Funding - what about the second year?

Stephen,

>For starters, NSI can determine easily how many multi-homes customers they
>have.  

Easily?  How?  And of those, how many will need additional resources
such that they'll need to become members?

>They also know how many backbone providers there are.

No, they don't, after all, it has proven impossible to come up with a
definition of "provider" much less "backbone provider".  However, lets
look at "backbone providers" that I think we all (or at least most)
can agree on:

I think we can all agree UUNet, MCI, and Sprint will likely become
members.  But how about BBN?  They've got (or had) more address space
than APNIC and RIPE.  How about PSI?  Similar situation.  I think they
will, but I don't think they actually need to.

Then you get into when they'll become members, since presumably they
won't become members until they need the registry's services, so if
they have enough address space or ASNs to last them a while, I'm sure
they'd prefer to spend the money on something else.

Looking at the wider audience, you get to guess how many of the smaller
ISPs will decide the fees are too much to justify, how many will wait,
and how many will file anti-trust lawsuits.

Just a few variables, no?

>If we don't know the customer base, we better damn well better find out
>before we even think about how to organize the thing.

I think we know the customer base -- the people who need resources.
Unforunately, as with most businesses I'm aware of, determining the
exact magnitude of your customer base tends to be a bit tricky.

Regards,
-drc