* Erik Bais

> -        Assignments smaller than the minimal allocation size, can’t be
> split into smaller assignments, but can be re-assigned as a complete
> assignment.
>
> My reasoning is that it would disallow cutting up small assignments into
> even smaller assignments.

That is would be somewhat illogical IMHO. Assignments and allocations
are two different things. The minimum allocation size has never been
applied to assignments, so why start now? We've never had a minimum
assignment size, at least not in recent years.

> The question that I have is, would the community prefer a transfer
> policy proposal for PI with or without the above stated rule or
> limitation in freedom in transfers of PI.

Without.

I am not at all concerned about the routing table here. There is nothing
in policy nor in the RIPE database software that prevents people from
adding /32 route objects and attempting to advertise them into the DFZ.
There are at the moment 3888 route objects in the database with that are
/25 or longer, but the routing community seem to be able to ignore them
just fine. I don't see how "nano-PI" would be any different, the routing
community won't have any difficulty ignoring those either. After all, a
router couldn't care less if a route is from an inetnum with status
"ASSIGNED PA" or "ASSIGNED PI".

Or to put it another way, we don't need policy to forbid every bad idea
under the sun. Let the routing community decide how they want to deal
with this one.

Tore