[AusNOG] NBN Co Petition

[Mark ZZZ Smith]

----- Original Message -----
> From: Matthew Moyle-Croft <mmc at mmc.com.au>
> To: Mark ZZZ Smith <markzzzsmith at yahoo.com.au>
> Cc: Phillip Grasso <phillip.grasso at gmail.com>; Chris Hurley <chris at minopher.net.au>; "AusNOG (AusNOG at lists.ausnog.net)" <AusNOG at lists.ausnog.net>
> Sent: Thursday, 12 September 2013 6:51 AM
> Subject: Re: [AusNOG] NBN Co Petition
> 
> I think it's not "14 POIs" vs "100+ POIs".  It's 
> finding the right
> middle ground:
> 

Agree.

I don't know if 200 POIs is the right answer, but I think 14 is a really wrong answer. The people who object to the current model seem to commonly characterise their argument as that the original 14 POIs was the right answer because it would be cheap(er) to connect to. The position they seem to be arguing from is that cost of connection to the NBN is the only consideration. 

The right number of POIs, given they are the interconnection points for all of the services being delivered over the NBN, is the number of POIs that can provide the required availability levels for all of the services that are going to be delivered over it. Assuming that what exists today (e.g., the six facilities in Adelaide) is an indication of what is required to deliver the required service availability, 14 POIs isn't enough.

> Creating 14 POIs to do 10M+ services is dumb.  But building a model
> where less than 100k services are available at each and some have very
> high costs to reach is dumb as well.  Especially considering how small
> Australia is as a market and how concentrated the ownership of
> customers is.
> 
> NBNCo is/was/may still be building a "transit" (their word) network
> between all FAN sites to connect them all for management.  The cost to
> extend the remote POIs to more central ones is really just the cost of
> DWDM transponder cards.  (My understanding is it's in the order of 10s
> of millions).
> 
> If we built a model to allow people to connect where they wanted it
> might work - allow NBNCo to charge higher aggregation fees for those
> who aggregate to a central POI, then fine.  But that makes NBNCo run a
> much more complex network.  Maybe do it at "cost" which'll drive 
> down
> equivalent wholesale costs for people to offer their own aggregation.
> ie. no one is going to charge more.
> 
> The current model is dumb because it does allow a very small number of
> players (Telstra, Optus, Nextgen), some of whom have conflicting
> interests in the retail space, to control access to the remote POIs
> (non-metro) and reinvent the games played with wholesale costs for
> transmission played with places like Tas, Darwin, etc.
> 
> My point being that there's no perfect model.  But little work has
> been done to provide a better transition to the NBN (in whatever form)
> nor ensure that current market dominance isn't reenforced unfairly due
> to circumstances that arose previously.  (See my argument for buying
> back Telstra from earlier).
> 
> (I didn't even get into the whole "last mile residential is a SPOF
> anyway so why worry so hard, but that's a can of worms too hard to
> keep shut once opened).
> 
> MMC
>