In your previous mail you wrote: > => globally unique IIDs give a full guarantee as I explained > in a previous mail (with what is this guarantee). globally unique IIDs can give you full guarantees, right, but the implementation of globally unique IIDs can not.
=> if I understand your argument, your concern is for instance we can't guarantee uniqueness of MAC addresses. I agree but MAC address collisions make impressive damage independently (in fact before) of IID stuff so I consider uniqueness is guaranteed by my sledgehammer (to shutdown definitively the device) and my legal department (for the NIC seller). In any case, I do agree with you that globally unique IIDs is a good solution but today it is not available and it don�t seems to be in a short time => they are available... > => no, the real question is who has to cleanup the mess when it fails. > I prefer this guy has the control over the acceptable options too. This is also an important aspect but if the probability of that mess is very low => very low but not zero. and the "solution" to avoid the mess is expensive (for example for handovers), what is better? => if I am in the charge of the network and the price of the solution is not for me I may answer I'd like to get the solution. I believe we have exchanged enough messages to see our opinions are not different about the probability argument validity but are different about who should decide to apply it of not. For me the network manager should get the control of this choice, not the mobile user. Regards [EMAIL PROTECTED] -------------------------------------------------------------------- IETF IPng Working Group Mailing List IPng Home Page: http://playground.sun.com/ipng FTP archive: ftp://playground.sun.com/pub/ipng Direct all administrative requests to [EMAIL PROTECTED] --------------------------------------------------------------------
