• Up to 200 metres per house (three houses down a right-of-way for instance get 600 metres);
• Multi-unit houses up to three storeys are covered;
• The UFB fibre companies will fund the first $1,000 of installation costs per tenancy for houses with more than three storeys.
• Covers installations by Chorus, and Northpower, Enable and Ultrafast Fibre have already agreed to provide free UFB deployment to end of 2019.
Despite the odd bad installation, and Chorus rolling out the fibre network in a weird fashion with some streets getting only partial coverage, there's little doubt that the UFB was the right choice.
For evidence of that, look across the Tasman.
Australia's National Broadband Network (NBN) that was devised by Labor to use fibre to the premises but it is now "multi technology mix" based instead.
That sounds much better than what it is. The tweaked NBN is better described as a hodge-podge of old and new tech and it didn't end up being cheaper and faster to rollout, despite Liberal-National coalition's promises.
The idea behind the rejigged NBN was to use existing copper, and networks already reaching users like Telstra and Optus' hybrid fibre coaxial cable ones, similar to Vodafone's service in Wellington, the Kapiti Coast and Christchurch.
NBN's hoop-jumping to avoid fibre to the premises at all costs is almost comical.
HFC was cool tech in 1990s, and if the networks are looked after and upgraded, it could be coaxed (sorry) into a few more years' of service. NBN thought the same and bought both Telstra and Optus' HFC networks but are they up to the task? Optus one isn't: last month, NBN said "this isn't going to work" and dumped plans for using most of the Optus HFC network.
It's back to broadband over copper for NBN with new expensive lines being laid, but how to make it go fast enough?
That's tricky, because those pesky laws of physics get in the way, but if you pull fibre to the pits where phone lines to houses are, that's a start. That makes the line to the house shorter, so that the electric broadband signal can reach it.
You're still stuck with VDSL2 tech for the last bit of copper, which is quick, 50-100 megabit per second over short cable runs for downloads, but not a patch on fibre of course. Fibre also goes further, requires less power and can be upgraded almost without a limit.
Copper will increasingly become stranded tech, like Telecom's mobile CDMA network a few years ago.
Telco equipment vendors such as Nokia which swallowed up Alcatel-Lucent are trying to help out the NBN here. They've devised a new complicated technology called XG.FAST that requires the entire wire sheath to be managed by one provider, and which features some amazing noise-cancelling algorithms, wideband high frequency signals and more.
It's fast: recent trials shows that you can squeeze out 5 to 8 Gbps from copper phone lines, for downloads. Imagine that!
That result was achieved in controlled lab trials however, on new copper that's 30-70 metres long and not those decades-old wires coming into people's houses.
XG.FAST uses as much as 500MHz of bandwidth (normal VDSL2 uses 17.5 to 30MHz) and no vendor wants to reveal to me how much power is used by the technology but it will require an electricity feed from customer premises as well as from the distribution node.
There are no commercially available XG.FAST modems yet, and there won't be for a few years. It'll be interesting to see if NBN can obtain XG.FAST modems and at what price in a world that's shifted to optical fibre components instead. Copper will increasingly become stranded tech, like Telecom's mobile CDMA network a few years ago.
Meanwhile, New Zealand UFB customers can get up to 1 Gbps speeds today and fibre broadband delivery in cities at least is no longer a battle. Looks like we got it right then.
