Re: [v6ops] IPmix version 01 modification.

[Nick Hilliard <[e-mail]>]

Khaled Omar wrote on 18/11/2018 00:22:
> Look, electronically, what is 0 and what is 1, 0 is 0 voltage, and 1 is 
> 5 voltage, below 5 voltage is 0, 5 or over voltage is 1, so the values 
> of voltages that can represent a 0 are listed on the file 0, 0.1, 0.2, 
> etc and the values of voltages that can represent a 1 are 5, 5.1, 5.2, 
> etc, so the 32 bits can be represented by 5 bits.
> 
> For example, the output can be 00000 for many number of inputs, and each 
> 00000 has a different value of voltage that represents each 0.
> 
> Now it is very clear, looking forward for more discussions anywhere.

Right, I see what you're proposing: fractional voltage to encode 
multiple intermediate values between 0 and 1.  Nice!

Could I make a helpful suggestion?  It would probably be easier if you 
could write an internet draft to encode fractional ipv4 addresses.  So 
instead of having just 192.168.1.0 and 192.168.1.1, you could have a 
range of values in-between, e.g. 192.168.1.(1/2) or 192.168.1.(7/13). 
You'd need to work on the notation a bit because we're already using "/" 
to denote netmask, but I think the idea is sound, and it would solve a 
lot of problems with ipv4 deployment as well as providing a foundation 
for full ipv4 / ipv6 interoperability.  I.e. you'd solve two problems at 
once.

It would be a lot easier to handle this fractional idea at a higher 
level rather than trying to encode it at the electrical signaling level 
(see RFC 1925, Section 6).  The reason for this is that you'd probably 
need to make some changes at the electrical interface layer on IP 
routers to allow your signaling differentiation idea to work, and that 
could present some logistical problems.  Also, if you could make your 
proposal more generic by allowing floating point ipv4 addressing, that 
would probably be an advantage, but I don't think it would be necessary 
to go this far.  Fractional would be more than enough.

Nick