4 ms·

Is it time to just admit IPv6 is a failure and move on to new standard? Adoption is very slow though economical needs are already here. Instead of complaining a

Is it time to just admit IPv6 is a failure and move on to new standard? Adoption is very slow though economical needs are already here. Instead of complaining about users maybe we should do something so they actually want to switch?

Faaak 8y ago
Why would IPv6 be a failure ? It sounds like you don't know what you're talking about
- d3ckard 8y ago
  It's terrible from UX point of view. 1. 128 bit is likely an overkill. 64 bit would give more or less 2 billion addresses per every human for earth. Likely enough. 2. It's unreadable. Why the hell use hex? 3. It's been proposed in 1998. It became a standard last year. Not exactly quick adoption and it tells something. Yes, it will probably be deployed out of necessity, but it's weird nobody noticed that if something doesn't get traction for 20 years, there is probably something wrong with it.
  - majewsky 8y ago
    2. is because of 1. 3. Eh, what? IPv6 was published in 1998, as you say. What makes you think it became a standard only last year?
    - urda 8y ago
      Not OP, but for 3) I believe they are referring to: https://www.internetsociety.org/blog/2017/07/rfc-8200-ipv6-has-been-standardized/ https://www.internetsociety.org/blog/2017/07/rfc-8200-ipv6-h...
  - tambre 8y ago
    1. Easier organization. Allows logical subnetting and more efficient routing. 2. It allows you to see subnetting. It makes managing IPv6 so much easier than IPv4. 3. It became "Internet Standard" last year, because the working group decided that it was about time to combine all the RFCs with the errata corrected in a single RFC. Just a formality. As for traction... blame humans. If there's something better to replace the old thing, but the improvements aren't dramatic enough (e.g. saving tons of money) and the old thing still works, then a huge portion of lazy people won't switch.
  - urda 8y ago
    1. Perhaps it seems too much now, but if we're taking the time to migrate our global network to a new standard, we might as well get a ton of addresses out of it for many years to come. 2. Why are you reading it? Networking addresses do NOT need some special UI / UX interface. For most users, they are never going to see it. 3. Neat, but that's not really a point. If you find yourself setting time limits on about everything in life, well let's just say you're boxing yourself in for failure. Over the past 20 years we've seen the explosion of growth in devices, networking, and communication. It comes with no shock (to experts and experienced technical people alike) that it's taking a long time to switch over. There isn't some magical "press this button for everyone to be IPv6" FYI.
vbezhenar 8y ago
IPv4 worked for a long time and works for majority of installations now. It's hard to migrate when advantages are not obvious. I'm not sure that IPv6 is very good, I'm always having problem with it, but any standard will have adoption problems when current solution is good enough. I don't think that there's need for new standard and I'm sure that new standard won't be adopted better than IPv6.
- kalleboo 8y ago
  I'm not sure that IPv4 is very good, I'm always having issues with NAT etc
  - vbezhenar 8y ago
    What issues? NAT always worked fine for me. It's one iptables rule and one sysctl.
    - Rjevski 8y ago
      NAT breaks the original design of the internet where every endpoint should be reachable from anywhere. Try using protocols that rely on this (like SIP) and see how far you go.
sp332 8y ago
I think you'd be surprised if you looked up actual adoption numbers. http://www.worldipv6launch.org/measurements/ http://www.worldipv6launch.org/measurements/
- d3ckard 8y ago
  Just did. 7 operators over 90%. On 7th page you get under 50% adoption if sorted. Well, given how old is IPv6, it's a failure.
  - icebraining 8y ago
    Well, given how old is IPv6, it's a failure. Why?

Faaak 8y ago

Why would IPv6 be a failure ? It sounds like you don't know what you're talking about

d3ckard 8y ago
It's terrible from UX point of view. 1. 128 bit is likely an overkill. 64 bit would give more or less 2 billion addresses per every human for earth. Likely enough. 2. It's unreadable. Why the hell use hex? 3. It's been proposed in 1998. It became a standard last year. Not exactly quick adoption and it tells something. Yes, it will probably be deployed out of necessity, but it's weird nobody noticed that if something doesn't get traction for 20 years, there is probably something wrong with it.
- majewsky 8y ago
  2. is because of 1. 3. Eh, what? IPv6 was published in 1998, as you say. What makes you think it became a standard only last year?
  - urda 8y ago
    Not OP, but for 3) I believe they are referring to: https://www.internetsociety.org/blog/2017/07/rfc-8200-ipv6-has-been-standardized/ https://www.internetsociety.org/blog/2017/07/rfc-8200-ipv6-h...
- tambre 8y ago
  1. Easier organization. Allows logical subnetting and more efficient routing. 2. It allows you to see subnetting. It makes managing IPv6 so much easier than IPv4. 3. It became "Internet Standard" last year, because the working group decided that it was about time to combine all the RFCs with the errata corrected in a single RFC. Just a formality. As for traction... blame humans. If there's something better to replace the old thing, but the improvements aren't dramatic enough (e.g. saving tons of money) and the old thing still works, then a huge portion of lazy people won't switch.
- urda 8y ago
  1. Perhaps it seems too much now, but if we're taking the time to migrate our global network to a new standard, we might as well get a ton of addresses out of it for many years to come. 2. Why are you reading it? Networking addresses do NOT need some special UI / UX interface. For most users, they are never going to see it. 3. Neat, but that's not really a point. If you find yourself setting time limits on about everything in life, well let's just say you're boxing yourself in for failure. Over the past 20 years we've seen the explosion of growth in devices, networking, and communication. It comes with no shock (to experts and experienced technical people alike) that it's taking a long time to switch over. There isn't some magical "press this button for everyone to be IPv6" FYI.

vbezhenar 8y ago

IPv4 worked for a long time and works for majority of installations now. It's hard to migrate when advantages are not obvious. I'm not sure that IPv6 is very good, I'm always having problem with it, but any standard will have adoption problems when current solution is good enough. I don't think that there's need for new standard and I'm sure that new standard won't be adopted better than IPv6.

kalleboo 8y ago
I'm not sure that IPv4 is very good, I'm always having issues with NAT etc
- vbezhenar 8y ago
  What issues? NAT always worked fine for me. It's one iptables rule and one sysctl.
  - Rjevski 8y ago
    NAT breaks the original design of the internet where every endpoint should be reachable from anywhere. Try using protocols that rely on this (like SIP) and see how far you go.

sp332 8y ago

I think you'd be surprised if you looked up actual adoption numbers. http://www.worldipv6launch.org/measurements/ http://www.worldipv6launch.org/measurements/

d3ckard 8y ago
Just did. 7 operators over 90%. On 7th page you get under 50% adoption if sorted. Well, given how old is IPv6, it's a failure.
- icebraining 8y ago
  Well, given how old is IPv6, it's a failure. Why?