A Year In Review Part One: Reintroducing The Invisible Internet
In March we had the good fortune to receive support from The OTF Usability Lab and Simply Secure. The download workflow was evaluated from first exposure to the site, then navigation of the download process, and successfully installing the software. After going through this process with Eileen Wagner, we began a months long process of deeper evaluation after the initial heuristic was presented.
What has been observed is that there is a poor success rate for completing the download process, and ambiguity with regards to what I2P/ the Invisible Internet Project actually is.
The next issue is something that other open source and privacy focused projects encounter: a very technical entry point. In the case of I2P, the documentation for new users assumes developer / network engineering knowledge.
At Data Terra Nemo 2, Gina Giampaolo talked about the importance of language. She perfectly illustrated the point about how language can be alienating and lose people. “What if Twitter called themselves an asynchronous pub-sub messaging service with opinionated formatting instead of offering users the ability to see what is happening in the world right now?.” I recommend anyone who has not heard Gina’s talk to listen to it in its entirety.
With more people becoming interested in privacy focused alternatives, we need to think about accessibility in terms of language and not alienating people with tech jargon, or who do not have computer science background. Seeing signing keys and headless installs makes sense to some of us. It might also not make much sense to some of us. While it is important to keep docs intact for the technical community, where it could be changed to accommodate more people, we made those changes.
The next issue is narrative. Traffic and data anonymizing networks have suffered greatly from linguistic triggers and presuppositions. Being called a “Deep” or “Dark” web creates a bias towards the character of the individuals who use these tools and the conversation about the technology and its motives unfairly. Sadly, these presuppositions often go unchallenged and are still being propped up by tech writers and researchers. We have all at some point come across stories about how any place where people can congregate online have been, and continue to be, used for dangerous or illicit activity. However, when your work is called a “dark web” or “deep web,” it applies an instant bias and blindspot to the overwhelmingly damaging behaviours and actions that occur on platforms run by tech giants.
Over the next few posts I will speak about each of these topics in more depth. First, we will start with some examples of how we have tried to address the download process and make I2P less ambiguous.
What is the Invisible Internet / I2P?
First of all , we must address the fact that nowhere on the site is the software download introduced. People visiting the site are prompted right away to download “I2P.” At no point is it clear what people will be downloading, or what that download is.
There is a promise of privacy for communication in the call to action (CTA), but how does it provide that? Is it an app? What kind of communication? Email, social network, messaging? Is there anything it compares to?
Is the Invisible Internet a project or a software? There are also calls to action to contribute as a developer, and research articles and materials.
It seems a bit complicated. This is where we saw the opportunity to “Get I2P” — the software download, and introduce the ecosystem as The Invisible Internet Project.
The ambiguous call to action (that I contributed too…)
Solution:
We formally introduced the project and its software in a new CTA, with new descriptions that explain both. In the second section of the landing page we get to point new users to the documentation they need, and the technically curious to specific documentation about the I2P transport and APIs for app developers. Reviewing the documentation and information architecture for developers and researchers will be the next project in this process.
When a person visited the About section, they were still given no clarity. Instead what is provided is an explanation assuming tech knowledge of networking and cryptography. It also introduces terms that seem to be specific to how I2P works ie “destinations”.
What does this mean?
And am I downloading this?
Four layers of encryption sounds great, I guess?
Messages, tunnels, latency?
I am confused.
The FAQ does not help to clarify. It assumes that a person has already downloaded the software, which is more confusing for people since this information is not backing up any information about what a person is downloading! If we felt that the terrain was already unfamiliar, we are probably lost now. More terminology and unfamiliar ideas for many people (eepsite, router console, exit node, outproxy). And more trust issues surface when a person sees “ why do I see my IP” — help! And what is this now about distributing and storing content? If I did not know how a peer- to -peer network operates, this is some pretty alienating information. And I still do not know what I2P is or what it means to download it.
Solution: Introducing the software and its download workflow.
The I2P Java software ships a whole stack. It includes a router and applications, as well as many options for configuration.
We created a list of everything that is included in the I2P download and gave it a brief description. This gets rid of the ambiguity of what people will be be downloading. We tell you exactly what kinds of communication tools you are going to get and the variety of things you can do while using the I2P network. It introduces people to the apps using familiar terminology. “Eepsites” have been renamed “I2P sites.” I understand website in relation to the Internet. An I2P site is makes more sense language wise as it relates to the I2P network or Invisible Internet.
Download and Installation
Getting people to download the software presents trust issues and assumes risk simply by requiring the downloading of Java in addition to the I2P software and a disrupted workflow. This is already potentially enough to make people turn away. Also, as it was pointed out in the feedback from Simply Secure, there was assumed developer knowledge and jargon on the download page. We needed to evaluate what could be done in order to better support new people through the process.
What does this mean to people who are not developers?
We gave the Download page a review and made some changes. This is where we took the opportunity to introduce again what is included, what to expect during the install process, and making downloading Java a little more friendly.
The first os specific step-by-step guide was created for Windows as well with a link directly from the download page. A guide for Mac users will be put together in the next release. We were very inspired by Mullvad VPN’s example for this.
Getting back to the menu and information architecture, we made changes here as well specifically for new users. A new introduction was written, a software guide created, and the FAQ updated. I went through every post on Reddit from the past three years and we included those questions and reorganized the content on this page. Some new Application guides were written as well.
What did we learn from this process?
It is easier to think about I2P in two ways: one a project, and the other as the software it provides. Once we separate these things more clearly in the documentation, people can just focus on the tools rather than have to interact too much with how the tools are created, or they can better find what they need if they are developers wanting to contribute to the network.
We took into account results that came from a Brand and Identity workshop and exploration with Ura Creative as a guide for how to communicate in the updated documentation.
We should not expect people to be of any skill level by default. The I2P documentation should be written in a way that can guide any user no matter their technical background. It is important to use the most familiar examples and explanations as possible. Finding my own blindspots was a great way to allow me to take the work even further than we initially thought in this process.
While we were exploring Brand and Identity and the information architecture, I participated with Simply Secure on the Decentralization, Off the Shelf (DOTS) research with many others in the decentralized web community. The Maxims that are presented in the report provide an incredibly valuable foundation, strategy and resource for the entire community to use and build on.
-sadie
