Re: controversial paper

From: Mok-Kong Shen (mok-kong.shen_at_t-online.de)
Date: 09/30/03 

Date: Tue, 30 Sep 2003 09:29:27 +0200

Mxsmanic wrote:
>
> Mok-Kong Shen writes:
>
> > For fixed length field, assigning sequentially is too
> > difficult to manage.
>
> I'm not convinced of that. Managing a handful of root servers for the
> entire Internet seems to work well enough.

The trouble is you would have to have one central place for
issuing the required number in order to avoid duplicates.
One needs decentralization, meaning that at each level
you have to have certain reserves in the address space
and at no time there is any consecutive part of the address
space that is fully (compactly) utilized (something you
desire, if I don't err). But this is just a fact one has to
live with in practice in my view. If the address space
were chosen really large enough, there wouldn't be any
practical problems till the sun burns out. But, actually,
one does want to save bits, e.g. employing only 128 bits.
If the address space later turns out to be insufficient,
then one enlarges it.

I think that the following analogy could be relevant in
some sense here: In programming for certain kinds of tasks,
one could either employ a large array that is only sparsely
occupied or employ a linked list that has all the entries
containing user data compactly stored together. Convenience,
if not also efficiency, considerations normally lead to the
employment of the first kind of data structure.

M. K. Shen

Quantcast