On Wed, 2005-02-02 at 09:24 -0500, Bound, Jim wrote: > Using AF_INET6 as only socket and handling both v4 and v6 can only be done > well if the implementation supports a hybrid v4-v6 stack. > A pure dual stack (code path for v4 and code path for v6 see URL pdf > below) is not friendly to use v4-mapped and I will assume all > understand that on this list.
Apparently you don't understand this, because this is exactly what a number of people have been trying to explain to you already. Maybe it now has popped into your mind, can you then also be so nice and to at least acknowledge that you have been trying to spread a lot of nonsense on this subject, not even speaking about the weird stuff you are trying to push into my face. > Basic examples that benefit from v4mapped are any app that runs via inetd > (e.g. ftp, telnet) and for more complex applications like Netscape browser > and Mozilla example removes the need for if and test conditions for address > types. Wrong, as mentioned before, which is why I brought up the example, Mozilla is exactly one of the apps that has a problem: https://bugzilla.mozilla.org/show_bug.cgi?id=175340 They had the issue that binding twice, which is what one is supposed to do would cause issues on a Linux box, as they default to v4-mapped, but on any other platform it would work. > Clearly if a vendor platform does not support a hybrid stack and thus > not optimal for v4mapped it creaates an engineering trade-off. As mentioned before, Windows does not have a hybrid stack, and the current KAME code also has it turned off. Thus most* of the deployed base has a non-hybrid stack. * That is under the assumption that Windows + KAME + Linux together are 'most', which could quite well be true with the number of XP's out there on the market and the fact that KAME is in all the BSD's and a big part of Linux. > The market should not be driven by a few implementations in any scenario is > my hope. Thus you agree that v4-mapped is bad because it does not exist on most deployed implementations? And thus that we could at least put a note on it > Most all implementations support the use and option of v4mapped. See my note above, Window+KAME+Linux. Unless you have a hidden OS somewhere that has ten millions of users. > We shall see how the application market develops and in progress now. > Another example is a large database application vendor can use AF_INET6 > to support both v4 and v6 as one release and it reduces that part of the > code path length to deal with both address types and the user would be > transparent to how v4 or v6 is used by the application. That is absolutely true, unless you try to run the code on a different platform, which does not have this. Thus having you to modify the code anyway, then finding out that it does not work the same everywhere, as one would expect and thus causing a lot of problems which are totally unnecessary. A simple addition to the doc saying: do not use it unless you do not want to be portable. Is enough. > See the IPv6 porting explanation at the URL below: > > http://gsyc.escet.urjc.es/~eva/IPv6-web/ipv6.html Which I mailed to the mailinglist before exactly because of the above, which underlines the issues which we are discussing. Yes, discussing again, because it was 'forgotten' the last time. Do note also that all Eva's examples use getaddrinfo(). On hybrid stack boxes this will thus fail the second bind to 0.0.0.0. Thus the first iteration will bind to "::", the second Maybe if you would read the above and the other messages I sent you would have actually understood this before. Sorry that I am not an native American speaker, maybe that is the issue, nevertheless I am really wondering what the _real_ problem is. On Wed, 2005-02-02 at 09:30 -0500, Bound, Jim wrote: > > > > > I emphatically disagree > > > with Itojun, cmetz, et al referenced and we had this debate > > many years > > > ago, then again had the debate, and that view lost and we > > should not > > > revisit it again. > > > > You mean some people shoved the arguments away without having any > > background in the subject? > > Your statement above was the insult. The insult is against all the people who are being ignored, like me, which is what I pointed you to. See your own message above, for a perfectly good example of this, you most likely did not even read the URL of Eva's site, which I put on the list before mentioning that is the way that it should be done and is being done and is causing problems, see above. The fact that you don't ever say _why_ you disagree doesn't help a bit btw. If you would at least once mention that, instead of trying to curse at people, maybe I could find out what you don't get of my messages and try to put it in another form. > We can discuss offline this is now > not an IETF issue but until we resolve this privately it is best you and > I not talk further in this forum. This more sounds like a "you shut up" than anything else. If you want to do any communications on this subject to me, don't forget to CC at least the IPv6 WG chairs + IAD's, then they can follow the subject and see what kind of weird things you will bring up next against me. Apparently you do not want to discuss this in public, I wonder why actually. It is about how the IETF works, thus it is certainly IETF related. I for sure have no personal relation with you and with the way you are trying to treat me, I am actually very inclined to not even think of doing that at all. If there is anybody who needs to feel insulted it is me, but with that attitude of yours and the messages you are sending I can not even feel far from insulted, only wondering a lot why you are acting in this way. Greets, Jeroen
signature.asc
Description: This is a digitally signed message part
-------------------------------------------------------------------- IETF IPv6 working group mailing list ipv6@ietf.org Administrative Requests: https://www1.ietf.org/mailman/listinfo/ipv6 --------------------------------------------------------------------
