Re: [Int-area] [v6ops] Still need to know what has changed.... Re: IPv10 draft

[Khaled Omar <[e-mail]>]

>> You have IPv4 address inserted into IPv6.

This is in the IP mix draft, not the IPv10 draft, IPv10 has no IPv4 embedded address.

Khaled Omar

-----Original Message-----
From: Vasilenko Eduard <[e-mail]> 
Sent: Monday, September 28, 2020 10:12 AM
To: Khaled Omar <[e-mail]>
Cc: IPv6 Operations <[e-mail]>; int-area <[e-mail]>
Subject: RE: [v6ops] [Int-area] Still need to know what has changed.... Re: IPv10 draft

Hi Khalid,
You have avoided to answer my primary question:
> There is a logical hole in your proposal:
> If it possible to upgrade every host from IPv4, then why not to 
> upgrade it to IPv6 directly? That's it - problem solved.
> Why anybody would need to upgrade hosts to combination of 2 protocols 
> in the data plane? (second header is IPv10) It is not logical. Does not make sense.

You have IPv4 address inserted into IPv6.
Believe you or not - industry does use "stateless translation" name for such solutions.
You have invented 9th version of address translation, that is not fully specified.
I do not see advantages against RFC 6144.

Eduard
> -----Original Message-----
> From: Khaled Omar [mailto:[e-mail]]
> Sent: 25 сентября 2020 г. 21:47
> To: Vasilenko Eduard <[e-mail]>
> Cc: IPv6 Operations <[e-mail]>; int-area <[e-mail]>
> Subject: RE: [v6ops] [Int-area] Still need to know what has 
> changed.... Re: IPv10 draft
> 
> >> If you look inside the draft - you would see that it is additional 
> >> version of IPv6
> Stateless translation with IPv4-embedded addresses again.
> 
> Eduard, There are no any kind of translation used, it is just mixing 
> the two version in the same header, one as a source the other as a destination.
> 
> >> Your solution already exist. It is RFC 6144 - April 2011. Please, read this RFC.
> It has a bit more details then yours.
> 
> This draft uses translators, where is the statement that two versions 
> can exist in the same header and achieve the communication?!
> 
> >> And what you propose to do in the next 20 years when part of host 
> >> would
> already support IPv10, but other part does not?
> 
> Good question, all the devices has to step by step be updated, then we 
> can switch on a flag day to IPv10, so this will give time to 
> developers to first write the code, then apply the code gradually on all OSs.
> 
> >> Additionally I need to inform you that people are not satisfied 
> >> with stateless
> translation, because long transition needs IPv4 per every host, but 
> IPv4 is in shortage.
> 
> We will not need more IPv4 addresses, as new hosts will be assigned 
> IPv6 addresses and still be able to communicate with the IPv4 hosts.
> 
> Khaled Omar
> 
> 
> -----Original Message-----
> From: Vasilenko Eduard <[e-mail]>
> Sent: Friday, September 25, 2020 8:22 PM
> To: Khaled Omar <[e-mail]>
> Cc: IPv6 Operations <[e-mail]>; int-area <[e-mail]>
> Subject: RE: [v6ops] [Int-area] Still need to know what has 
> changed.... Re: IPv10 draft
> 
> Hi all,
> I have looked to the draft. This hurricane is a little groundless.
> 
> Experts,
> What Khaled was trying to invent is not a new IP protocol. It is just 
> a mistake that it was called IPv10.
> If you look inside the draft - you would see that it is additional 
> version of IPv6 Stateless translation with IPv4-embedded addresses again.
> Hence, billions of man-hours that is needed for development of new IP 
> protocol is not needed. It is good news.
> 
> Khaled,
> Your solution already exist. It is RFC 6144 - April 2011. Please, read 
> this RFC. It has a bit more details then yours.
> Just nobody before you was so brave to propose stateless translation 
> directly from every host in the world. Everybody else was thinking 
> about some gateways to keep the majority of hosts intact.
> There is a logical hole in your proposal:
> If it possible to upgrade every host from IPv4, then why not to 
> upgrade it to IPv6 directly? That's it - problem solved.
> Why anybody would need to upgrade hosts to combination of 2 protocols 
> in the data plane? (second header is IPv10) It is not logical. Does not make sense.
> 
> By the way, it is not possible to upgrade every host in the world out 
> of IPv4 (nobody see money to do the job) - but it is the second 
> problem that you would not face because of previous problem.
> 
> And what you propose to do in the next 20 years when part of host 
> would already support IPv10, but other part does not?
> 
> Additionally I need to inform you that people are not satisfied with 
> stateless translation, because long transition needs IPv4 per every 
> host, but IPv4 is in shortage.
> Hence, other translation technologies: 464XLAT, MAP-T/E, DS-lite, lw4o6.
> If you trying to attack translation topic - you need to read all of these carefully.
> 
> Eduard