Philosophy (dist.ribut.us)

Waxnet was envisioned as a protection against siren servers. Allowing users to form communities without being coerced into trusting a central authority that cannot be replaced without dissolving the community.

Siren Servers and the Iceberg

Even a well intentioned siren server cannot avoid the Iceberg. Writing a server based social network is not a challenge. Making it scale is. And running it at scale is expensive. Charging for the service is not an option, the only model that has persisted is advertising. So a service needs to start advertising. Using an advertising service reduces control and returns a smaller cut. Writing an ad service is non-trivial.

The competitiveness of an ad service can be seen as a product of 1) the size of the user base 2) the desirability of the user base 3) information extracted from user base. This pressures the service to increase these dimensions, potentially at the cost of the quality of the service. It also increases the overall cost of running the service, eventually having a significant portion of the revenue raised from advertising going back into seeking advertising revenue.

Eventually, a small percent of revenue and therefore a small amount of influence in decisions is going into the quality of the core service.

Privacy and Security

Meta Privacy

Advertising has established that simply by knowing your friend graph, I can accurately advertise to you. Therefor, keeping the content of communications secure is not enough, the shape of the graph must be obfuscated.

High Security

Most users do not require high security. Instead it is our intention to provide "front-door security". Enough protection to prevent casual invasions of privacy and mass surveillance, particularly for economic profit. The analogy of "front-door security" is that bypassing the common front-door lock is trivial, but provides just enough of a barrier to deter and provide legal protection by showing intent.

For users that do require high security, it will come at a cost of performance. But the network of low-security users will still provide "crowd cover". In essence, we can compare Waxnet to something Vuvuzela, but where Vuvuzela uses dummy packets, Waxnet uses low-security packets. A less technical analogy - imagine trying to follow someone. If we're trying to follow someone on a crowded street without being noticed, even if they take no evasion, it will be difficult because of the crowd. If someone is evasive (by making several detours rather than heading straight to their destination) it will be even more difficult, but their behavior will not stand out.

The Market and the Park

I have no issue with commerce and strongly believe in free markets. But not to the exclusion of all other models. There is a park near where I live. You cannot engage in commerce there and you certainly cannot open your own shop there. Waxnet is park - software must be free and open. If it reaches any degree of success, there will be significant pushback on this point and the community must know from the start that they must help maintain these values.

Ease of Use

It is critically important that setting up Waxnet be easy. The specific goal is that 80% of internet users should be able to setup Speakeasy without any additional help.

A New Framework

Waxnet is a new protocol running atop udp rather than http. Http and the web provide too many hurdles to overcome from DNS to browser fingerprints.

Waxnet is not intended as an anonymizing network for connecting to public services (TOR) or for hidden services (TOR and I2P). It is intended as an anonymous P2P network for P2P applications and procs. In order for the p2p aspect to be evenly scaleable, there are no special peers, therefore, no routing nodes. Instead everyone must route. But to protect users from the risks of running an exit node, only Waxnet traffic can be routed. Scalability above all else is why Waxnet is developing it's own overlay network rather than using an existing infrastructure.

Application Unity

While some degree of forking is inevitable, one of the goals is to keep the number of applications and procs low, attempting to merge similar tools when they arise. A waxnet program should provide a single and specific utility.

For instance, Speakeasy is intended to be our flagship product. It will be seen as a social media network. But from an information or communications theory standpoint, Speakeasy should be platform for narrowcast communication between users with external relationships.

I would like to see applications developed based on other communication classifications. One application could handle all forms of "forum" like activity where discussion is topic oriented and users do not know each other. Another important application is authority based publication - which would encompass everything from Blogging to Wikipedia, and even brochure sites.

Public Good

There is an argument that Waxnet will be used for everything from piracy to criminal conspiracy. And while that is true, many technologies we live with everyday are regularly used in dangerous and criminal ways. Society has always balanced the benefits against the cost. I believe there is value in a platform where users can communicate securely and privately. There is increasing evidence about the negative social effects of a panopticon.

I will also urge the community, particularly the early adopters, to focus on building socially valuable tools and conventions. If the software can be built, the next biggest factor in it's success will be public perception. If Waxnet is viewed as a darknet harboring criminals, it will fail. If Waxnet is seen as the next web, free of advertising and collecting personal information

Side note, I just googled "advertising spying" to find the term "collecting personal information", and instead I came up with a full page of articles on how to use advertising to spy on competitors and users.

Flexibility

This is a networking service. It only works if there is network access. Therefor I want a solid upgrade protocol. And (unless disabled by the user) upgrades happen automatically. The importance of this is that we can focus on getting something that works knowing that we can go back and fix it later if it's not perfect.

Overhead

We incur a lot of network overhead for privacy. Particularly onion routing. For an "average" user (with front-door settings) I want this overhead to be low enough that it's not noticed. This may difficult but I think it's possible. A simple way of thinking of this, is that the overhead consumed by advertising and bad design should be about the same (or greater) than the overhead consumed by Waxnet security.

Close attention should also be paid to observable speed. For instance, email (Speakeasy messaging) rarely suffers from a delay of a few seconds. While voip is highly sensitive.