don at nz.net

From time to time, it comes up that some members wanted to see InternetNZ “stick to its knitting” and just do DNS as “it had been formed to do”. To which I have to react that the running the DNS was not what INZ was formed to do – it was a function that needed doing, but not the reason for its formation.

And given that we’re having long overdue, why are we here, anyway, discussions, perhaps it’s worth exploring where InternetNZ had in fact come from. I think that if as part of that we’re looking at the Society’s Objects, I think it’s worth going over where the existing Objects came from, and how the environment has changed since they were first framed.

So this piece is partly background for folk who don’t know, and partly some starters in thinking about what INZ is for and how its Objects might need to be updated. Obviously, this my personal view, and I’d be interested to see if there are any different takes.

Our parentage

First of all, let’s look at InternetNZ’s “parent” organisations: the Net Society and the Tuia Society.

Tuia was set up by the research & academic communities (basically Universities and CRIs) as an entity to develop the Internet and inter-institutional networking, so we had an umbrella organisation to run the technical needs of the member organisations. The main “output” of Tuia was Tuia Net, the original Internet backbone in NZ, and its interconnections to the rest of the Internet. By 1995, the telecommunications industry was finally starting to understand the Internet, and the clear trajectory was for Internet services to be provided by telecommunications/Internet industry providers, and Tuia members would buy from them; the universities in particular could see that Internet wasn’t something they needed to be doing in the long term. Tuia members ran the NZ DNS and most interconnections to early Internet service providers.

The Tuia Society was not an open organisation; members were institutions, not individuals.

The Net Society was formed out of opposition to Trevor Rogers’ private members bill, the Technology and Crimes Reform bill introduced in 1994, which attempted to regulate content on the Internet by requiring anyone providing Internet access to filter objectionable material. It was a purely volunteer effort, lacking any real resources (although there was a good deal of informal support from within Tuia member organisations).

In the 1993-1996 term, the National Government had an uncomfortably slim majority, and with new parties forming in the lead-up to the first MMP election in 1996, the party leadership was inclined to tread carefully and not just dismiss the bill out of hand lest it lose that majority, despite considerable distaste for the bill among senior Ministers. Thus, the bill passed its first reading in 1994, but by early 1995 it was well buried in procedures. Rogers defected from the National Party in June 1995 to join Ross Meurant’s Conservative Party; with the end of the electoral term, and the Conservatives’ dismal showing in the election, the bill died.

Thus, by early 1995, the Net Society’s primary raison d’être was receding. As often happens with organisations set up to address a specific threat, the Net Society was at risk of simply evaporating through lack of continuing momentum. Meanwhile, Tuia members had been rattled by the Rogers bill – in 1994, there had even been even talk of turning the tap off completely if it became law – and realised that its structure meant that it was incapable of bringing in the wider community and changing its focus to wider Internet issues than the operational considerations for which it was formed.

So the decision was made to wind up Tuia and use its funds to foster a new organisation – ISOCNZ (later renamed InternetNZ) – that could address the challenges faced by the developing Internet, and in the process provide a more substantial vehicle for the Net Society’s activities.

Environment in which ISOCNZ was formed

In 1995, most telecommunications was under the control of Telecom New Zealand Ltd (now Spark), which had been split from the NZ Post Office in the mid ’80s. Up until April 1989, it had been a legislated monopoly (that was two weeks before the first Internet connection went live); competing against it was actually illegal until then. By 1995, Telecom was still effectively a monopoly, owning pretty much all “last mile” (and the majority of “first mile”) infrastructure – and remained so to at least some degree until the early 2010s. So with Telecom moving into the Internet game, there was a great deal of unease about the idea of Telecom “owning” the Internet through its sheer weight.

That’s a big part of where “open and uncaptureable” comes from. The other big part was the fallout from the Rogers bill and the awareness that the Internet was no-longer something that would be left unnoticed by Government and others.

In 1995, the focus was on building rather than using. The Internet clearly had a long way to go; we’d got past getting people to notice it at all, through convincing people that it was something they they should at least think about having, but we were still a very long way from an Internet that was reliable, fast, secure and ubiquitous. Voice and video services were still very much not a part of the ecosystem (mainly because of the bandwidth required), and wouldn’t be for the best part of another decade. Most connections were through dial-up (typically maxing out at 28 kbps), or fairly expensive fixed line connections (typically 48-128 kbps), layered on what was primarily a network designed to carry telephone calls. E-mail was the most used application; WWW was fast catching up. “Social media” – the term hadn’t yet been invented – was e-mail lists, Usenet news, IRC, and a few other pioneering applications.

The focus on build-out meant that most of the people directly involved were from the tech community, and that community was largely young, white male, and relatively privileged. (That’s still true. We’re just not so young any more!) That’s because the people doing the work in the ’90s were born during or before the early-mid ’70s and grew up and were educated in far less diverse environments. Early adopters of the Internet similarly tended to be more white & male than the general population for the same reasons.

While the DNS was definitely on the radar as a thing the new organisation would have to do, it wasn’t the only thing and it wasn’t the main thing. DNS only became a major focus following the formation of the Society as first, name holders had to be convinced to pay for domain names (registrations were not previously charged), and then we had to work out how the service provider space would work technically and commercially. That process was difficult and messy, and only really settled down once the SRS was safely embedded in the early-mid 2000s.

Content issues at the time were largely restricted to dealing with “objectionable” material, and for the most part, “objectionable” was almost synonymous with “pornographic”. It’s worth noting that child sexual abuse panics such as that leading to the Peter Ellis case were in the recent past, and “protecting the children” was something of a rallying cry of Christian conservatives (like Rogers) pushing for censorship of the Internet. As well as the Rogers bill, the Films, Videos, and Publications Classification Act 1993 had fairly recently come into force, and there was some discussion of how that would relate to Internet content.

Security had been much less of an issue up until the mid ’90s. There was very little crime and Internet software often ignored potential abuse. The research community that had done much of the early work tended to be pretty well behaved; while criminality wasn’t unknown, it was much less common and had much less impact (partly because “important” stuff hadn’t moved onto the Internet yet). Encryption was rare; SSL had only just been released, and export of “strong” encryption products (including software) from the US was still under military control. Spam was just starting to become a problem.

There were a lot of things that had yet to be tested or developed. Electronic crime – hacking et c – was not specifically addressed in law in any way; you had to be able to show direct fraudulent loss for the criminal justice system to be interested. A lot of people thought the Internet was somehow outside the law – defamation was infamously tested in 1999-2001 – and things like cyber-bullying and other antisocial effects were yet to be recognised. The Privacy Act 1993 was also relatively new and untested.

Most Internet application services used local infrastructure – email servers, early web et c. Partly, this was because you needed to work with skilled people to provide these, so outsourcing to overseas providers wasn’t really a thing.

The Treaty was a becoming a part of the wider landscape by 1995, but it wasn’t something that anyone in the tech field felt a need to specifically address.

Comparing to now

We no longer have a telecommunications monopoly in the way Telecom was. We have instead created minimal, regulated monopolies around major infrastructure (i.e. Chorus and the LFCs). There are multiple mobile carriers, WISPs, and regional fibre operations on top of that, and for the most part, the regulatory environment allows these to operate relatively easily. While Spark is still the largest operation, it is not overwhelmingly so.

Technologically, the Internet is vastly different. Gigabit capabilities over purpose-built networks are the norm.

The Internet is now sufficiently close to being ubiquitous that most organisations are assuming that it is, which means that alternative ways into many organisations are being increasingly neglected. That presents challenges to reaching those who are still excluded, for geographical, economic, educational or cultural reasons. These challenges are quite different to those experienced in the 1990s.

Social media is vastly more significant than it was, and with it comes the echo chamber effect, and the effect of the loudest voices drowning out the crowd. Extremism has come to dominate discourse.

Since the Internet is now so much closer to ubiquity than it was, many more minorities are present than previously, and are finding that many aspects of the Internet do not suit them – issues include socio-economic ones, plus extremist and intolerant views being directed towards them that they hadn’t previously had to endure, or at least, not to the same extent. They want their voices heard and their issues addressed.

DNS policy is basically done. We’re tidying infrastructure and policy up, but nobody is talking about rolling Councils over its fundamental DNS business model any more, as actually happened in 2000.

Questions around “objectionable” content have less focus on pornography and more on extremist and violent content.

Crime and anti-social activity are now major problems. The traditional image of a teenage hacker in a darkened basement doing it for fun is a long way from reality. Scams, spam, malware, penetrations and disinformation campaigns are perpetrated by sophisticated criminal operations and state actors.

Legislation has moved on. Spam, hacking, bullying and harassment are now the subject of at least some legislation. The threat of mandatory filtering still rears its ugly head from time to time.

Giant international companies dominate the Internet application services space. The nature of the Internet means that it is hard to differentiate between local and international players. This has major implications to trans-national data sovereignty and privacy laws that haven’t been fully explored. Social media in particular is heavily dominated by Facebook and Google, becoming in effect the default regulators of social media content, a role they are ill-suited and not really incentivised to fill.

The Treaty is viewed very differently now to how it was in 1995.

Conclusion

The current Objects seem to be very much a product of their time, and I think our challenge if we’re going to review them is to re-frame them to be relevant not just to the current environment but to any likely future. Also, I think the .nz Principles and Objects should align closely – ideally, we could remove the Principles entirely and simply refer to the Objects, or at least strip them down to a reference and any necessary clarifications.