Hi Warren, On Tue, Sep 11, 2012 at 8:12 PM, Warren Kumari <[email protected]> wrote:
> > On Sep 11, 2012, at 12:44 PM, Pars Mutaf wrote: > > > > > > > On Tue, Sep 11, 2012 at 5:38 PM, Warren Kumari <[email protected]> > wrote: > > > > On Sep 11, 2012, at 10:31 AM, Pars Mutaf wrote: > > > > > Hi all, > > > > > > Instead of IP translator, I use the new term "IP payload copier". It > copies payload between different IP versions. > > > > Right… And who runs these? And who pays for them? And how many do you > need? And how do you scale them? And how does discovery work? And how are > they different to current NAT64 type things? > > > > W > > > > > > > > > > They are like other routing devices. ISPs pay. > > And their incentive for this is? > Same incentives for buying a router. The copier is like a router but it is not a router. > > > > > NAT64 works by translating the headers if I am not wrong (I just > checked). IP payload copiers that I propose, copy payload from one version > to another. > > > > We need 1 or 2 between this Internet and China's IPv6 Internet. Then > another one, etc. As I see they are not different from backbone routers. > > > > The real question is not how it works, for the moment. The real question > is: What do we want: > > > > 1. Worry about the whole core Internet, others' IP version choices? (a > question that has no end no answer), or > > 2. Change the end nodes. > > You left out: > 3. Not have a small number of devices that break the end-to-end principle, > cause performance issues, end up as monitoring / choke points and make > debugging hard. > I don't understand what you mean exactly, it is not clear. In any case, this is their problem, you cannot intervene in others' affairs. > > W > > > > > > > > > > Pars > > > > > > > > > > > > > > > > > TCP does not work on this because it assumes same IP version for > source and destination. > > > > > > We probably need a new host identifier, and change TCP to use this > identifier but the IP version transition problem is solved forever. > Everybody use the IP version of their preference. This is the end-to-end > princple, we do not care about the core Internet, we change the end-hosts. > > > > > > See details in the paper: > > > > > > http://www.scribd.com/doc/105448105/Discrete-IP > > > > > > > > > On Tue, Sep 11, 2012 at 2:58 PM, Pars Mutaf <[email protected]> > wrote: > > > By "translation" I mean something completely new. It is written in the > paper: > > > > > > Splitting the path to two (or more) parts. One part is IPv4, the other > part is IPv6. > > > > > > Source IPv4 <-> Translator IPv4 > > > Translator IPv6 <-> Destination IPv6 > > > > > > We take the data found in the IPv4 packet, we put it in an IPv6 packet > (and vice versa). > > > > > > By translation, I do not mean translating the information found the in > the headers. We only copy data. (from IPv4 packet to IPv6 packet and vice > versa). See the paper for illustration (Figure 2). > > > > > > How TCP works on this, I don't know yet. I don't have to know it right > now. What is important is "What we want". > > > > > > 1. Worry about the core network, i.e. others' IP version preferences > forever, or > > > 2. Change the end-node TCP design, if necessary to solve this problem. > > > > > > Sending a packet from an IPv4 host to an IPv6 host, connected using > such a translator, is a routing problem. The problem is that TCP assumes > same IP version for source and destination. We need another type of > identifier, perhaps. Such proposals exist already. But again, what is > important here is not the solution. What is important is "What we want". It > is one of the two possibilities mentioned above. The second one, better > respects the end-to-end principle. > > > > > > Pars > > > > > > > > > > > > On Tue, Sep 11, 2012 at 1:38 PM, Arturo Servin <[email protected]> > wrote: > > > > > > The big flaw of your idea is that you suppose that translation > magically works, which it is not true due to the complexity of some > protocols. That is why I am sceptical. > > > > > > You are very welcome to try your idea. Let me know how it goes. > > > > > > /as > > > > > > > > > On 11 Sep 2012, at 11:21, Pars Mutaf wrote: > > > > > >> Hi Arturo, > > >> > > >> On Tue, Sep 11, 2012 at 1:04 PM, Arturo Servin <[email protected]> > wrote: > > >> > > >> On 11 Sep 2012, at 10:55, Pars Mutaf wrote: > > >> > > >>> > > >>> "Do you have some running code?" > > >>> > > >>> These are all ideas and behavior that we adopted without > questioning. We use them in wrong situations generally. > > >> > > >> Which already exists in many forms of translation mechanisms > that we know are not the best way to solve the problem. There are plenty of > documents on the Internet that will tell you why. > > >> > > >> > > >> I believe that people should be able to choose their own solutions. > Would you transition to IPv6 if Discrete IP worked for you? Some entities > may, others may not, others may do it when they have money, yet others may > invent and use IPv7. We cannot know. > > >> > > >> This means that China may use IPv6, Uruguay may not. We see what > people decide, we cannot force them. They use translators if they wish. I > want to develop and test a new IPvX in my university and be reachable to > the world, who can say no? > > >> > > >> Not knowing what is good for others, not knowing the future, helps > technology develop. It also reduces our suffering because we do not try to > know the unknowable. > > >> > > >> > > >>> > > >>> We first discuss the idea, then implement. > > >>> > > >>> The fact that there is running code doesn't mean that it is a good > idea. > > >> > > >> > > >> But if it were, it would help to convince the sceptical like me. > > >> > > >> > > >> I don't see why this would convince you. > > >> > > >> > > >> Also, the fact that there is not running code could mean that > the idea is impractical and infeasible. > > >> > > >> > > >> No you can implement anything. Even the worst idea. > > >> > > >> Pars > > >> > > >> > > >> Cheers, > > >> as > > >> > > >>> > > >>> > > >>> On Tue, Sep 11, 2012 at 12:22 PM, Arturo Servin <[email protected]> > wrote: > > >>> > > >>> On 11 Sep 2012, at 08:20, Pars Mutaf wrote: > > >>> > > >>>> Discrete IP: > > >>>> > > >>>> 1. Can be a very good transition mechanism to worldwide IPv6 > without enforcement (meaning that we have no idea what is good for others) > > >>> > > >>> Do you have some running code? > > >>> > > >>>> 2. Can solve the IP transition problem forever. > > >>> > > >>> Forgive for being sceptical of those all magical technologies > that solve all the problems of the world. Of course, if you had running > code that would change my view. > > >>> > > >>> Cheers, > > >>> as > > >>> > > >> > > >> > > > > > > > > > > > > _______________________________________________ > > > Int-area mailing list > > > [email protected] > > > https://www.ietf.org/mailman/listinfo/int-area > > > > -- > > After you'd known Christine for any length of time, you found yourself > fighting a desire to look into her ear to see if you could spot daylight > coming the other way. > > > > -- (Terry Pratchett, Maskerade) > > > > > > > > > > > > -- > "Working the ICANN process is like being nibbled to death by ducks, > it takes forever, it doesn't make sense, and in the end we're still dead > in the water." > -- Tom Galvin, VeriSign's vice president for government relations. > > > >
_______________________________________________ Int-area mailing list [email protected] https://www.ietf.org/mailman/listinfo/int-area
