Hi, after now almost 12 years using, working and teaching[1] IPv6 I've come to the conclusion that IPv6 is a mistake and will not work. Therefore the RIPE IPv6 WG should be disbanded and replaced with a new WG that MUST investigate all possible solutions to artificially prolong the live of IPv4 till the day a new successor for IPv4 is created and implemented! Some great ideas[2] are already proposed, some of them already implemented: - Use of NAT - Use of the first Class-A network 0.0.0.0[3] - Use of parts of localhost Class-A network 127.0.0.0 - Use of (parts) of Class-D address space (multicast) - Use of Class-E address space (future use) - Using part of the UDP / TCP port range as extension for the address. Some of the reserved address spaces could also be used. E.g. nobody is using 192.0.2.0/24 for documentation anyway. It should also be investigated to take back legacy IPv4 resources, although the "owners" of these resources might already selling them on the open market. It MUST also be considered not filtering on Class-C[4] bounderies but going for something smaller like /26 or /27 in the global routing table. Also new Class Designations for these prefixes MUST be created. The new successor to IPv4 should not make the same mistakes as IPv6. - IT MUST have NAT - It MUST have Classes - IT MUST have DHCP - It MUST have ARP - It should be possible to drop ICMP the same impact as in IPv4. Many experts I talked to over the years told me that blocking ICMP has no negative impacts. - It MUST only have numbers and dots "." - There should be absolutly no reasons to use "[ ]" in URLs Probably the best way to proceed is to just add one or two octets to the address. One of the reasons for the above is that there are so is so many good documentation already written about IPv4! And people already know about IPv4! Why waste this knowledge and experience? There is also plenty of good software out there that can't work with IPv6[5] Change is bad! People don't want to learn! IPv4! MUST! NOT! DIE! Jens [1] at least trying to teach, as one can see from the great number of people actually using IPv6 with little success [2] https://netdevconf.info/0x13/session.html?talk-ipv4-unicast-expansions [3] https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/commit/?i... [4] a Class-C network is the equivalent of an /24. I was told by experts that the definition of some bit set in the first octet of an IPv4 address is complete and utter nonsense [5] like a 20 year old shell script that is so important for $university that it would be hard for them to implement IPv6!
Hi Jens, Wow, first I had to look at today's date, I thought this was a April Fools joke mail. But to go forward seriously, a couple of questions to maybe clarify your thinking - from bullet points: 1. WHY should it have NAT 2. What do you understand under class, IPv4 "Classes" are just defined subnet groups (simply put) 3. AFAIK DHCPv6 is defined in RFC (3319,3646,4704,5007,6221,6355,6939,8415) 4. Partly agree on this one 5. Partly agree on this one, but probably with the right set of firewall rules you could achieve the same effect you are going after 6. Dots and colon, what's the difference? 7. Use DNS to resolve - no [] needed then. And for the "footprints": [4] you want classes in "IPvX" but negate the same with this point [5] what does the script have to do with network layer? Just my 2c. Uros On Thu, Oct 3, 2019 at 12:35 PM Jens Link <lists@quux.de> wrote:
Hi,
after now almost 12 years using, working and teaching[1] IPv6 I've come to the conclusion that IPv6 is a mistake and will not work.
Therefore the RIPE IPv6 WG should be disbanded and replaced with a new WG that MUST investigate all possible solutions to artificially prolong the live of IPv4 till the day a new successor for IPv4 is created and implemented!
Some great ideas[2] are already proposed, some of them already implemented:
- Use of NAT - Use of the first Class-A network 0.0.0.0[3] - Use of parts of localhost Class-A network 127.0.0.0 - Use of (parts) of Class-D address space (multicast) - Use of Class-E address space (future use) - Using part of the UDP / TCP port range as extension for the address.
Some of the reserved address spaces could also be used. E.g. nobody is using 192.0.2.0/24 for documentation anyway.
It should also be investigated to take back legacy IPv4 resources, although the "owners" of these resources might already selling them on the open market.
It MUST also be considered not filtering on Class-C[4] bounderies but going for something smaller like /26 or /27 in the global routing table. Also new Class Designations for these prefixes MUST be created.
The new successor to IPv4 should not make the same mistakes as IPv6.
- IT MUST have NAT - It MUST have Classes - IT MUST have DHCP - It MUST have ARP - It should be possible to drop ICMP the same impact as in IPv4. Many experts I talked to over the years told me that blocking ICMP has no negative impacts. - It MUST only have numbers and dots "." - There should be absolutly no reasons to use "[ ]" in URLs
Probably the best way to proceed is to just add one or two octets to the address.
One of the reasons for the above is that there are so is so many good documentation already written about IPv4! And people already know about IPv4! Why waste this knowledge and experience? There is also plenty of good software out there that can't work with IPv6[5] Change is bad! People don't want to learn!
IPv4! MUST! NOT! DIE!
Jens
[1] at least trying to teach, as one can see from the great number of people actually using IPv6 with little success
[2] https://netdevconf.info/0x13/session.html?talk-ipv4-unicast-expansions
[3] https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/commit/?i...
[4] a Class-C network is the equivalent of an /24. I was told by experts that the definition of some bit set in the first octet of an IPv4 address is complete and utter nonsense
[5] like a 20 year old shell script that is so important for $university that it would be hard for them to implement IPv6!
On 3 Oct 2019, at 12:58, Uros Gaber <uros@ub330.net> wrote:
Hi Jens,
Wow, first I had to look at today's date, I thought this was a April Fools joke mail.
Did you also look at the From?, because that’s not the one I expected if I instinctively expanded the name to that of someone I know, like the wg co-chair or so. Cheers Joao
But to go forward seriously, a couple of questions to maybe clarify your thinking - from bullet points: 1. WHY should it have NAT 2. What do you understand under class, IPv4 "Classes" are just defined subnet groups (simply put) 3. AFAIK DHCPv6 is defined in RFC (3319,3646,4704,5007,6221,6355,6939,8415) 4. Partly agree on this one 5. Partly agree on this one, but probably with the right set of firewall rules you could achieve the same effect you are going after 6. Dots and colon, what's the difference? 7. Use DNS to resolve - no [] needed then.
And for the "footprints": [4] you want classes in "IPvX" but negate the same with this point [5] what does the script have to do with network layer?
Just my 2c.
Uros
On Thu, Oct 3, 2019 at 12:35 PM Jens Link <lists@quux.de <mailto:lists@quux.de>> wrote: Hi,
after now almost 12 years using, working and teaching[1] IPv6 I've come to the conclusion that IPv6 is a mistake and will not work.
Therefore the RIPE IPv6 WG should be disbanded and replaced with a new WG that MUST investigate all possible solutions to artificially prolong the live of IPv4 till the day a new successor for IPv4 is created and implemented!
Some great ideas[2] are already proposed, some of them already implemented:
- Use of NAT - Use of the first Class-A network 0.0.0.0[3] - Use of parts of localhost Class-A network 127.0.0.0 - Use of (parts) of Class-D address space (multicast) - Use of Class-E address space (future use) - Using part of the UDP / TCP port range as extension for the address.
Some of the reserved address spaces could also be used. E.g. nobody is using 192.0.2.0/24 <http://192.0.2.0/24> for documentation anyway.
It should also be investigated to take back legacy IPv4 resources, although the "owners" of these resources might already selling them on the open market.
It MUST also be considered not filtering on Class-C[4] bounderies but going for something smaller like /26 or /27 in the global routing table. Also new Class Designations for these prefixes MUST be created.
The new successor to IPv4 should not make the same mistakes as IPv6.
- IT MUST have NAT - It MUST have Classes - IT MUST have DHCP - It MUST have ARP - It should be possible to drop ICMP the same impact as in IPv4. Many experts I talked to over the years told me that blocking ICMP has no negative impacts. - It MUST only have numbers and dots "." - There should be absolutly no reasons to use "[ ]" in URLs
Probably the best way to proceed is to just add one or two octets to the address.
One of the reasons for the above is that there are so is so many good documentation already written about IPv4! And people already know about IPv4! Why waste this knowledge and experience? There is also plenty of good software out there that can't work with IPv6[5] Change is bad! People don't want to learn!
IPv4! MUST! NOT! DIE!
Jens
[1] at least trying to teach, as one can see from the great number of people actually using IPv6 with little success
[2] https://netdevconf.info/0x13/session.html?talk-ipv4-unicast-expansions <https://netdevconf.info/0x13/session.html?talk-ipv4-unicast-expansions>
[3] https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/commit/?i... <https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/commit/?id=96125bf9985a>
[4] a Class-C network is the equivalent of an /24. I was told by experts that the definition of some bit set in the first octet of an IPv4 address is complete and utter nonsense
[5] like a 20 year old shell script that is so important for $university that it would be hard for them to implement IPv6!
On 10/3/19 1:11 PM, Joao Luis Silva Damas wrote:
Did you also look at the From?, because that’s not the one I expected if I instinctively expanded the name to that of someone I know, like the wg co-chair or so.
of course Jens Link is not Jen Linkova :-) -- antonio
On 3. Oct 2019, at 13:16, Antonio Prado via ipv6-wg <ipv6-wg@ripe.net> wrote:
On 10/3/19 1:11 PM, Joao Luis Silva Damas wrote:
Did you also look at the From?, because that’s not the one I expected if I instinctively expanded the name to that of someone I know, like the wg co-chair or so.
of course Jens Link is not Jen Linkova :-)
-- antonio
No worries, looking at the headers and seeing that nearly all received-by mail hops in the original mail use IPv6 addresses -- and considering that IPv6 doesn't work anyway -- the whole email obviously must have been a complete and utter illusion … ;-) -Andi
Andreas Härpfer <ah@v6x.org> writes:
On 3. Oct 2019, at 13:16, Antonio Prado via ipv6-wg <ipv6-wg@ripe.net> wrote:
No worries, looking at the headers and seeing that nearly all received-by mail hops in the original mail use IPv6 addresses -- and considering that IPv6 doesn't work anyway -- the whole email obviously must have been a complete and utter illusion …
Thanks for noticing. A new mail server setup is on my todo list, running a dual stack system is too much work to ipv6 will go away. Jens -- ---------------------------------------------------------------------------- | Delbrueckstr. 41 | 12051 Berlin, Germany | +49-151-18721264 | | http://blog.quux.de | jabber: jenslink@quux.de | --------------- | ----------------------------------------------------------------------------
On Thu, Oct 3, 2019 at 13:57 Jens Link <lists@quux.de> wrote:
Andreas Härpfer <ah@v6x.org> writes:
On 3. Oct 2019, at 13:16, Antonio Prado via ipv6-wg <ipv6-wg@ripe.net> wrote:
No worries, looking at the headers and seeing that nearly all received-by mail hops in the original mail use IPv6 addresses -- and considering that IPv6 doesn't work anyway -- the whole email obviously must have been a complete and utter illusion …
Thanks for noticing. A new mail server setup is on my todo list, running a dual stack system is too much work to ipv6 will go away.
Even worse, delivering email over ipv6 to the mail giants is a far worse experience than via ipv4. More emails arrive when you disable ipv6 on your mail servers. Kind regards, Job
On Thu, Oct 3, 2019 at 9:12 PM Joao Luis Silva Damas <joao@bondis.org> wrote:
Did you also look at the From?, because that’s not the one I expected if I instinctively expanded the name to that of someone I know, like the wg co-chair or so.
Yeah....Jens, maybe before we start discussing the renaming the group we shall first decide who of us two needs to change the name? ;) People are getting confused... -- SY, Jen Linkova aka Furry
Jen Linkova wrote : As a co-chair I'm excited to see some discussion happening here, especially after the list has been quiet for a while.
As a matter of fact, it was so quiet that I forgot that I once subscribed to it. Don't worry; I am about to unsubscribe.
I'm less excited to see that some people have started giving up hope and telling the rest of use we shall give up to..
You are fighting for your survival. Although I agree with the most recent analysis that IPv6 will not become an orphan, you have to understand that the IPv4 ecosystem will survive no matter what you do, and that a war will not be in your favor. I do not represent the IPv4 zealots, but I have become one of them by economic necessity. I'm not trying to run your WG. All that I am saying is that the IPv6 zealots have been a total pain in my backside, and that the FUD they have been spreading for the last 20 years has come out of style. IPv4 will survive forever. I can not say the same of IPv6. You have to change your tune if you want to survive.
Convince everyone that IPv6 is good and IPv4 shall be turned off?
This is what you have to stop. You work for Google. Do you want to bet your career that Google is going to switch to IPv6 only ? Think carefully. Think about your career. I personally know the big shots at big vendor who bet their careers on IPv6 and have become irrelevant. You are too young to be a politician who turns their coat in the wind. I say again : are you willing to bet your career on IPv6 ? Michel.
Disclaimer: this is a personal opinion, all hats off. On Sat, Oct 5, 2019 at 12:52 PM Michel Py <michel@arneill-py.sacramento.ca.us> wrote:
I'm less excited to see that some people have started giving up hope and telling the rest of use we shall give up to..
You are fighting for your survival.
I'm not sure it's about survival at all. I could definitely survive w/o IPv6 (or without any form of IP) but it's much more fun with it. I'm not even sure it's a fight even. Fight sounds too romantic or too dramatic.
Although I agree with the most recent analysis that IPv6 will not become an orphan, you have to understand that the IPv4 ecosystem will survive no matter what you do,
Well, I've seen DECNET and Appletalks "ecosystems" in the wild long after most of the world forgot those words. Complete IPv4 extinction has never been my goal, I do not really care if some isolated networks are going to run it.
and that a war will not be in your favor.
I do not think there is a war. People do what they need to do and sometimes they talk about what they have experienced, just in case others might learn smth new.
I'm not trying to run your WG. All that I am saying is that the IPv6 zealots have been a total pain in my backside, and that the FUD they have been spreading for the last 20 years has come out of style.
IPv4 will survive forever.
There will be a long tail indeed. Then the economics would make dual-stack suboptimal, ex. for the very special cases almost nobody cares about. What you might want to keep in mind is that different networks are in different situation re: IPv6. Some of them badly need it. Some of them got rid of Ipv4. Some of them will need it sooner or later. Some could probably survive forever using a single Ipv4 address.
I can not say the same of IPv6. You have to change your tune if you want to survive.
I appreciate people care about our survival. I even appreciate free advices ;)
Convince everyone that IPv6 is good and IPv4 shall be turned off?
This is what you have to stop.
First of all please note it was listed as a question. An option. Not necessary the best one.
You work for Google.
In case you have not noticed there is no reference to my employer anywhere, so I'd suggest we do not talk about whom I work for. It does not really matter.
Do you want to bet your career that Google is going to switch to IPv6 only ?
I'm afraid to answer your question I need to go through some legal approval process.
Think carefully. Think about your career. I personally know the big shots at big vendor who bet their careers on IPv6 and have become irrelevant.
As I've said before, I'm deeply touched that someone in the Internet cares about my career ;)
You are too young to be a politician who turns their coat in the wind.
Lol, I'll take this a compliment, thank you sir!
I say again : are you willing to bet your career on IPv6 ?
I have done it. No regrets so far. -- SY, Jen Linkova aka Furry
Uros Gaber <uros@ub330.net> writes:
1. WHY should it have NAT
NATs are good. They provide security.
2. What do you understand under class, IPv4 "Classes" are just defined subnet groups (simply put)
Things need names. Numbers are hard to remember. We have Class-A for /8, Class-B for /16 and Class-C for /24. We need names for the others as well.
3. AFAIK DHCPv6 is defined in RFC (3319,3646,4704,5007,6221,6355,6939,8415)
But it's DHCPv6. Not DHCP! It works differently. And Android does not support it. Enterprise Customers want DHCP!
6. Dots and colon, what's the difference?
I have do change my regex.
7. Use DNS to resolve - no [] needed then.
DNS is to hard, to complex and fails to often. And in enterprise networks it probably done by another team.
[5] what does the script have to do with network layer?
The script was just an example for software breaking when you implement something that looks completely different like IPv4. I this case the script is parsing log files and netflow data and we are back to the regex. Jens -- ---------------------------------------------------------------------------- | Delbrueckstr. 41 | 12051 Berlin, Germany | +49-151-18721264 | | http://blog.quux.de | jabber: jenslink@quux.de | --------------- | ----------------------------------------------------------------------------
Hi, On Thu, 3 Oct 2019, Jens Link wrote:
Uros Gaber <uros@ub330.net> writes:
1. WHY should it have NAT
NATs are good. They provide security.
No, they provide "Translation". Not the same thing.
2. What do you understand under class, IPv4 "Classes" are just defined subnet groups (simply put)
Things need names. Numbers are hard to remember. We have Class-A for /8, Class-B for /16 and Class-C for /24. We need names for the others as well.
CIDR. People that thought that terminology to students over the years really fumbled...
3. AFAIK DHCPv6 is defined in RFC (3319,3646,4704,5007,6221,6355,6939,8415)
But it's DHCPv6. Not DHCP! It works differently. And Android does not support it. Enterprise Customers want DHCP!
Is it unfixable...?
6. Dots and colon, what's the difference?
I have do change my regex.
The world is all about changes :-)
7. Use DNS to resolve - no [] needed then.
DNS is to hard, to complex and fails to often.
No, it's really the most robust planetary system. It can suffer attacks (it did, it does) but is still pretty much does the job.
And in enterprise networks it probably done by another team.
So? Teams inside the same organisation are supposed to speak :-) Cheers, Carlos
Hi, On Thu, Oct 03, 2019 at 01:03:07PM +0100, Carlos Friaças via ipv6-wg wrote:
On Thu, 3 Oct 2019, Jens Link wrote:
Uros Gaber <uros@ub330.net> writes:
1. WHY should it have NAT
NATs are good. They provide security. No, they provide "Translation". Not the same thing.
You need to listen to the experts here https://www.youtube.com/watch?v=v26BA Gert Doering -- NetMaster -- have you enabled IPv6 on something today...? SpaceNet AG Vorstand: Sebastian v. Bomhard, Michael Emmer Joseph-Dollinger-Bogen 14 Aufsichtsratsvors.: A. Grundner-Culemann D-80807 Muenchen HRB: 136055 (AG Muenchen) Tel: +49 (0)89/32356-444 USt-IdNr.: DE813185279
Hi, On Thu, Oct 03, 2019 at 02:17:24PM +0200, Gert Doering wrote:
You need to listen to the experts here
I could use an expert that explains to me this click-and-paste stuff with modern browsers... anyway, the correct video is https://www.youtube.com/watch?v=v26BAlfWBm8 Gert Doering -- NetMaster -- have you enabled IPv6 on something today...? SpaceNet AG Vorstand: Sebastian v. Bomhard, Michael Emmer Joseph-Dollinger-Bogen 14 Aufsichtsratsvors.: A. Grundner-Culemann D-80807 Muenchen HRB: 136055 (AG Muenchen) Tel: +49 (0)89/32356-444 USt-IdNr.: DE813185279
Am Thu, 03 Oct 2019 schrieb Gert Doering:
I could use an expert that explains to me this click-and-paste stuff with modern browsers...
Uuuuh, Ooooh. Let's roll out 100% IPv6 globally first. That's low hanging fruit, compared to that click-and-paste stuff ;-) bbu
Carlos Friaças via ipv6-wg <ipv6-wg@ripe.net> writes:
CIDR. People that thought that terminology to students over the years really fumbled...
years ago? It's till taught. Did a entry level Juniper Certification some time ago and they asked which Class 10.10.10.10 belongs to. Sure CIDR turned 26 last month but hey.
But it's DHCPv6. Not DHCP! It works differently. And Android does not support it. Enterprise Customers want DHCP!
Is it unfixable...?
The Android side? No. Use google. You should find a discussion with a couple thousand posting on this topic. On the Enterprise side? I wouldn't bet on it.
6. Dots and colon, what's the difference?
I have do change my regex.
The world is all about changes :-)
"I've been working in the IT buissnes for 25 years and nothing has changed!" - I heared this more then once.
No, it's really the most robust planetary system. It can suffer attacks (it did, it does) but is still pretty much does the job.
People remember wired problems with DNS and people don't understand DNS. "I need to put the IP into my DNS Server" - The guy meant %systemroot%\sytem32\drivers\etc\hosts
And in enterprise networks it probably done by another team.
So? Teams inside the same organisation are supposed to speak :-)
Supposed is the right word. In theory yes, in reality most often not. Jens -- ---------------------------------------------------------------------------- | Delbrueckstr. 41 | 12051 Berlin, Germany | +49-151-18721264 | | http://blog.quux.de | jabber: jenslink@quux.de | --------------- | ----------------------------------------------------------------------------
Hello, Classes "A" "B" "C" mean not only /8 /16 /24 but '0' '0' '0', '1' '0' '0', '1' '1' '0' in first 3 bits... :-) __________________________________ Валерий Солдатов, ЗАО Бэст-Телеком ----- Исходное сообщение ----- От: "Jens Link" <lists@quux.de> Кому: ipv6-wg@ripe.net Отправленные: Четверг, 3 Октябрь 2019 г 14:55:32 Тема: Re: [ipv6-wg] Disband IPv6 WG ...
2. What do you understand under class, IPv4 "Classes" are just defined subnet groups (simply put)
Things need names. Numbers are hard to remember. We have Class-A for /8, Class-B for /16 and Class-C for /24. We need names for the others as well.
Hello Jen Link, Il 03/10/2019 13:55, Jens Link ha scritto:
Uros Gaber <uros@ub330.net> writes:
1. WHY should it have NAT NATs are good. They provide security. Are you sure you are about networking?
2. What do you understand under class, IPv4 "Classes" are just defined subnet groups (simply put) Things need names. Numbers are hard to remember. We have Class-A for /8, Class-B for /16 and Class-C for /24. We need names for the others as well.
3. AFAIK DHCPv6 is defined in RFC (3319,3646,4704,5007,6221,6355,6939,8415) But it's DHCPv6. Not DHCP! It works differently. And Android does not support it. Enterprise Customers want DHCP!
6. Dots and colon, what's the difference? I have do change my regex.
7. Use DNS to resolve - no [] needed then. DNS is to hard, to complex and fails to often. And in enterprise networks it probably done by another team.
[5] what does the script have to do with network layer? The script was just an example for software breaking when you implement something that looks completely different like IPv4. I this case the script is parsing log files and netflow data and we are back to the regex.
Jens
__ Riccardo
Riccardo Gori <rgori@wirem.net> writes:
1. WHY should it have NAT NATs are good. They provide security.
Are you sure you are about networking?
People believer this so it must be true.I wont question the believes and wisdom of thousands of networking experts. Jens -- ---------------------------------------------------------------------------- | Delbrueckstr. 41 | 12051 Berlin, Germany | +49-151-18721264 | | http://blog.quux.de | jabber: jenslink@quux.de | --------------- | ----------------------------------------------------------------------------
Hi Jens, On 03/10/2019 12:34, Jens Link wrote:
after now almost 12 years using, working and teaching[1] IPv6 I've come to the conclusion that IPv6 is a mistake and will not work.
Please elaborate. We are well underway with IPv6. In al honesty, your suggestions hurt my eyes. -- Marco
Hi, (Respectfully) IPv6 _is_ working. Check your DNS, please. You should see AAAA records. Google is using it. Facebook is using it. Cloudflare is using it. We are using it. And so on... Cheers, Carlos On Thu, 3 Oct 2019, Jens Link wrote:
Hi,
after now almost 12 years using, working and teaching[1] IPv6 I've come to the conclusion that IPv6 is a mistake and will not work.
Therefore the RIPE IPv6 WG should be disbanded and replaced with a new WG that MUST investigate all possible solutions to artificially prolong the live of IPv4 till the day a new successor for IPv4 is created and implemented!
Some great ideas[2] are already proposed, some of them already implemented:
- Use of NAT - Use of the first Class-A network 0.0.0.0[3] - Use of parts of localhost Class-A network 127.0.0.0 - Use of (parts) of Class-D address space (multicast) - Use of Class-E address space (future use) - Using part of the UDP / TCP port range as extension for the address.
Some of the reserved address spaces could also be used. E.g. nobody is using 192.0.2.0/24 for documentation anyway.
It should also be investigated to take back legacy IPv4 resources, although the "owners" of these resources might already selling them on the open market.
It MUST also be considered not filtering on Class-C[4] bounderies but going for something smaller like /26 or /27 in the global routing table. Also new Class Designations for these prefixes MUST be created.
The new successor to IPv4 should not make the same mistakes as IPv6.
- IT MUST have NAT - It MUST have Classes - IT MUST have DHCP - It MUST have ARP - It should be possible to drop ICMP the same impact as in IPv4. Many experts I talked to over the years told me that blocking ICMP has no negative impacts. - It MUST only have numbers and dots "." - There should be absolutly no reasons to use "[ ]" in URLs
Probably the best way to proceed is to just add one or two octets to the address.
One of the reasons for the above is that there are so is so many good documentation already written about IPv4! And people already know about IPv4! Why waste this knowledge and experience? There is also plenty of good software out there that can't work with IPv6[5] Change is bad! People don't want to learn!
IPv4! MUST! NOT! DIE!
Jens
[1] at least trying to teach, as one can see from the great number of people actually using IPv6 with little success
[2] https://netdevconf.info/0x13/session.html?talk-ipv4-unicast-expansions
[3] https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/commit/?i...
[4] a Class-C network is the equivalent of an /24. I was told by experts that the definition of some bit set in the first octet of an IPv4 address is complete and utter nonsense
[5] like a 20 year old shell script that is so important for $university that it would be hard for them to implement IPv6!
Carlos Friaças via ipv6-wg <ipv6-wg@ripe.net> writes:
Hi, (Respectfully)
IPv6 _is_ working.
github, stackexchange, twitter, amazon.com.... As a small test: Try to see if you can get all the packages needed for Linux from Scratch on an IPv6 only host (with out any form of NAT / proxy). There was as presentation at RIPE about this some time ago and I guess not much has changed.
Cloudflare is using it.
Ah I forgot that in my original mail: They break IPv6. Had to disable my IPv6 VPN tunnel in order to access sites hosted by CF. I hate clicking that I'm not a robot and pretty sure that installing an anti virus software will not help. That was when I was still actively using IPv6. Jens -- ---------------------------------------------------------------------------- | Delbrueckstr. 41 | 12051 Berlin, Germany | +49-151-18721264 | | http://blog.quux.de | jabber: jenslink@quux.de | --------------- | ----------------------------------------------------------------------------
On Thu, 3 Oct 2019, Jens Link wrote:
Carlos Friaças via ipv6-wg <ipv6-wg@ripe.net> writes:
Hi, (Respectfully)
IPv6 _is_ working.
github, stackexchange, twitter, amazon.com....
As a small test: Try to see if you can get all the packages needed for Linux from Scratch on an IPv6 only host (with out any form of NAT / proxy). There was as presentation at RIPE about this some time ago and I guess not much has changed.
If that doesn't work _now_, it will be fixed at some point. "IPv6-only" is different than "IPv6". If you go through the threads, the claim that some IPv4 will be needed is repeated many times. The real issue is not that IPv6 "doesn't work", the real issue is some people insist in rejecting it (i.e. those mentioned above, just to name a few).
Cloudflare is using it.
Ah I forgot that in my original mail: They break IPv6. Had to disable my IPv6 VPN tunnel in order to access sites hosted by CF. I hate clicking that I'm not a robot and pretty sure that installing an anti virus software will not help. That was when I was still actively using IPv6.
I don't work for Cloudflare, so i hope some from Cloudflare can comment :-) Cheers, Carlos
Jens -- ---------------------------------------------------------------------------- | Delbrueckstr. 41 | 12051 Berlin, Germany | +49-151-18721264 | | http://blog.quux.de | jabber: jenslink@quux.de | --------------- | ----------------------------------------------------------------------------
On 3 Oct 2019, at 13:30, Carlos Friaças via ipv6-wg <ipv6-wg@ripe.net<mailto:ipv6-wg@ripe.net>> wrote: The real issue is not that IPv6 "doesn't work", the real issue is some people insist in rejecting it (i.e. those mentioned above, just to name a few). https://www.youtube.com/watch?v=v26BAlfWBm8 (Surprised we’re having this conversation in 2019, as the final fumes of IPv4 address space disappear from Europe…) Tim
Tim Chown <Tim.Chown@jisc.ac.uk> writes:
(Surprised we’re having this conversation in 2019, as the final fumes of IPv4 address space disappear from Europe…)
If you had told me 10 or even 5 years ago that I would be having the conversation in 2019 I would have laughed at you. Now it's a very sad situation. IPv4 has won. I had a discussion over lunch about v6 yesterday (which is part of the reason I started this today) and all I heard "but that is different then IPv4. I don't like this!" Jens -- ---------------------------------------------------------------------------- | Delbrueckstr. 41 | 12051 Berlin, Germany | +49-151-18721264 | | http://blog.quux.de | jabber: jenslink@quux.de | --------------- | ----------------------------------------------------------------------------
On 3 Oct 2019, at 16:02, Jens Link <lists@quux.de> wrote:
Tim Chown <Tim.Chown@jisc.ac.uk> writes:
(Surprised we’re having this conversation in 2019, as the final fumes of IPv4 address space disappear from Europe…)
If you had told me 10 or even 5 years ago that I would be having the conversation in 2019 I would have laughed at you. Now it's a very sad situation. IPv4 has won.
I had a discussion over lunch about v6 yesterday (which is part of the reason I started this today) and all I heard "but that is different then IPv4. I don't like this!"
There will always be a legacy tail. The dinosaurs can wallow in their swamp. Those who deploy v6 will benefit from it. Others will feel the heat of not moving; here in the UK it’s Sky and BT who have between them ~10M households on IPv6. That’s not failure. New communities will benefit. For example, the largest science experiments are now migrating to IPv6, e.g., CERN and WLCG is 70% there, SKA will use it. Tim
Am 03.10.19 um 17:11 schrieb Tim Chown:
On 3 Oct 2019, at 16:02, Jens Link <lists@quux.de> wrote:
Tim Chown <Tim.Chown@jisc.ac.uk> writes:
(Surprised we’re having this conversation in 2019, as the final fumes of IPv4 address space disappear from Europe…) If you had told me 10 or even 5 years ago that I would be having the conversation in 2019 I would have laughed at you. Now it's a very sad situation. IPv4 has won.
Well, the source for "new" IPv4 addresses is finally drying out in the RIPE region, so I do not agree with "IPv4 has won"; it lived an amazing life so far and is, since several years, transitioning into it's evening of life. I wouldn't bet on a date when IPv4 in the public Internet will be shut down, though. Not even a decade, to be honest ...
I had a discussion over lunch about v6 yesterday (which is part of the reason I started this today) and all I heard "but that is different then IPv4. I don't like this!"
There will always be a legacy tail. The dinosaurs can wallow in their swamp.
Some of those dinosauers are still in their diapers, though.
Those who deploy v6 will benefit from it. Others will feel the heat of not moving; here in the UK it’s Sky and BT who have between them ~10M households on IPv6. That’s not failure.
No, it's a start; over here in Germany, most mobile operators give you RFC6890 or RFC1918 addresses, still. Cable operators hand out DS (-Lite, mostly) for consumers, (semi-) fixed IPv4 (no DS) for commercial clients. FritzVPN, the VPN solution of popular CPE maker AVM, still fails completely with IPv6, both as transport and as payload. All in all, it's more failure than success (and even progress is fscking slow; Vodafone is allegedly starting somethings like DS-lite on mobile these days, o2 on mobiles uses public v6 a long time already — for VoLTE, but not data). But then it's Germany, where anything IP is Neuland anyway.
New communities will benefit. For example, the largest science experiments are now migrating to IPv6, e.g., CERN and WLCG is 70% there, SKA will use it.
But will they go the whole way, i. e. make their stuff accessible from the outside, including informational webservers and other infrastructure (DNS, MX), v6-only? Until much used resources go v6-only, there's no chance in hell that "[o]thers will feel the heat of not moving", as everyone still makes everything available via v4. So, why not make ripe.net v6-only by 2020-01-01, as RIPE NCC's IPv4 pool will have run dry by then anyway? Am 03.10.19 um 13:11 schrieb Joao Luis Silva Damas:
On 3 Oct 2019, at 12:58, Uros Gaber <uros@ub330.net <mailto:uros@ub330.net>> wrote:
Hi Jens,
Wow, first I had to look at today's date, I thought this was a April Fools joke mail.
Did you also look at the From?, because that’s not the one I expected if I instinctively expanded the name to that of someone I know, like the wg co-chair or so.
Well, off-topic, but more noteworthy, the RIPE NCC mailservers are, as of today, still running Exim 4.92.2, remotely exploitable according to CVE-2019-16928. Am 03.10.19 um 12:34 schrieb Jens Link:
Hi,
after now almost 12 years using, working and teaching[1] IPv6 I've come to the conclusion that IPv6 is a mistake and will not work. According to the mailing list archive, "[t]he IPv6 Working Group is for anyone with an interest in the next generation Internet Protocol. The activities of the WG include education and outreach, sharing deployment experiences and discussing and fixing operational issues". So, Jens shared his IPv6 deployment experiences ("isn't happening"), maybe there's something the IPv6 WG can do to enforce IPv6 deployment? BTW, at least in terms of availability v6 is the current, v4 the legacy Internet Protocol, maybe that wording should be updated?
Regards, -kai
Am 04.10.2019 um 00:22 schrieb Kai 'wusel' Siering <wusel+ml@uu.org<mailto:wusel+ml@uu.org>>: So, why not make ripe.net<http://ripe.net> v6-only by 2020-01-01, as RIPE NCC's IPv4 pool will have run dry by then anyway? I like this idea. This message and any attachment are confidential and may be privileged or otherwise protected from disclosure. If you are not the intended recipient, you must not copy this message or attachment or disclose the contents to any other person. If you have received this transmission in error, please notify the sender immediately and delete the message and any attachment from your system. Merck KGaA, Darmstadt, Germany and any of its subsidiaries do not accept liability for any omissions or errors in this message which may arise as a result of E-Mail-transmission or for damages resulting from any unauthorized changes of the content of this message and any attachment thereto. Merck KGaA, Darmstadt, Germany and any of its subsidiaries do not guarantee that this message is free of viruses and does not accept liability for any damages caused by any virus transmitted therewith. Click http://www.merckgroup.com/disclaimer to access the German, French, Spanish and Portuguese versions of this disclaimer.
On Fri, 4 Oct 2019, Alexander Koeppe wrote:
Am 04.10.2019 um 00:22 schrieb Kai 'wusel' Siering <wusel+ml@uu.org>:
So, why not make ripe.net v6-only by 2020-01-01, as RIPE NCC's IPv4 pool will have run dry by then anyway?
The fact the pool runs dry doesn't mean IPv4 packets will stop flowing. It only means growth of public address usage through the exclusive usage of IPv4 becomes more difficult and possibly more expensive. The "scarcity age" of IPv4 in the RIPE service region started back in September 2012. More than 7 years ago. Everyone should have deployed IPv6 by now, but that didn't really happen. Carlos
I like this idea.
This message and any attachment are confidential and may be privileged or otherwise protected from disclosure. If you are not the intended recipient, you must not copy this message or attachment or disclose the contents to any other person. If you have received this transmission in error, please notify the sender immediately and delete the message and any attachment from your system. Merck KGaA, Darmstadt, Germany and any of its subsidiaries do not accept liability for any omissions or errors in this message which may arise as a result of E-Mail-transmission or for damages resulting from any unauthorized changes of the content of this message and any attachment thereto. Merck KGaA, Darmstadt, Germany and any of its subsidiaries do not guarantee that this message is free of viruses and does not accept liability for any damages caused by any virus transmitted therewith.
Click http://www.merckgroup.com/disclaimer to access the German, French, Spanish and Portuguese versions of this disclaimer.
According to the mailing list archive, "[t]he IPv6 Working Group is for anyone with an interest in the next generation Internet Protocol. The activities of the WG include education and outreach, sharing deployment experiences and discussing and fixing operational issues". So, Jens shared his IPv6 deployment experiences ("isn't happening"), maybe there's something the IPv6 WG can do to enforce IPv6 deployment? BTW, at least in terms of availability v6 is the current, v4 the legacy Internet Protocol, maybe that wording should be updated?
My experience: I was consulting with a small ISP on the opportunity to deploy dual stack IPv6 to the end-users. IPv6 is already functional with the peers. RIPE published its recommendations: https://www.ripe.net/publications/docs/ripe-690. Basically, delegate a /48 for each customer and give the ONT an IP from a different pool. Which comes with some routing questions: how do you map the /48 with the CPE? Is there some kind of OSPF-like protocol that allows the ONT to advertise the delegated class? Do you add all possible routes in the router at the beginning, but then end up with 100k routes in an area with 5k customers? Do you programmatically add the routes when activating a client? Is there a standard for that? So, at this year's MENOG, I asked the person next to the IPv6 banner whom I could talk to. She said she will forward all my questions to their IPv6 expert. Sure enough, I got a response from the IPv6 expert: a few very polite paragraphs that boiled down to: "Huh?!". And a link to this list, which for the past 8 months was comatose. The result: I asked my customer if he has enough IPv4 addresses for the next 3 years. He said yes, so my recommendation was: wait for a couple of years. Takeaway: If we want more IPv6 at the end-user level, we need standards to do that, not just some recommendations. Best regards, Dan Craciun
Hi, On Sun, Oct 06, 2019 at 10:49:02AM +0300, dan@barletford.com wrote:
Which comes with some routing questions: how do you map the /48 with the CPE? Is there some kind of OSPF-like protocol that allows the ONT to advertise the delegated class? Do you add all possible routes in the router at the beginning, but then end up with 100k routes in an area with 5k customers? Do you programmatically add the routes when activating a client? Is there a standard for that?
This is no different from IPv4, effectively. In IPv4, a client is handed "a single IPv4 address", by whatever means (RADIUS backend, DHCP backend, local pool on the PE router, static routes...) and this IPv4 address is then either injected into OSPF/iBGP or aggregated into "network, please send me only the /23 supernet". Same for IPv6 - depending on your gear, you either do the /48 assignment on the PE router (BRAS) by means of "static" or "DHCPv6" or "pool", or you involve a backend server (DHCPv6 relay, RADIUS, ...) to tell the PE what to send where. Then, either aggregate on the PE ("send this /40 to me, no specific routes") or redistribute to iBGP - which depends on network structure, aggregation boundaries, number of customers (if you have 5k customers, just send to iBGP, and be done with it, if you have 5m customers, you will want multiple layers of internal aggregation). Since this depends on what your ISP gear can do *and* how the ISP is generally set up (business customers with static assignments, or only dynamic assignments for residents, ...) and how their IPv4 provisioning tools operate, the "correct" answer depends. So it's not trivial to tell "this is the standard" - there are multiple, and they all have their upsides and downsides. As Lee said "just labour". [..]
The result: I asked my customer if he has enough IPv4 addresses for the next 3 years. He said yes, so my recommendation was: wait for a couple of years.
I think you are a bit lazy. Nobody else will be able to tell you what the right answer is for your customer network, because nobody *knows* your customer network. We can tell you what options and protocols exist, but only you and your ISP customer know the equipment, the equipment's capabilities, and how the existing network is operating. If you have specific questions ("I have a PE router from vendor X, and when I try to redistribute DHCPv6-relay into RIPng, the ethernet plugs all fall out"), there are people here or on the more vendor-specific lists (https://puck.nether.net/mailman/listinfo/) which will be happy to help. Gert Doering -- NetMaster -- have you enabled IPv6 on something today...? SpaceNet AG Vorstand: Sebastian v. Bomhard, Michael Emmer Joseph-Dollinger-Bogen 14 Aufsichtsratsvors.: A. Grundner-Culemann D-80807 Muenchen HRB: 136055 (AG Muenchen) Tel: +49 (0)89/32356-444 USt-IdNr.: DE813185279
I think he's on to something. Perhaps the IETF can start working on IPv4bis. Something with, say, 128 bit addresses, fixed-size subnets and heavier use of multicast instead of broadcast. We'll leave NAT4bis4bis as an implementation detail for router vendors. Alex
On Oct 3, 2019, at 12:34 , Jens Link <lists@quux.de> wrote:
Hi,
after now almost 12 years using, working and teaching[1] IPv6 I've come to the conclusion that IPv6 is a mistake and will not work.
Therefore the RIPE IPv6 WG should be disbanded and replaced with a new WG that MUST investigate all possible solutions to artificially prolong the live of IPv4 till the day a new successor for IPv4 is created and implemented!
Some great ideas[2] are already proposed, some of them already implemented:
- Use of NAT - Use of the first Class-A network 0.0.0.0[3] - Use of parts of localhost Class-A network 127.0.0.0 - Use of (parts) of Class-D address space (multicast) - Use of Class-E address space (future use) - Using part of the UDP / TCP port range as extension for the address.
Some of the reserved address spaces could also be used. E.g. nobody is using 192.0.2.0/24 for documentation anyway.
It should also be investigated to take back legacy IPv4 resources, although the "owners" of these resources might already selling them on the open market.
It MUST also be considered not filtering on Class-C[4] bounderies but going for something smaller like /26 or /27 in the global routing table. Also new Class Designations for these prefixes MUST be created.
The new successor to IPv4 should not make the same mistakes as IPv6.
- IT MUST have NAT - It MUST have Classes - IT MUST have DHCP - It MUST have ARP - It should be possible to drop ICMP the same impact as in IPv4. Many experts I talked to over the years told me that blocking ICMP has no negative impacts. - It MUST only have numbers and dots "." - There should be absolutly no reasons to use "[ ]" in URLs
Probably the best way to proceed is to just add one or two octets to the address.
One of the reasons for the above is that there are so is so many good documentation already written about IPv4! And people already know about IPv4! Why waste this knowledge and experience? There is also plenty of good software out there that can't work with IPv6[5] Change is bad! People don't want to learn!
IPv4! MUST! NOT! DIE!
Jens
[1] at least trying to teach, as one can see from the great number of people actually using IPv6 with little success
[2] https://netdevconf.info/0x13/session.html?talk-ipv4-unicast-expansions
[3] https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/commit/?i...
[4] a Class-C network is the equivalent of an /24. I was told by experts that the definition of some bit set in the first octet of an IPv4 address is complete and utter nonsense
[5] like a 20 year old shell script that is so important for $university that it would be hard for them to implement IPv6!
Hi, On Thu, Oct 03, 2019 at 01:53:47PM +0200, Alex Le Heux wrote:
Perhaps the IETF can start working on IPv4bis. Something with, say, 128 bit addresses, fixed-size subnets and heavier use of multicast instead of broadcast.
Surely global multicast could replace routing? That would be great for all those poor BGP-table-entry challenged routers! Gert Doering -- NetMaster -- have you enabled IPv6 on something today...? SpaceNet AG Vorstand: Sebastian v. Bomhard, Michael Emmer Joseph-Dollinger-Bogen 14 Aufsichtsratsvors.: A. Grundner-Culemann D-80807 Muenchen HRB: 136055 (AG Muenchen) Tel: +49 (0)89/32356-444 USt-IdNr.: DE813185279
On Thu, Oct 3, 2019, at 12:34, Jens Link wrote:
after now almost 12 years using, working and teaching[1] IPv6 I've come to the conclusion that IPv6 is a mistake and will not work. ......... IPv4! MUST! NOT! DIE!
Did "someone" had a look at the wrong page in the calendar (we're Oct 03, not Apr 01), or did time travel just happen for the first time ( for a duration of N+0.5 years, round trip) ? Anyway, waiting to see the related presentation at RIPE79 *IPv6* working group (unless the agenda has been shuffled these last days). -- Radu-Adrian FEURDEAN
Beside the part of sending SMTP email over IPv6... Jens, what did you drink this lunch time? It seems to be good :-) I take your email as a good joke: see the amount of replies ! -éric On 03/10/2019, 12:35, "ipv6-wg on behalf of Jens Link" <ipv6-wg-bounces@ripe.net on behalf of lists@quux.de> wrote: Hi, after now almost 12 years using, working and teaching[1] IPv6 I've come to the conclusion that IPv6 is a mistake and will not work. Therefore the RIPE IPv6 WG should be disbanded and replaced with a new WG that MUST investigate all possible solutions to artificially prolong the live of IPv4 till the day a new successor for IPv4 is created and implemented! Some great ideas[2] are already proposed, some of them already implemented: - Use of NAT - Use of the first Class-A network 0.0.0.0[3] - Use of parts of localhost Class-A network 127.0.0.0 - Use of (parts) of Class-D address space (multicast) - Use of Class-E address space (future use) - Using part of the UDP / TCP port range as extension for the address. Some of the reserved address spaces could also be used. E.g. nobody is using 192.0.2.0/24 for documentation anyway. It should also be investigated to take back legacy IPv4 resources, although the "owners" of these resources might already selling them on the open market. It MUST also be considered not filtering on Class-C[4] bounderies but going for something smaller like /26 or /27 in the global routing table. Also new Class Designations for these prefixes MUST be created. The new successor to IPv4 should not make the same mistakes as IPv6. - IT MUST have NAT - It MUST have Classes - IT MUST have DHCP - It MUST have ARP - It should be possible to drop ICMP the same impact as in IPv4. Many experts I talked to over the years told me that blocking ICMP has no negative impacts. - It MUST only have numbers and dots "." - There should be absolutly no reasons to use "[ ]" in URLs Probably the best way to proceed is to just add one or two octets to the address. One of the reasons for the above is that there are so is so many good documentation already written about IPv4! And people already know about IPv4! Why waste this knowledge and experience? There is also plenty of good software out there that can't work with IPv6[5] Change is bad! People don't want to learn! IPv4! MUST! NOT! DIE! Jens [1] at least trying to teach, as one can see from the great number of people actually using IPv6 with little success [2] https://netdevconf.info/0x13/session.html?talk-ipv4-unicast-expansions [3] https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/commit/?i... [4] a Class-C network is the equivalent of an /24. I was told by experts that the definition of some bit set in the first octet of an IPv4 address is complete and utter nonsense [5] like a 20 year old shell script that is so important for $university that it would be hard for them to implement IPv6!
Dear Jens, you're right. I also think that TCP-IP is unable to cope with today's Internet traffic and we should use a modern protocol like X.25: the problem with IPv6 is not the v6 part, it's the IP. Kindest regards, Olivier On 03/10/2019 11:34, Jens Link wrote:
Hi,
after now almost 12 years using, working and teaching[1] IPv6 I've come to the conclusion that IPv6 is a mistake and will not work.
Therefore the RIPE IPv6 WG should be disbanded and replaced with a new WG that MUST investigate all possible solutions to artificially prolong the live of IPv4 till the day a new successor for IPv4 is created and implemented!
Some great ideas[2] are already proposed, some of them already implemented:
- Use of NAT - Use of the first Class-A network 0.0.0.0[3] - Use of parts of localhost Class-A network 127.0.0.0 - Use of (parts) of Class-D address space (multicast) - Use of Class-E address space (future use) - Using part of the UDP / TCP port range as extension for the address.
Some of the reserved address spaces could also be used. E.g. nobody is using 192.0.2.0/24 for documentation anyway.
It should also be investigated to take back legacy IPv4 resources, although the "owners" of these resources might already selling them on the open market.
It MUST also be considered not filtering on Class-C[4] bounderies but going for something smaller like /26 or /27 in the global routing table. Also new Class Designations for these prefixes MUST be created.
The new successor to IPv4 should not make the same mistakes as IPv6.
- IT MUST have NAT - It MUST have Classes - IT MUST have DHCP - It MUST have ARP - It should be possible to drop ICMP the same impact as in IPv4. Many experts I talked to over the years told me that blocking ICMP has no negative impacts. - It MUST only have numbers and dots "." - There should be absolutly no reasons to use "[ ]" in URLs
Probably the best way to proceed is to just add one or two octets to the address.
One of the reasons for the above is that there are so is so many good documentation already written about IPv4! And people already know about IPv4! Why waste this knowledge and experience? There is also plenty of good software out there that can't work with IPv6[5] Change is bad! People don't want to learn!
IPv4! MUST! NOT! DIE!
Jens
[1] at least trying to teach, as one can see from the great number of people actually using IPv6 with little success
[2] https://netdevconf.info/0x13/session.html?talk-ipv4-unicast-expansions
[3] https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/commit/?i...
[4] a Class-C network is the equivalent of an /24. I was told by experts that the definition of some bit set in the first octet of an IPv4 address is complete and utter nonsense
[5] like a 20 year old shell script that is so important for $university that it would be hard for them to implement IPv6!
On Thu, 3 Oct 2019 at 17:28, Olivier MJ Crépin-Leblond <ocl@gih.com> wrote:
you're right. I also think that TCP-IP is unable to cope with today's Internet traffic and we should use a modern protocol like X.25: the problem with IPv6 is not the v6 part, it's the IP.
Surely the fix is IPv4 over HTTPS (IOH) Aled
Olivier MJ Crépin-Leblond <ocl@gih.com> writes:
Dear Jens,
you're right. I also think that TCP-IP is unable to cope with today's Internet traffic and we should use a modern protocol like X.25: the problem with IPv6 is not the v6 part, it's the IP. Kindest regards,
Back to OSI? I like that! Protocol designed by a real standard body, not some random people from the internet that reach decisions by humming. Jens -- ---------------------------------------------------------------------------- | Delbrueckstr. 41 | 12051 Berlin, Germany | +49-151-18721264 | | http://blog.quux.de | jabber: jenslink@quux.de | --------------- | ----------------------------------------------------------------------------
Hi, I support the proposal. Once a protocol has reached mainstream deployment (as IPv6 has) a dedicated WG might no longer be needed. I mean, there's no IPv4 WG either, right? now back to my day job, full of #IPv6 cheers Enno -- Enno Rey Cell: +49 173 6745902 Twitter: @Enno_Insinuator
Enno Rey <erey@ernw.de> writes:
Hi,
I support the proposal. Once a protocol has reached mainstream deployment (as IPv6 has) a dedicated WG might no longer be needed. I mean, there's no IPv4 WG either, right?
now back to my day job, full of #IPv6
So you are done with your work and Bjoern can get mail from $mailserver run by your new employer? Jens -- ---------------------------------------------------------------------------- | Delbrueckstr. 41 | 12051 Berlin, Germany | +49-151-18721264 | | http://blog.quux.de | jabber: jenslink@quux.de | --------------- | ----------------------------------------------------------------------------
* Jens Link (lists@quux.de) [191004 11:26]:
Enno Rey <erey@ernw.de> writes:
now back to my day job, full of #IPv6
So you are done with your work and Bjoern can get mail from $mailserver run by your new employer?
Nope. At least all their MXes are still legacy. But these changes need time. I haven't given up hope yet. On the bright side: One of the german universities has added IPv6 to their MXes a few weeks before. And since yesterday it really started working. Now listen and repeat: If you deploy IPv6 on your Mailservers: PLEASE check your setup from an IPv6-only site! If your Spamfilter depends on Legacy IP, it WILL break. If your DNS depends on Legacy IP, it WILL BREAK. If your outgoing MX depends on Legacy IP, it WILL break. Bjørn P.S: I know. This looks like preaching to the converted. But the sad truth is: To my current experience, even IPv6 professionals seem to fail on these simple checks. I will never understand, why. If you run Dualstack, also do a FULL check on both protocolls and don't rely on IPv4.
Hi, On Fri, Oct 04, 2019 at 04:21:12PM +0200, Bjoern Buerger wrote:
* Jens Link (lists@quux.de) [191004 11:26]:
Enno Rey <erey@ernw.de> writes:
now back to my day job, full of #IPv6
So you are done with your work and Bjoern can get mail from $mailserver run by your new employer?
Nope. At least all their MXes are still legacy. But these changes need time. I haven't given up hope yet.
so I'll give you guys the generic answer to that one: it's being worked on & it will be there at some point, but I won't be able to tell you about our progress in the interim. I guess the main part here, and for the present discussion, is the "being worked on, will eventually be done" element. On a personal note let me state that Bjoern's mail capabilities are rly present in my scheme of things (they've not yet started to creep into my dreams, but let's see...). You won't be surprised though that not everybody in my daily meetings shares the perspective that they're a top priority ;-). Still rest assured that many smart people over here work hard on things aligned with the IPv6 WG's objectives. #itstimeforIPv6 Also, in all seriousness, feel free to let me know on any channel incl. f2f in Rotterdam what you think we/I should prioritize on, or where things don't work, IPv6-wise. Chris thankfully continues to organize the IPv6 Practitioners' Dinner so that might be an occasion, too. everybody have a great weekend Enno -- Enno Rey Twitter:@Enno_Insinuator
* Enno Rey (erey@ernw.de) [191004 19:09]:
they've not yet started to creep into my dreams, but let's see...)
I wouldn't want that to happen. Take your time.
You won't be surprised though that not everybody in my daily meetings shares the perspective that they're a top priority ;-).
Indeed. Thanks for trying anyway :-) Bjørn
On Fri, Oct 4, 2019 at 4:04 AM Enno Rey <erey@ernw.de> wrote:
I support the proposal. Once a protocol has reached mainstream deployment (as IPv6 has) a dedicated WG might no longer be needed. I mean, there's no IPv4 WG either, right?
Actually it's exactly what I said after I became a co-chair. Our ultimate goal should be to get rid of this group. Let' get IPv6 deployed and go home. Or to the new adventure. As you've said, we do not have IPv4 WG. Unfortunately we are not there yet. -- SY, Jen Linkova aka Furry
On Sat, 5 Oct 2019, Jen Linkova wrote:
On Fri, Oct 4, 2019 at 4:04 AM Enno Rey <erey@ernw.de> wrote:
I support the proposal. Once a protocol has reached mainstream deployment (as IPv6 has) a dedicated WG might no longer be needed. I mean, there's no IPv4 WG either, right?
Actually it's exactly what I said after I became a co-chair. Our ultimate goal should be to get rid of this group. Let' get IPv6 deployed and go home. Or to the new adventure. As you've said, we do not have IPv4 WG. Unfortunately we are not there yet.
Hi, "We" as in "everyone" -- 110% agree. And i must add: there is still a longgggg road ahead. Should the WG try to define new goals, or work more around some measurements in order to support goal evaluation...? Carlos
-- SY, Jen Linkova aka Furry
Jens Link wrote on 03/10/2019 11:34:
- IT MUST have NAT - It MUST have Classes - IT MUST have DHCP - It MUST have ARP - It should be possible to drop ICMP the same impact as in IPv4. Many experts I talked to over the years told me that blocking ICMP has no negative impacts. - It MUST only have numbers and dots "." - There should be absolutly no reasons to use "[ ]" in URLs
There's nothing wrong with a good volte-face presentation, but I'd suggest you avoid positions of opinion-dressed-up-as-sound-technical-argument. This would be relevant to your bullet points about NAT, DHCP, ARP, and the wisdom of using pseudo-in-line signalling protocols and how they should be managed. IPv6 suffers from a good deal of second system effect, and many of the "improvements" it brings to the table have in retrospect turned out to be pointless or some cases quite harmful, e.g. heavy dependence on multicast and how this scales in large networks, the complexity of ND, extension headers, the DHCP va RA debacle, the pathological antipathy of many people towards NAT and many other things. Also, in case anyone is under the mistaken impression that ipv6 is classless, you will probably want to mention that the ietf 6man working group is hopelessly divided on draft-bourbaki-6man-classless-ipv6 (of which I am a coauthor), which has ended up with the draft being blocked. Unless and until that draft - or something similar - makes it through to rfc status, IPv6 remains de facto a fully classful addressing protocol. Otherwise, I'm sure everyone can agree on your only remaining point, namely that colons are better than dots in every conceivable way. This is a sound technical position btw. I will argue the case in exchange for beer and peanuts. Nick
On 4 Oct 2019, at 15:31, Nick Hilliard <nick@foobar.org> wrote:
I'm sure everyone can agree on your only remaining point, namely that colons are better than dots in every conceivable way. This is a sound technical position btw. I will argue the case in exchange for beer and peanuts.
Because beer and peanuts would be at stake, I'll be happy to claim that at signs are better than colons. FWIW I agree "colons are better than dots in every conceivable way". Since we're entertaining bad ideas and non-sequiturs on this thread, it's only fair to add another one. :-)
On Thu, Oct 3, 2019 at 8:34 PM Jens Link <lists@quux.de> wrote:
after now almost 12 years using, working and teaching[1] IPv6 I've come to the conclusion that IPv6 is a mistake and will not work.
Therefore the RIPE IPv6 WG should be disbanded
I suspect that the fact that one member has lost his faith in technology might not be a sufficient reason to close the working group...
and replaced with a new WG that MUST investigate all possible solutions to artificially prolong the live of IPv4 till the day a new successor for IPv4 is created and implemented!
AFAIR you need to have a BoF first before you can get a new working group ;-P https://www.ripe.net/participate/ripe/bof -- SY, Jen Linkova aka Furry
On 03/10/2019 12:34, Jens Link wrote:
Hi,
after now almost 12 years using, working and teaching[1] IPv6 I've come to the conclusion that IPv6 is a mistake and will not work.
Therefore the RIPE IPv6 WG should be disbanded and replaced with a new WG that MUST investigate all possible solutions to artificially prolong the live of IPv4 till the day a new successor for IPv4 is created and implemented!
Dear Jens, I would like to encourage you to submit a proposal to RIPE PC for a BoF that would discuss your topic. I understand that there is no real focused place at RIPE to discuss IPv4 right now (well, there is plenary, but that's more generic) and if you would like to discuss whatever you wish around extending IPv4 - then a BoF would be the right way to go. If you get enough like-minded people to the BoF and present the usefulness of more focused IPv4-enhancement work to the community at closing plenary - then you might get a Task Force or even a Working Group with a goal of discussing and extending IPv4 - well, who knows, right? :) However, I have a feeling that the topic of this working group is IPv6 and I would love to see this WG work and discussion to be focused on the IPv6 topic and its charter. My $.2 worth ;) With best regards, Jan Zorz (no hats ;) )
Jan Zorz - Go6 <jan@go6.si> writes: Jan,
I would like to encourage you to submit a proposal to RIPE PC for a BoF that would discuss your topic.
This is a great idea. I will do so right after finishing this mail
If you get enough like-minded people to the BoF and present the usefulness of more focused IPv4-enhancement work to the community at closing plenary - then you might get a Task Force or even a Working Group with a goal of discussing and extending IPv4 - well, who knows, right? :)
As the feedback on (and off) this list has show there are several people agreeing with my arguments that further work on IPv6 is a waste of time and we need to prolong the live of IPv4 until we find a real successor. Maybe should also discuss if this successor should be defined by IETF or by a real standard body like the ITU. Jens -- ---------------------------------------------------------------------------- | Delbrueckstr. 41 | 12051 Berlin, Germany | +49-151-18721264 | | http://blog.quux.de | jabber: jenslink@quux.de | --------------- | ----------------------------------------------------------------------------
Hi, On Sat, Oct 05, 2019 at 06:52:25PM +0200, Jens Link wrote:
As the feedback on (and off) this list has show there are several people agreeing with my arguments that further work on IPv6 is a waste of time and we need to prolong the live of IPv4 until we find a real successor. Maybe should also discuss if this successor should be defined by IETF or by a real standard body like the ITU.
There will not be anything else on the public Internet in our lifetime. Either it's "IPv6" (with a infinite heavy-tail of IPv4 inside enterprise networks, shielded via application gateways from the Internet anyway) or "IPv4 plus NAT". With the way the Internet is evolving today, IPv4+NAT might just be good enough anyway. End users want lots of TV channels, the big content networks are providing. Everything (including DNS) is done over HTTPS today, which is very NAT friendly. CGN in the eyeball ISP world can easily achieve 10:1 or 50:1 IPv4 oversubscription, and with that, we have enough IPv4 for ever... Well, yes, end-to-end communication will be lost forever. But since the "EVERYONE MUST HAVE A FIREWALL!" crowd broke that for the normal household anyway, it's lost anyway. I still think IPv6 is a more reasonable way forward, but I *expect* to make a shitload of money by fixing people's NAT4444 setups in the next decade... ("our multi-million dollar machines can no longer work because this network thingie broke, and we have neither config nor documentation, so can you please figure out which networks it's trying to connect to and make it work again, QUICKLY?") *sigh* Gert Doering -- NetMaster -- have you enabled IPv6 on something today...? SpaceNet AG Vorstand: Sebastian v. Bomhard, Michael Emmer Joseph-Dollinger-Bogen 14 Aufsichtsratsvors.: A. Grundner-Culemann D-80807 Muenchen HRB: 136055 (AG Muenchen) Tel: +49 (0)89/32356-444 USt-IdNr.: DE813185279
Hello Gert, In-line On 05/10/2019, 20:19, "ipv6-wg on behalf of Gert Doering" <ipv6-wg-bounces@ripe.net on behalf of gert@space.net> wrote: ----%<----%<------ With the way the Internet is evolving today, IPv4+NAT might just be good enough anyway. End users want lots of TV channels, the big content networks are providing. Everything (including DNS) is done over HTTPS today, which is very NAT friendly. CGN in the eyeball ISP world can easily achieve 10:1 or 50:1 IPv4 oversubscription, and with that, we have enough IPv4 for ever... Until the police declares that 1000:1 oversubscription (combine with encryption everywhere) makes their job impossible and enforce some limitation. Well, yes, end-to-end communication will be lost forever. But since the "EVERYONE MUST HAVE A FIREWALL!" crowd broke that for the normal household anyway, it's lost anyway. And, the above (combined with uneducated ISP blocking IPv6 extension headers) is even more frightening.... it forces the Internet in the hands of a couple of cloud providers and is a real ossification of the Internet ----%<--------%<----------- Now, I do not have any solution -éric
Hi Eric, two quick questions (which might be of interest for the group as well):
And, the above (combined with uneducated ISP blocking IPv6 extension headers)
do you have newer numbers than those of http://www.ipv6conference.ch/wp-content/uploads/2016/06/T03-Vyncke-20160616-... (or RFC 7872 which, iirc, is based on research performed earlier than your's)
is even more frightening.... it forces the Internet in the hands of a couple of cloud providers and is a real ossification of the Internet
Can you elaborate on the line of argumentation here? thanks in advance & everybody have a great Sunday Enno -- Enno Rey @Enno_Insinuator https://theinternetprotocol.blog
Hi, On Sun, Oct 06, 2019 at 04:52:09AM +0200, Enno Rey wrote:
is even more frightening.... it forces the Internet in the hands of a couple of cloud providers and is a real ossification of the Internet
Can you elaborate on the line of argumentation here? thanks in advance
& everybody have a great Sunday
Well, the way the Internet has evolved in the past few years is (for "95+% of everything", so those few of you that do "ssh $home" are all on this list, but not relevant in the grand scheme) - all traffic is https - browser vendors are trying to push even DNS on https - all traffic is between "users" and "content providers" - "end-to-end" traffic is relayed via cloud services, because there is no end-to-end anymore (IPv6 routers ship with firewalls on-by-default, IPv4 routers with NAT - which can be circumvented, but every new application would have to deal with it) - due to DoS etc., most important content is hosted by a small handful of very large CDNs and/or anti-ddos providers which, taken all together, means "the Internet works nicely and smoothly for 99% of the users, but rolling out anything *new* is near to impossible unless you happen to be a browser vendor or major content network" ossified... (And the fact that it *does* work nicely and smoothly for most users means it's fairly hard to convince anyone that this evolution might not be what "we" - for some definition of "we" - really want) Gert Doering -- NetMaster -- have you enabled IPv6 on something today...? SpaceNet AG Vorstand: Sebastian v. Bomhard, Michael Emmer Joseph-Dollinger-Bogen 14 Aufsichtsratsvors.: A. Grundner-Culemann D-80807 Muenchen HRB: 136055 (AG Muenchen) Tel: +49 (0)89/32356-444 USt-IdNr.: DE813185279
On 5/10/19 21:52, Enno Rey wrote:
Hi Eric,
two quick questions (which might be of interest for the group as well):
And, the above (combined with uneducated ISP blocking IPv6 extension headers)
do you have newer numbers than those of http://www.ipv6conference.ch/wp-content/uploads/2016/06/T03-Vyncke-20160616-...
(or RFC 7872 which, iirc, is based on research performed earlier than your's)
I can re-run the experiment if folks would find that useful. P.S.: Little pearl not included in the RFC: the filtering also applies to.... IPsec EHs :-o -- Fernando Gont SI6 Networks e-mail: fgont@si6networks.com PGP Fingerprint: 6666 31C6 D484 63B2 8FB1 E3C4 AE25 0D55 1D4E 7492
On 5/10/19 13:18, Gert Doering wrote: [....]
With the way the Internet is evolving today, IPv4+NAT might just be good enough anyway. End users want lots of TV channels, the big content networks are providing. Everything (including DNS) is done over HTTPS today, which is very NAT friendly. CGN in the eyeball ISP world can easily achieve 10:1 or 50:1 IPv4 oversubscription, and with that, we have enough IPv4 for ever...
Well, yes, end-to-end communication will be lost forever. But since the "EVERYONE MUST HAVE A FIREWALL!" crowd broke that for the normal household anyway, it's lost anyway.
It's worse than that: Most IPv4 CPE devices have UPnP support, but IPv6 ones often lack the hooks to punch holes into the fw. SO at the end of the day you get better end-to-end connectivity with IPv4 than with IPv6. e.g., see: https://searchnetworking.techtarget.com/tip/Ensuring-P2P-apps-dont-cause-net... Thanks, -- Fernando Gont SI6 Networks e-mail: fgont@si6networks.com PGP Fingerprint: 6666 31C6 D484 63B2 8FB1 E3C4 AE25 0D55 1D4E 7492
On 23 Oct 2019, at 14:26, Fernando Gont <fgont@si6networks.com> wrote:
On 5/10/19 13:18, Gert Doering wrote: [....]
With the way the Internet is evolving today, IPv4+NAT might just be good enough anyway. End users want lots of TV channels, the big content networks are providing. Everything (including DNS) is done over HTTPS today, which is very NAT friendly. CGN in the eyeball ISP world can easily achieve 10:1 or 50:1 IPv4 oversubscription, and with that, we have enough IPv4 for ever...
Well, yes, end-to-end communication will be lost forever. But since the "EVERYONE MUST HAVE A FIREWALL!" crowd broke that for the normal household anyway, it's lost anyway.
It's worse than that: Most IPv4 CPE devices have UPnP support, but IPv6 ones often lack the hooks to punch holes into the fw. SO at the end of the day you get better end-to-end connectivity with IPv4 than with IPv6.
e.g., see: https://searchnetworking.techtarget.com/tip/Ensuring-P2P-apps-dont-cause-net...
Isn’t this a we broke the network so we must further break the network scenario ? If you remove PAT a lot of the UPnP needs go away and can be replaced by a mix of straightforward fw rules and stateful peeking like PAT residential CPEs do already. Going forward there’s nothing really stoping UPnP being implemented over IPv6 anyway is there ?
Thanks, -- Fernando Gont SI6 Networks e-mail: fgont@si6networks.com PGP Fingerprint: 6666 31C6 D484 63B2 8FB1 E3C4 AE25 0D55 1D4E 7492
On 23/10/19 10:41, Carlos Morgado wrote:
On 23 Oct 2019, at 14:26, Fernando Gont <fgont@si6networks.com> wrote:
On 5/10/19 13:18, Gert Doering wrote: [....]
With the way the Internet is evolving today, IPv4+NAT might just be good enough anyway. End users want lots of TV channels, the big content networks are providing. Everything (including DNS) is done over HTTPS today, which is very NAT friendly. CGN in the eyeball ISP world can easily achieve 10:1 or 50:1 IPv4 oversubscription, and with that, we have enough IPv4 for ever...
Well, yes, end-to-end communication will be lost forever. But since the "EVERYONE MUST HAVE A FIREWALL!" crowd broke that for the normal household anyway, it's lost anyway.
It's worse than that: Most IPv4 CPE devices have UPnP support, but IPv6 ones often lack the hooks to punch holes into the fw. SO at the end of the day you get better end-to-end connectivity with IPv4 than with IPv6.
e.g., see: https://searchnetworking.techtarget.com/tip/Ensuring-P2P-apps-dont-cause-net...
Isn’t this a we broke the network so we must further break the network scenario ?
?
If you remove PAT a lot of the UPnP needs go away and can be replaced by a mix of straightforward fw rules and stateful peeking like PAT residential CPEs do already.
At the end of the day, there's not much of a difference. In the IPv4 world you map external ports to internal ports. And in the IPv6 world you need to punch holes into the firewall, even when the port is not translated.
Going forward there’s nothing really stoping UPnP being implemented over IPv6 anyway is there ?
There isn't, indeed. But in many cases support is simply not there. Thanks, -- Fernando Gont SI6 Networks e-mail: fgont@si6networks.com PGP Fingerprint: 6666 31C6 D484 63B2 8FB1 E3C4 AE25 0D55 1D4E 7492
Am 23.10.19 um 15:26 schrieb Fernando Gont:
It's worse than that: Most IPv4 CPE devices have UPnP support, but IPv6 ones often lack the hooks to punch holes into the fw. SO at the end of the day you get better end-to-end connectivity with IPv4 than with IPv6
That assumes you have at least a public ipv4 address at your home router. A a lot of people (in Germany) cannot fulfill this requirement. Some of them hope/pray for pcp instead of UPnP. (good luck) On the other hand, despite most home routers have a simple firewall - people may change their routers. And cheap router may get an software update, as they got it in former time for dyndns and port forwarding... -- There’s no place like ::1 Thomas Schäfer (Systemverwaltung) Ludwig-Maximilians-Universität Centrum für Informations- und Sprachverarbeitung Oettingenstraße 67 Raum C109 80538 München ☎ +49/89/2180-9706 ℻ +49/89/2180-9701
On Wed, Oct 23, 2019 at 6:45 AM Thomas Schäfer <thomas@cis.uni-muenchen.de> wrote:
Am 23.10.19 um 15:26 schrieb Fernando Gont:
It's worse than that: Most IPv4 CPE devices have UPnP support, but IPv6 ones often lack the hooks to punch holes into the fw. SO at the end of the day you get better end-to-end connectivity with IPv4 than with IPv6
That assumes you have at least a public ipv4 address at your home router.
A a lot of people (in Germany) cannot fulfill this requirement. Some of them hope/pray for pcp instead of UPnP. (good luck) On the other hand, despite most home routers have a simple firewall - people may change their routers. And cheap router may get an software update, as they got it in former time for dyndns and port forwarding...
The only hopes I have for innovation along the edge breaks down into people retaining control over the software in their routers, the elimination of binary blobs, working on improving wifi, and making ipv6 better. I wish more ISPs realized that wifi was the major thing keeping their market alive in light of cellphones everywhere, and despite working on openwrt as much as I can, don't see many ISPs making any investment into better CPE, just riding their rental fees. Binary blobs - well, as examples, I'd like an open source ONT so I could wedge the sch_cake algorithm into one - or just one dsl driver - or - gasp! a cable modem. Coping with the binary blob in just one wifi chip just took years to sort out. ( https://www.spinics.net/lists/netdev/msg615203.html ) As for public IPs, only the gaming market, really, is left with sufficient clout to do anything about it, and even there e2e is dying due to people mounting ddos attacks against visible participants. And nobody seems to know how 5G will implement IPv6. Peering into my cloudy crystal ball, I see a whole generation not knowing what an end to end experience is like, the online gaming experience becoming more like farmville, and routing ssh over https. The internet could have been so much more, and I despair.
--
There’s no place like ::1
Thomas Schäfer (Systemverwaltung) Ludwig-Maximilians-Universität Centrum für Informations- und Sprachverarbeitung Oettingenstraße 67 Raum C109 80538 München ☎ +49/89/2180-9706 ℻ +49/89/2180-9701
-- Dave Täht CTO, TekLibre, LLC http://www.teklibre.com Tel: 1-831-205-9740
On Thu, 3 Oct 2019 at 11:34, Jens Link <lists@quux.de> wrote:
The new successor to IPv4 should not make the same mistakes as IPv6.
[...]
- It MUST only have numbers and dots "."
Maybe we should allow IPv6 addresses to be written in dotted quad format with the quads not being limited to 0..255 but instead 0..4294967295 So instead of writing 2001:4860:4860::8888 we would write 536954976.1214251008.0.34952 which is clearly easier to use. bonus: your old regex ^[0-9]*\.[0-9]*\.[0-9]*\.[0-9]*$ will still work Aled
- It MUST only have numbers and dots "."
Maybe we should allow IPv6 addresses to be written in dotted quad format with the quads not being limited to 0..255 but instead 0..4294967295
So instead of writing 2001:4860:4860::8888 we would write 536954976.1214251008.0.34952 which is clearly easier to use. Why not just use the pure decimal form? 168427777 -> 10.10.1.1 For IPv4 the BSD/Posix-Network stack accepts them anyway
bonus: your old regex ^[0-9]*\.[0-9]*\.[0-9]*\.[0-9]*$ will still work bonus: regex is just: ^[0-9]*$ or even ^\d+$
But joking aside: What I'm missing here is the beauty of IPv6 subnet topology: Because we have so much quibbles it's easier to organize your networks: 2001:db8:<customer>:<vlan>::/64 e.g. 2001:db8:4321:7::/64 In the IPv4 net many are scattered (right now) Anton
Anton Rieger <inrin@jikken.de> writes:
In the IPv4 net many are scattered (right now)
This is good from a security perspective. Confuses the attacker! Jens -- ---------------------------------------------------------------------------- | Delbrueckstr. 41 | 12051 Berlin, Germany | +49-151-18721264 | | http://blog.quux.de | jabber: jenslink@quux.de | --------------- | ----------------------------------------------------------------------------
On Sat, Oct 05, 2019 at 07:10:31PM +0200, Jens Link wrote:
In the IPv4 net many are scattered (right now)
This is good from a security perspective. Confuses the attacker!
Hope the exclamation mark is a ⸮(irony sign) here. All prefixes are public and many tools exist to gather them. For example bgp.he.net. Security through obscurity should not be an pro or contra argument ;) Anton
So why do not representing IPv6 in a base64 encoding? 20 chars + 3 numbers [0-255] for the last 8 bits organized in 4 group 10 char; 8 char; 2 char; 3 numbers example: CompanyNet;inMilano;hr;121 pros: - buzzword or company name preference for choice as happened for dns name; - network partition inside a company would use the last two base64 char + 3 numbers (like ipv4) for a total of 20 bits of space (2^20 should be enough); - optimized usage of bit space, no needs for /64 as minimum for a network; - considering that it is only a representation with a well know alhorithm should be simple to implement. -CIDR mechanism can work as well. Best regards. Mirko Mancini - Telecommunication Engineer ________________________________ Da: ipv6-wg <ipv6-wg-bounces@ripe.net> per conto di Aled Morris via ipv6-wg <ipv6-wg@ripe.net> Inviato: sabato 5 ottobre 2019, 17:51 A: Jens Link Cc: ipv6-wg@ripe.net IPv6 Oggetto: Re: [ipv6-wg] Disband IPv6 WG On Thu, 3 Oct 2019 at 11:34, Jens Link <lists@quux.de<mailto:lists@quux.de>> wrote: The new successor to IPv4 should not make the same mistakes as IPv6. [...] - It MUST only have numbers and dots "." Maybe we should allow IPv6 addresses to be written in dotted quad format with the quads not being limited to 0..255 but instead 0..4294967295 So instead of writing 2001:4860:4860::8888 we would write 536954976.1214251008.0.34952 which is clearly easier to use. bonus: your old regex ^[0-9]*\.[0-9]*\.[0-9]*\.[0-9]*$ will still work Aled Ottieni Outlook per Android<https://aka.ms/ghei36>
participants (33)
-
Aled Morris -
Alex Le Heux -
Alexander Koeppe -
Andreas Härpfer -
Anton Rieger -
Antonio Prado -
Bjoern Buerger -
Carlos Friaças -
Carlos Morgado -
dan＠barletford.com -
Dave Taht -
Enno Rey -
Eric Vyncke (evyncke) -
Fernando Gont -
Gert Doering -
Jan Zorz - Go6 -
Jen Linkova -
Jens Link -
Jim Reid -
Joao Luis Silva Damas -
Job Snijders -
Kai 'wusel' Siering -
Marco Davids (Private) -
Michel Py -
Mirko M -
Nick Hilliard -
Olivier MJ Crépin-Leblond -
Radu-Adrian FEURDEAN -
Riccardo Gori -
Thomas Schäfer -
Tim Chown -
Uros Gaber -
Валерий Солдатов