Hi everyone,
We are the original authors of the post and thought we'd offer some very brief clarifications since there seems to be some good discussion about both the intent of the piece and our approach to the technology. Our hope is to let you know where we wanted to see the conversation go in. Where it actually goes is, of course, something of a collective choice.
We'll keep this short since some people have already provided good responses in some threads that we don't have much to add to.
First, on useful but ultimately tangential technologies:
fsflover writes:
> I2P does not require a special infrastructure and provides strong anonymity on top of existing networks.
mackrevinack writes:
> the safe network is going to be autonomous so it won't even need anyone to act as an admin. all you will need to do is log in to access your files
Munksgaard writes:
> I think gemini is an interesting alternative to the regular internet. It's deliberately designed to avoid many of the things making the web terrible
Others have suggested similar overlay networks, including GNUnet, Tor, and so on.
As we state in the article, overlay networks are useful, but they are by their very definition reliant upon existing hardware devices to provide the underlying network over which these networks lay. One of the goals of our article is to point out that a hyper-focusing on such overlay networks misses the point of "owning your own infrastructure" at a deeper (hardware) level. What good is an overlay network if your ISP refuses to relay your network frames to the next Layer 2 device?
In other words, LockAndLol said it best in their comment:
> Isn't [Gemini, etc.] just an application layer protocol? How will that create a new hardware network?
It won't, of course. That's not to discount such higher layer protocols. It's just that we're pointing to a side of the mountain that, in our opinion, too few have considered climbing: layer 1, the hardware.
Second, on the path from here to there and self-hosted services as being a better first step for most people than self-hosted infrastructure.
dnautics wrote:
> I think there's a market for something that is like "plug in to your local network, do minimal configuration, you can set up shared folders a la dropbox, and you can access it over the net".
Yes, we agree. In fact, that's how we teach things in our more involved classes and workshops: first, learn to run a service for yourself. Then learn how to build an internetworking connection to someone else, if you're still motivated and find doing this sort of thing fun. (If you don't find it fun, don't keep doing it. Replace "internetworking" with any other vocational trade, like plumbing or residential electrical wiring, or gardening, and the same advice would apply.) For most people, running Ethernet cable from one building to another or creating wireless links from one rooftop to another is a heavier lift than simply running a semi-private service on an RPi or similar with existing network infrastructure. But doing so doesn't resolve the infrastructure ownership problem, of course.
It's important to recognize these two problems as the two related but distinct problems that they are. In our article, we point out also that they are technically orthogonal from one another: one can "own one's own infrastructure" and use it to connect to Facebook, for example, or one can install something like a Mastodon instance on a spare workstation in one's home closet and ask one's "tribe" to use access it over connections originating from the existing (capital-I) Internet infrastructure. From the perspective of an autonomous community, neither of these scenarios is bad per se, they are both simply incomplete.
The point we made in our article is that unlike the situation 20 years ago, installing something like a Mastodon instance (or any of a bazillion other self-hostable services) is far easier and less expensive than it was then, yet most people we encounter (especially politically-inclined but not especially technologically experienced folks) tend to think the opposite is true. The perception that telecoms autonomy is less possible when the capabilities required to achieve such autonomy are actually more accessible is worth countering with more energy than we generally see it countered with. Hence, the article.
Put another way: given that so much more is possible at Layer 1 at far lower price points than has been possible ever before, and also that Layer 7 has seen an arguable overabundance of focus, we feel that hacker mindshare is better spent solving the part of the problem that hasn't received as much attention or honest community investment, which is "who owns the infrastructure," not "what software are you using to share files."
Finally, a few notes on the obvious political implications of this:
* "Who owns the infrastructure" is ultimately a political question, as it is fundamentally a matter of what "ownership" does (or should) mean. There are cynics who seem to think that any prolonged human activity will inevitably result in the same situation as the one we have now. We'll agree to disagree on the grounds that such assertions are at best Stop Energy and leave it at that.
* A number of people identified several sociopolitical forces at play here. For example, helen___keller wrote: "A large part of the complacency of the modern web is that most people (that I can see) are tired of it, whether or not they realize." Again, we agree, which is why our article tried to emphasize the utility and importance of "local services," where "local" means geographically local, not merely located on the same Ethernet broadcast domain or IP subnet as your NIC's current config.
By way of example, most restaurants in the Oklahoma City suburbs do not need a Web site accessible to visitors (or bots) in Tokyo, yet that is effectively what they are paying for when they pay their monthly Squarespace subscription (aka "Web rent"). We're not trying to argue that a global communications network is not useful, because it clearly is. Rather, we're simply pointing out that when a global(ized) infrastructure is the only available option for local coordination, or commerce, it is also clear that there are some obvious misalignments between layperson expectations, economic influences, and governance models that cause conflicts that can be resolved by investing in smaller-scale and geographically-conscious "local" networking solutions. Moreover, these two models (global and geographically local telecoms) do not, technologically speaking, need to be in conflict with one another. They are "not an either-or situation, nor a zero-sum game," as we say in the article.
To borrow a phrase from another recent post of ours ( https://c4ss.org/content/53915 ) that the Trump supporters in this thread are sure to like even less than this one: Reject the idea that successful mobilizations must be large, or that to do anything meaningful you must first do it "at scale." Instead, build coalitions with neighbors and others in your locality by building on relationships already established through earlier work building physical infrastructure together. Coalition means scaling out, not scaling up.
That can begin as simply as simply as sharing your Wi-Fi password with a neighbor in exchange for splitting the monthly Internet bill. This is true physical infrastructure coalition on a tiny, almost imperceptible scale that was damn near impossible just ten years ago. We think that's notable enough to write an article encouraging people to think moderately more ambitiously about what's possible.
That's all.