Hi Khaled, see comments inline.
On 30.03.2017 at 17:25 Khaled Omar wrote: > This way too is not working, keeping repeating same answers to the > same questions is ridiculous. Since your proposal hasn't changed, I can only give the same answers to the same questions... >> Your IPv10 proposal doesn't solve the IPv6 deployment problems, >> you > basically get an additional IPv10 deployment problem. > > IPv10 does not depends on clients (ISPs or Enterprises) like IPv6 > migration process, it depends on software modification on all hosts' > OSs and routers' OSs to be able to encapsulate both IP version on the > same IP header. How is this different from IPv6? Moreover, for high speed routers, firewalls and so on, IPv6 also required new hardware (so will "IPv10"), thus a router OS update isn't enough in this case. >> IPv10 doesn't allow an IPv6-only host to communicate to an >> IPv4-only > host and vice versa as stated in the I-D. Hint: an IPv4-only host has > got no idea what an IPv6 address is, let alone an "IPv10 address". > Think about a host treats a different version destination as a host > in a different subnet, it will send the packet to the gateway, the > gateway (router) will check the destination packet, and based on its > version (whether it is an IPv4 packet or and IPv6 packet) it will > checks its IP routing table (whether the IPv4 routing or the IPv6 > routing table respectively) to make a routing decision, and each > router on the road to the destination will do the same till reaching > the destination. This isn't the problem I was referring to. An IPv4-only host cannot simply use an IPv6/IPv10 address as destination address since it doesn't have any notion of such addresses. >> As others already pointed out: the proposal is technically flawed > and does not work. > > Please speak with your own tongue, others who stated that had no full > information about how IPv10 works, later they believed it is the best > solution for IPv4 and IPv6 coexistence until reaching a full > migration to IPv6. I already spoke out for myself (https://mailarchive.ietf.org/arch/msg/ipv6/zyoQNbqFLFP_dgXyBKbnWyI5I3U), but several others seemed to have found the same issues, just reread the thread. >> Repeating this over and over again does not work. IMHO you only can >> move forward with a _technically sound_ proposal, otherwise many >> people will regard it as waste of time. > > Simply, if IPv10 is not of your interest, peacefully don't > participate in the discussion, as I stated on my proposal, the Sorry, I tried to help politely - but if you prefer to dismiss any well intended advice and are not open to criticism of your idea, I better stop now. > internet will be divided into two divisions, and that will not be > good for all of us when some new users will be assigned (or already > assigned) IPv6 only addresses and has no access to all internet > resources or a limited access(Google and Facebook). > >> There are various WGs in the IETF that try to work towards better >> solutions. You may not like them, but they are at least rough >> community consensus. > > Science doesn't understand consensus, science understands ideas, > better ideas and the best idea, that can technically deployed easily But science is based on discussion, facts, experiments and proofs. > and in a short time, I can make a solution like IPv6 and think about > forcing people to migrate, but this will not work practically in a > short time, and this is what we are experiencing, also, thinking > about a solution that requires giving training everywhere to let > people deploy it is not a good solution, because still depends on > clients (users), but to make a solution that depends on a few number > of companies developing networking OSs and can be deployed in a short > time with no interventions of a helping technology (like protocol > translation or asking DNS for more efforts). I still don't see why you wouldn't have exactly the same problems for IPv10. > Eventually, IMHO, you will never reach a full consensus by all IETF > members, so it is the time for the responsible ADs and Chairs with > some help from interested participants to make a good impact for > all. Yes, full consensus isn't required in the IETF, rough consensus suffices. If there is enough interest in the IETF, you'll get a working group to move this forward. Best regards, Roland > > -----Original Message----- From: Bless, Roland (TM) > [mailto:[email protected]] Sent: Thursday, March 30, 2017 5:03 PM > To: Khaled Omar; [email protected] Subject: Re: [Int-area] Continuing > IPv10 I-D discussion. > > Dear Omar, > > Am 30.03.2017 um 15:16 schrieb Khaled Omar: >> I think all of you now know about IPv10 and what is the problem and >> how IPv10 can solve it and how it can be deployed in a short >> time. > > - Your IPv10 proposal doesn't solve the IPv6 deployment problems, > you basically get an additional IPv10 deployment problem. - IPv10 > doesn't allow an IPv6-only host to communicate to an IPv4-only host > and vice versa as stated in the I-D. Hint: an IPv4-only host has got > no idea what an IPv6 address is, let alone an "IPv10 address". - As > others already pointed out: the proposal is technically flawed and > does not work. > >> You can ask any question and I'll do my best to give you answers to >> make it clear for everyone so the IPv10 I-D can go forward through >> the IETF standardization process and be published. > > Repeating this over and over again does not work. IMHO you only can > move forward with a _technically sound_ proposal, otherwise many > people will regard it as waste of time. > >> If there is a better solution for this problem I can participate >> freely on its discussion. > > There are various WGs in the IETF that try to work towards better > solutions. You may not like them, but they are at least rough > community consensus. > > Best, Roland > _______________________________________________ Int-area mailing list [email protected] https://www.ietf.org/mailman/listinfo/int-area
