All things being equal, could waypoint based ipv4 have been a viable alternative to ipv6?

  • 🐕 I am attempting to get the site runnning as fast as possible. If you are experiencing slow page load times, please report it.

Betonhaus

Irrefutable Rationality
kiwifarms.net
Joined
Mar 30, 2023
right now ipv6 requires all sorts of translation systems to work with ipv4 and still isn't fully supported or adopted. What if you could simply add the gateway to the ip address, so that the first ip address would be the ip to your ISP and the second ip address would be to your home network, so it would look like 23.54.234.132;54.42.124.234? All traffic from your local isp to your neighbourhood would have to have to go through the same hub anyways, so migth as well have an ip address associated with that hub then a secondary one for your home. you could possibly also have a third partial one inside your network, so 23.54.234.132;54.42.124.234;53 so that you could have multiple servers on your home network with their own dns name and ip address that your router knows how to route (so traffic that goes to ;53 would get redirected to the device at 192.168.0.5, or however your router is configured to send it to as giving direct access seems a bad idea)

how viable would a concept be?
 
So, there are quite a few potential issues and risks:

Compatibility and Standardization

  1. Protocol Standards: The Internet runs on standardized protocols. Introducing a new IP format would mean making significant changes to these standards, which would be a massive challenge globally.
  2. Legacy Systems: Many existing devices wouldn't recognize or correctly interpret this new format, leading to compatibility problems and possibly causing network failures.

Routing Complexity

  1. Routing Table Management: Routers would need to handle this new format, complicating their management and lookup processes.
  2. Increased Latency: More complex addresses mean routers need more processing time, which could slow down network performance.

Security Implications

  1. Address Spoofing: Multiple address segments could make it easier for attackers to spoof addresses, complicating network security.
  2. Attack Surface: A more complex addressing scheme might increase the potential points of attack, making it easier for hackers to find and exploit weaknesses.

Network Administration

  1. Configuration Complexity: Network administrators would have a harder time setting up and managing devices with this new addressing scheme.
  2. Human Error: The more complex the configuration, the higher the chance of mistakes, leading to potential network outages and misconfigurations.

Interoperability Issues

  1. Transition Mechanisms: The new format would need mechanisms to ensure it works with both IPv4 and IPv6, adding another layer of complexity.
  2. Vendor Support: Networking equipment and software vendors would need to update their products to support this new format, which would take time and resources.

Scalability

  1. Address Space: While IPv6 provides a vast address space, this new scheme doesn’t solve the problem of IPv4 address exhaustion.
  2. Hierarchical Addressing: This idea might not scale well with the rapidly increasing number of internet-connected devices.
 
It would have been nice if the next ipv standard was just the old one but with bigger addresses and maybe a hex based notation system to keep the length down (like 4 hex blocks of 4 for 64 bit so it would be like dead.beef.1488.c0de or omething which is actually reasonably sized/writable instead of the horrendous BS they came up with)

Instead it was like 'what if we stuffed all this faggot oppression aparatus stuff in that seems like the biggest problem to solve is figuring out who is making rude internet posts and where they are located'
 
IPv6 was designed for really fast hardware-based routing. Core routers generally drop more complicated packets, but the basic fields are aligned for very fast identification. Your proposal does not have that benefit.
 
IPv6 was designed for really fast hardware-based routing. Core routers generally drop more complicated packets, but the basic fields are aligned for very fast identification. Your proposal does not have that benefit.
It identifies the waypoint that the core router needs to send the packet to and the waypoint does the rest. It would be faster then having to look up billions more up addresses Individually
 
  • Thunk-Provoking
Reactions: Vecr
Could you elaborate on this? I'm a retarded caveman though so I'd appreciate it if you could make it easy to understand, please and thank you.
Off hand, every device on your house will have its own wan facing IP address so they can determine how many devices you own and what you are accessing on each one. Since ipv6 IP addresses are so long, a shorthand gets developed to assign one which can determine the make and maybe even model of various devices
 
Off hand, every device on your house will have its own wan facing IP address so they can determine how many devices you own and what you are accessing on each one.

Since ipv6 IP addresses are so long, a shorthand gets developed to assign one which can determine the make and maybe even model of various devices
So the first sentence is already the case, right? - and in order to make the newer ipv6 addresses long enough, they (potentially (probably)) will also include information about its device?

I don't like that but I wouldn't exactly call it spook shit.
I could see malicious actors pivoting to specific exploits for different makes though. But I don't know enough about how those work in the first place.
 
IPv6 wouldn't even be a thing if retards years ago hadn't decided everything that shits electrons also needed its own unique IP address. My fucking toaster does not need to be globally routable.

Fortunately my ISP still has a nice chunk of IPv4 to itself (no CGNAT), so I've nuked IPv6 on every device where I'm able to do so. My router is the only IP visible to the outside, and nothing else can be reached via that address without opening a port. And that's the way I like it.
 
My previous ISP went with CGNAT, toredo won't work at all, so I can't even get an IP address even if I wanted to. At least I get a semi functional ipv6 address from my new ISP now, though they don't officially support it.
IPv6 NAT exists, you can still do that if you want.
 
how viable would a concept be?
What you describe is similar to source routing.

With source routing you directly specify the path a packet should take, e.g. 143... = router A port 1 -> router B port 4 -> router C port 3 -> ... like a phone number.
The biggest issue is that it can be abused to create routing loops, e.g. 111111... = router A port 1 -> router B port 1 -> router A port 1 -> ...

(IIRC IPv4 has an extension for source routing, not sure though)


The biggest issue with IPv6 is that it isn't just "more bits" but it's designed by a committee which assumed that the real world is nice and tidy and ISPs dance the kumbaya with their customers and will definitely not do retarded shit like handing out /64s so that it becomes impossible to subnet while using SLAAC.
Fuck you google add support for DHCPv6 you fucking niggers.
 
  • Like
Reactions: Vecr
so that the first ip address would be the ip to your ISP and the second ip address would be to your home network
If you're just going to glue 2 addresses together to get a longer address, you're just slowly re-inventing ipv6
you could possibly also have a third partial one inside your network
... so the headers all have variable sizes? Good luck working with that at scale, and you'll go back to ...
still isn't fully supported or adopted
There is a way to "wrap" a local network behind a gateway. It's called a VPN.

Also, paragraphs nigger, use them.
 
  • Dumb
Reactions: Betonhaus
Back