Background:
How is consensus found in small groups that negotiate important Internet protocols, just using public email listservers? This project analyses three important threads on a public email listserver that contain the email exchanges of a small group of people who were chartered by the IETF with updating our current HTTP protocol, that underlies all our traffic on the Internet. It was an update that was long overdue, and is still in the wake of being finalized. The Internet Engineering Task Force (IETF) is an international Internet standard setting organisation. In general, its mandate is to identify and develop technical standards for the entire Internet, anywhere in the world. It is one of the most unregulated, independent, transnational, and anti-authoritarian Internet governance levels. Given the considerable number of users affected by these standards, vested interests emerge and are vigorously pursued, at times. Only a very small number of studies have shed light on the socio-economic and political role of this organisation, exploring how the IETF works, who benefits from their procedures and publications, and how much deference is paid to corporate interests that claim patents or intellectual property rights.
The selected email threads contain discussions on the required changes to make the next generation HTTP a success. One particularly contested point was how much of the world’s next HTTP should be based on Google’s SPDY (read ‘speedy’), developed by Mike Belshe and Robert Peon while they were employed by Google (both have since left). SPDY was launched by Google in 2009 without IETF standardisation, and included some business enhancing possibilities for their social media platforms (see https://tools.ietf.org/html/draft-mbelshe-httpbis-spdy-00). Their corporate search and social media business model is straightforward. While people are engaged in entering searches or exchanging their views and opinions online automatically all their online entries and machine data are tracked, stored, analysed, bought and sold without public oversight (Hoofnagle & Whittington 2014). Although the possibility for online exchanges is usually viewed positively, online users are divided over the merits of the corporate use of their personal data (PEW 2015, Goodman 2006, Peacock under review).
Additionally, pervasive monitoring (PM) with its continuous analysis of web traffic by governmental agencies was identified as a problem that needed attention in any HTTP update: “Pervasive monitoring is a technical attack that should be mitigated in the design of IETF protocols, where possible” (Farrell & Tschofenig 2014, p. 1). In the wake of this development, the HTTP working group who was chartered with the design of the new HTTP framed PM as relevant to their work. These and other discussions lead to increased their independence from the initial corporate push to merely standardise SPDY.
Objective:
In international organisations people selectively share important knowledge to strengthen in-group coherence, enhancing social status of some while producing social exclusion of others. The current study offers evidence of the social tension between corporate interests and the public goods character of the Internet. Given the salience of the HTTP/2 charter corporate actors, online agency and user advocacy feature prominently in select online discussions.
The current work is not so much about the merits of our new HTTP/2 but focuses on the way rough consensus in this working group is produced using tools of social media. The extent of rough consensus is at the very heart of this project and the use of social media – in this case, an early form of social media, namely public email listservers – serves as my empirical case study. According to an earlier study by Froomkin (2003) discussion in IETF working groups are an ideal case of an inclusive practical discourse that Habermas seems to have envisioned as a public sphere (Habermas 1989, Froomkin 2003). The questions answered in this study are twofold: When does rough consensus succeed or fail and who are the most central people in this process? Furthermore, in how far do online discussions emulate an inclusive practical discourse, particularly if the politics of engineering Internet standards for the entire world is at the heart of the discourse?
As one of the first institutional bodies of standardising best practise on the Internet, a dated but proven email listserver tool is used to exchange ideas and opinions. In the current high-stakes case to produce the next generation HTTP, I assess behavioural objectives and social interactions amongst the working group (WG) members who are all volunteers, vis-à-vis dominant corporate agents to analyse how consensus is build or fails to be build. Most email inputs in the archive contain technical specifications or experiments (testing applications in the wild), but the launch of the charter to construct a new HTTP/2 saw deeply social topics covered like the degree of independence of the IETF which shows the political nature of Internet engineering. In the end, what is made possible or inhibited by technology are very often political decisions made by a small number of people which then go on to affect a large number of people, in this case the entire world (if adoption is successful).
Of particular interest are the initial three months when the group’s charter is discussed. Members weigh into the merits of SPDY versus a more secure HTTP, and debate the extent of their options to have more independence. Given the corporate presence of WG members with heavy ties to Google, Mozilla and a number of other large Internet firms and an already existing ‘new’ HTTP (SPDY) members felt heavily nudged towards ‘rubberstamping’ corporate online interests. But other opinion leaders emerged to insist on broadening their assignment. Eventually, this lead to a more consensual move: SPDY was used as a mere backdrop and while the group independently improved the current HTTP. A small number of people dominated the discourse resulting in the outcome we have today – an improved HTTP with backward compatibility and better options to circumvent PM. According to Haberman, this can be viewed as an outcome of delegation when important decisions are delegated to people who are considered more capable than others. As might be expected, a considerable number of emails focused on the very nature of the WG’s assignment and the envisioned end product, given the urgent need for more thorough HTTP updates. Humour, allcaps, hilarity, and rhetoric are used as members engage in highly contentious exchanges, which is usually a good indication of the political nature of their work. Taken together, a theoretical framework of the public sphere seems appropriate to advance my analysis of the empirical data produced by this working group.
Methods:
My research includes the representation of knowledge, rules, clout, and classifications in the publicly accessible email contributions from working group members (WG). These public contributions are of a very peculiarly nature, because public access to the email list is obscure while at the same time, full public access is granted, if anyone finds them. Currently, I am finishing the analysis of emailed contributions from members of the HTTP/2 WG on the contentious discussions about the extent of their charter and questions surrounding the inclusion of Google’s SPDY for the new HTTP/2. To wit, SPDY had been in use since 2009, and the number of company heads who were adopting it without IETF standardisation was increasing (e.g., Amazon in September 2009 or Netty in February 2012). Discussions on how to recharter the HTTP/2 working group started in January 2012. The first http proposal was officially launched in February 2012 (based fully on SPDY). In the next two months at three different discussion threads ensued containing lively debates on the question how much of our next generation HTTP should be based on Google’s SPDY.
The choice of this case study is based on its importance for the online public sphere, particularly regarding user agency and the fact that social media were used to negotiate technical choices. Contentious issues are identified by the comments of group members and in published articles outside the WG that describe this particular piece of engineered standard setting (e.g., Kristol 2001, Kamp 2015). The full list of email discussions on the HTTP/2 standard is hosted by the IETF HTTP WG (see http://httpwg.github.io/). Additionally, an IETF meeting took place in Paris in March 2012 (IETF 83) that offers some insights of how the discussion of the charter continued face-to-face. All of these materials are publicly accessible, although members expect little public scrutiny beyond those immediately contributing and involved in the discussions. Therefore the archives are characterised as quasi-public in my current work.
The contex
...
