swordfishBob

* Spam and viruses are a big problem, not easy to address. However, if the investment of cash, time and effort proposed for the censor filter were invested in improving security and safety at the ISP level, we might achieve something very significant.

* Most discussion on NBN has related customer connection speeds, with an implication of sufficient backhaul to service those speeds. What about the actual structure and efficiency of the network and of reseller access? If each ISP must still have all their customer traffic routed through their own tunnel heads, and pay the infrastructure owner for specific capacity on backhaul, then the "national" aspect of the network will be carry a lot of waste. The base network should be able to route efficiently on behalf of any/all providers, making peer traffic far more efficient than currently possible. Note that peer traffic includes VPNs for multi-site businesses, and for teleworkers. In the case of teleworkers, the home-office ISP may not be the employer's ISP. Flexible peer routing also facilitates distributed content caching, so major providers (esp video / IPTV) can increase their capacity with minimal impact on backhaul, by putting a cache in each region. Local resources applicable to a town would be accessible at full speed within that town. Traffic between two nearby towns or suburbs need not be routed via capitol city central exchanges.

* There is no easy answer to fairness. Having Telstra build the NBN will further entrench a monopoly that we already don't know how to handle fairly and legally in a "competitive" market. Going apart from Telstra means a fragmented base network and poor peering between Telstra's networks and customers and the new NBN. However, Telstra's comment that they would compete with any other provider suggests that is the most effective way to solve the "incumbent monopoly" question.

* If the NBN is to maximise efficiency and cost-effectiveness, it must be a monopoly. In that respect it should be government-owned. At the very least, it should be prevented from operating in competition with its own customers (retail ISPs), but even then, without competition, it may lose incentive for furture improvements.

* Could the aims be met with a new kind of USO, that specifies good data rates with future increases? Set a requirement for all new services from 2010, with (say) a 5 year window to upgrade all existing services.

* USO should stipulate that where a provider uses mobile/wireless technologies to deliver a fixed-location service, the client is not penalised financially for this decision. This means wireless can be used where sensible (e.g. avoiding long fibre runs per customer in remote locations, or temporarily until a new node is installed in the area) while mantaining impetus for good infrastructure development. Access costs for fixed installs should be in proportion to speed/performance, not access method.

Thanks for listening. Here's hoping someone is able to digest all the comments and achieve something useful in response.