powers combined we ended up this fucked up flipper baby of a monstrosity that did everything except fix the problem it needed to.

Probably went something like -
"We need more IPs!!"
"I know, we can make it 64 bits!"
"No we should make it 128 bits and drop half the range on other technologies that are not at all needed!"
"We should also add in all kinds of other shit that no one wants or needs!"
"Guys, we just needed more ip add.."
"FUCK YOU GET OUT! MY IDEA IS BETTER THAN YOURS!"
"NO FUCK BOTH OF YOU, MY IDEA IS BETTER THAN ANYONE ELSES AND I WILL TAKE MY BALL AND GO HOME IF I DON'T GET MY MAN BABY WAY!!!"

What if we had a browser extension that translated names into IP addresses, except that the table would be stored in a blockchain where users could register a name by spending coin and mine coin by keeping a server up that would also serve the blockchain itself. Then we could bypass DNS servers altogether in a decentralized manner?

What is the problem with ipv6?

Good idea. Similar to the wiki system for advertising some hidden services

>wiki system for advertising some hidden services
Can you elaborate? I don't know much about Tor.

Point one and two are literally arguments FOR not against.
As for point three, you could design in it a way to say
domain.subdomain.com, which makes more logical sense anyways.

You wouldn't be able to claim the same domain, theoretically. You could make it work like tor domains where you have to mine for a domain name if you don't like the default rng one.

Hex is good desu. But they should have just had it 4x4 or 8x2 instead of 8x4. (2001.0db8.85a3.0000). We literally don't need more addresses since devices can interconnect limitlessly in the same local network.
I remember one of the things being shilled was muh IoT devices need ipv6, but you can already use those fine in ipv4 and there's no clutter since every router supports at least 16 devices on the same network, and if it doesn't you can just get another router or repeater.

Pretty much everything. There are no doubt millions of articles on it at this point, but most of those lead to the points of implementation cost, lack of compatibility, and over complication in design

There is at least one wiki that you can easily find that lists other sites you can go to
Alright, I'll look into it. I've just been adopting it on my home network and the experience has been pretty good except for Alpine not supporting dhcpv6 (??)

IP packets have to be routed, which means they have to be assigned to an endpoint somewhere.

In your system, the hash of your domain would be fixed on 1 IP. You'd be stuck with whoever's your host. Now you just have IPs but with words instead of numbers.

We don't *need* more addresses if we get clever with NAT but it just creates more and more problems. NAT also sets a hard limit on # of connections with the port allocation system. It also limits the ability of hobbyists to host services.

128 bits was a good idea and right thing to do

Don't get me wrong, if it works, it works. Just don't be surprised if there are some compatibility issues, and the setup is more complicated than it needs to be.

>and over complication in design
it's way simpler than IPv4 and cuts lots of corners in the design

It's hard to judge whether the setup is "too complicated" for me. I just use pfsense in assisted mode and it does all the RA settings for me. I agree that it could be more complicated than it needs to be to support a given set of needs but ipv6 is designed with the aim of supporting a lot of features that ip is missing, and for that matter all the software for it is a lot less mature and thus simple to set up (vs something complicated like NAT being mature and autoconfigured like everywhere)

>Point one and two are literally arguments FOR not against.
I'm saying why it won't be done
Not if it's good or bad.

riiiiight. Yeah, you have fun with that.

>switch hosting provider
>can no longer use your domain name.

>upgrade your server
>can no longer use your domain name.

>server breaks, have to use backup
>can no longer use your domain name.

Yeah, great plan....

>What is the problem with ipv6?

Not much apart from privacy issues (which are fixable).

You said pfsense, then you said NAT and complicated.

I assume you must be running an HA cluster with multiple VIPs and routes with required levels of separation for services and such both inbound and outbound.

Because otherwise I have no idea what you are doing with NAT that is making it so complicated. pfSense has the most functional and simplistic NATs I have ever used. It works for everything even if you stack the NAT like 30 times and hairpin the shit out of it across multiple interfaces - which I have done and it still worked like a champ.

NAT is internally complicated
Setup is easy as fuck to set up but the actual system is arcane imo. I'm saying that NAT is complicated but implementations areature enough to make it easy even for me.

Yeah I privacy extensions are on by default in Ubuntu and nixos, I assume most other desktop distros

You mean 65000 sessions per remote host?

While I have no argument that this is a scalability issue at some level, for most instances where you would be actually using NAT this would not be an concern.

You're legitimately retarded

Based zoomers.
If future generations are that stupid they will never be any real competition in the job market.

Compared to the multi headed dick hydra that is IPv6 and the required additional hardening and implementation at all other points such as operating systems, switching and routing planes, ect - NAT is crazy simple.

namecoin called, she said youre 8 years late
and i2p is glaring at you like shes gonna kill you

>there is a global push since the early 2000s to turn internet into the new television, aka the big producers produce and you consume (see netflix)
How is this being pushed and what can we do to fight it?

You want to fight Jow Forums? Perhaps reddit, fark, netflix, youtube, twitter, ect ect ect?

They produce a platform, you both consume and fill the platform at the same time. The game has changed, but it is still the game.

They didn't need to make it shit like TV, they just needed to jump on board the train and buy the pieces they wanted, then slowly add the shit for their business interests. Why do you think you see so many instances of sites and such trying to demonize ad blockers as stealing and criminal? The fight to stop them was over LONG ago, now all you can do is whip up the mob to go after them for trying to make it worse - and just a lesson from history here, they will just re-brand it, change it a tiny bit and try again and again until it sticks.

But how can we fight attempts to limit user upload rights? Net neutrality I know is one issue

complications from trying to achieve compatibility with IPv4
without this reasonable goal, IPv6 would be strictly simpler and better than IPv4

... you serious? Seems like you might be fuckin' with me. screw it

just don't use platforms that do that, bam, you fought it.

Well that or you could resort to violence or destruction of property (I am pretty sure you wouldn't accomplish anything good with this), or find a way to make their efforts self destructive to such a degree that they don't realize it before its too late - then again the woke movements already appear to be doing this, and actually seems to only makes things worse in general as everyone flails and shits their stupidity all over everything.

You can short cut hex though with ::

>but it also addresses ISP/MITM privacy concerns by taking the resolution step out of the equation.
it doesn't such network doesn't prevent MITM in any way

It tries to fix everything at once and ends up being just another huge clusterfuck.

who would assign the IP addresses? A DNS server, perhaps?

Could you give some examples of hardening one has to do?

we could modify the protocol so that you still own subdomains

not him but:
filtering of ICMP packets is overall a topic in IPv6 networks
translation of from local IPv6 network to IPv4-only service on Internet requires manipulating DNS requests. Citing Satrapa's IPv6 book:
>everything starts request of type AAAA sent from local machine to local DNS server
>that first tries to forward the request in unchanged form. if it succeeds then communication can continue in IPv6 form
>on failure, DNS server that implements DNS64 tries to send the request of type A for the same name
>obtained answer with IPv4 address is translated to type AAAA and modifies the address using the Prefix64::/94 mechanism (prefix reserved by network manager for purposes of NAT64, can only be single address if only single NAT64 server on the network)
>client obtains an address, NAT64 will translate it to the IPv4 address
issues:
>DNS64 modifies DNS request, thus is not compatible with DNSSEC verification directly on client
>there is workaround flag that causes DNS64 server to return the type A request directly without translation, this allows the client to verify the DNS request and implement DNS64 on the client, but this requires to know the Prefix64::/n
>there is mechanism for its discovery, described in RFC 7050: Discovery of the IPv6 Prefix Used for IPv6 Address Synthesis
there is also described abandoned predecessor NAT-PT, which used to have many issues with the DNS. however NAT-PT was attempting the provide the option to open the connection from outer network to inner.

rest is done stateless with SIIT mechanism, which is simple
it's just OS devs are unable to manage their growing feature creeps and implement IPv6 support without bugs, not that there would be some serious issues in IPv6 itself

This doesn't sound routable.

IPv6 have unlimited network capacity
what do you suggest?

You'd still need a new IP system.

No I mean upload restrictions from regulation and providers. The current trend is to cut user bandwidth and reallocate that to big services, creating cable 2.0. What can we do?

Not him, but WC3 adding DRM to HTML5 comes to mind.

It doesn't prevent all forms of MITM attacks, but it sure prevents DNS spoofing.