As an avid geek, one of the things I like to do is test things, in good faith. Because of this good faith, I have kept the details below to myself, although I did send a message to a Telecom employee who is active on these forums informing him of my findings. Unfortunately they were a little late in the piece. Now that most users are off the plan, I would like to share the findings of my experimentations.

I thought it was a really good plan and it was only flawed in small ways, namely it lacked even a loose fair use policy & their technology couldn't keep up with the technology of users. I welcomed the opportunity to shift my usage to offpeak times, in exchange for A) more overall usage and B) the freedom of not having to monitor usage like a hawk.

The most obvious problem with their technology was outlined in a number of forums while Big Time was still in operation, though they patched these holes, there are only so many fingers you can stick in the dam. I can't say I ever tried those policies - I was more than happy to have my non browsing traffic shifted to offpeak times.

What I found while I was using the plan, was that their technology also had a terminal flaw in how it affected torrents; in particular, µTP traffic & Teredo. As many people who use bittorrent will possibly know, µTP is:

"an open source cross-platform protocol used in the bittorrent networks and is designed to provide reliable, ordered delivery while maintaining minimum extra delay. It is implemented on top of UDP protocol."

While Teredo is:

"a tunneling protocol designed to grant IPv6 connectivity to nodes that are located behind IPv6-unaware NAT devices. It defines a way of encapsulating IPv6 packets within IPv4 UDP datagrams that can be routed through NAT devices and on the IPv4 internet"

Having many different versions of bittorrent clients available @ my disposal, I eventually tested the throughput during peak & offpeak periods and found that utp & toredo connections were much less affected by Telecom's shaping than vanilla (or even encrypted) TCP connections.

Utorrent clients released prior to the introduction (in the 1.8 version) of the two technologies were completely cut off by Telecom's shaping technology, while the UDP connections allowed transfer rates of in excess of 150kB/s at any time of the day. Now granted, it's not full speed by any stretch of the imagination, you're still looking @ 8.5GB per day of onpeak download traffic as a result if used consecutively, at up to 230GB of onpeak download traffic per month. Furthermore, UL speed was similarly affected; the only thing affecting UL speed was the line speed.

Given that I would estimate that a substantial majority, if not close to 100% of utorrent users would be using a version newer than 1.8.3, and utorrent is by far the most popular bittorrent client in use, you are looking at a substantial population who were able to download excessive amounts of onpeak data.

This combined with the publicised difficulties Telecom had in differentiating between "good" FTP and HTTP (say a driver update for your gfx card) & bad FTP and HTTP downloads (such as from remote severs or from file hosting services), and it was evidently very easy for people to effectively bypass Telecom's traffic management software and use copious amounts of data each month, particularly at onpeak times (potentially 500GB on peak(!)).

I would finally also like to add that after discovering my findings, I stopped using the UDP based technologies for my bittorrenting. I strongly approved of being able to leave my client running with the Ubuntu images seeding in the background for it to communicate with NZ peers during peak hours (and thus alleviate some of the international traffic) and then take advantage of spare capacity offpeak.

K1wi