RE: IPv10 (Temp. name IPmix).

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

 



From: ietf [mailto:ietf-bounces@xxxxxxxx] On Behalf Of Khaled Omar
Sent: Tuesday, November 22, 2016 1:53 PM
To: ietf@xxxxxxxx
Subject: IPv10 (Temp. name IPmix).

Dear all,

I would like to ask all of you something that is important for me, when
someone has an opinion, please feel free to say it regarding my proposal,
but you shouldn?t speak with the name of others and telling their opinion,
this happened two times till now.

First, someone was claiming that most of people are not comfortable with the
name IPv10 because it uses a version number and it confuses most people and
it also harm them, and after the conversation I suggested another name
because I will never aim to do that to the participants of my proposal, and
it became IPmix (as a temporary name until it will be standardized with the
IPv10 name), and I still see some people still not comfortable after this
surface modification.

Second, another one was claiming that most of people will not be comfortable
with the modification that will be made due to the change in the packet
format and that it will contain different IP versions in the same packet
header, and about this I would like to say if you don?t feel comfortable
with the efforts that should be made for that new solution, you can simply
stop participating, no one is forcing you I think, but do not talk with the
tongue of others, they have a tongue as well.

I write here for all because I followed what some people suggested for me
and I got some other suggestions conflicts with the first one, and after
following the first suggestion, I got no response till now, so the technical
discussion almost stopped, but I can feel some progress in the background. 

My request was asking for a WG to discuss topics related to IPv10 because I
feel tired from explaining the same things to everyone asking or thinking
the same way and trying to understand the full idea, questions are
repeatable, and if there is a WG we will discuss every issue once at a time.
  

Thanks for all positive participants, and I will be waiting for
collaboration with all of you in the near future.

Best regards,

Khaled Omar



Khaled,

I believe part of the reason you are not getting much response is that your
proposal looks to define something new, while it overlooks existing
definitions that do much of what you are trying to do. Embedding IPv4
addresses in the IPv6 128-bit field exists in many forms already. Packet
format is the trivial part. 

Things your proposal does not address:

- Why IPv4-compatible and IPv4-mapped addresses are not directly applicable
to the problem you are trying to solve?
(hint: you are recreating this particular set under a new version number,
but offering a very incomplete outline of what you are trying to accomplish)

- How this approach mitigates the problem you identified as 'enterprises are
afraid of the service outage...'?
(your approach is even more problematic than dual-stack IPv4 / IPv6, as
there are 2 more choices for src/dst )

- IPv4 applications have been written since the dawn of time to "know" that
an address is ONLY 32-bits, so even if they got an answer from DNS, they
would buffer-overflow on the 128-bit response. How does your approach solve
this when an IPv4 app tries to talk to an IPv6-only app?
(this is one reason that IPv6 was given its own record type)

- How the network is supposed to route a packet?
(Tables for IPv4 and IPv6 are separate, but you expect them to know which
table to look in when the new version number shows up, and it could be
either based on contents rather than version ID.)



In short, creating a new version number does nothing to mitigate the
adoption problems. The primary issue is retiring out a generation of people
that refuse to learn something new, and it doesn't matter if that is IPv6,
IPv10, or IPv99999. This is the primary reason that during the Windows XP
beta, I insisted that there would be no knob to turn IPv6 off. I knew that
if enterprise IT departments had the ability to turn it off, they would.
Unfortunately enough of them complained to management to get the switch
installed before production shipment.  In any case, IPv6 is more widely
deployed than it is given credit for, because it has been in most stacks for
more than a decade now. The gating item now is getting the ISPs off their
collective backsides to turn it on in the networks. That is happening now
that the price of continuing down the IPv4 path is clearer to their
management. 

Tony






[Index of Archives]     [IETF Annoucements]     [IETF]     [IP Storage]     [Yosemite News]     [Linux SCTP]     [Linux Newbies]     [Fedora Users]