Liquid mesh
Translating common carriage to the 21st century internet is difficult, because companies and regulators are getting used to cutting across the protocol stack. So much more of this will happen, of course, in the absence of any rules saying discrimination is unlawful, and it will take years to write any such rules.
“Transport” wants to do more discrimination so as to make money for itself, and so is demanding the right to be free to run its “own” network. Many people think we should create and enforce the norm of neutrality, thus ensuring that no applications or content or devices are ever discriminated against.
But applications want to discriminate too. Should email providers be forced to provide address portability? Should search engines be forced to reveal the algorithms that they use, so that they can be checked for fairness? No, you say, shocked. But shouldn't they be treated as common carriers if the transport layer is? Where does the principle of nondiscrimination stop? Can't you imagine law enforcement requiring end-to-end encryption so that no discrimination could possibly occur (no packet inspection possible) but then saying that as a trade it will need the keys to all of this within two hours after any encryption scheme is released?
And what if the transport layer sees itself as a speaker — doesn't it have First Amendment claims (however specious) about wanting to support its own video and gaming services that could tie us up in court for years?
What happened to all of the deep concerns about technical mandates that came out in the broadcast flag context? Do they just disappear at the lower layers? Are the lower layers incapable of innovation? Should they be fixed in one form? Wouldn't it be better just to force access providers to be truthful about what they did, so that consumers could figure out what was going on and decide for themselves?
I am as committed to the ideal of the open internet as the next guy, and my dream is to have OneWebDay support that goal. But the mischief that can be done to our future (in so many unexpected ways) by insisting on statutory and regulatory definition of neutrality seems to outweigh the possible benefits of this path. There is so much nonsense, so much horse-trading, between where we stand now and the glorious goal of neutrality. The sad fact is that Americans don't mind vertical integration one bit, and the duopolists know that. Not only that, but price discrimination in a competitive market is actually a good thing. Now all we need is a competitive market.
I'd rather see a future that doesn't depend on a “third pipe” but that includes broadband internet access that is neither cable nor DSL. I can imagine a network owned by its users, or by a cooperative, or subsidized by a large company that has no interest in controlling use of the network. Our devices will be doing most of the computation, so there will be no way to tell the difference between devices and routing. We'll have network-aware applications, too.
This admittedly techno-determinist view fits with how the internet was supposed to work. Routing is not supposed to be centrally determined, and the idea of mesh networks pushes this even more to the edge — individual devices will make decisions about routing. As long as we don't make this kind of broadband provision illegal (even by accident, by some casual legislative drafting), it will likely emerge in time.
As David Reed might say, we're in a phase transition of sorts, and there are many people who want to force us into hierarchical and rigid solids. Stay liquid, and the outcome will be extremely positive — as humans, we're good at being liquid.
Comments
4 Responses to “Liquid mesh”
Got something to say?
Susan, could you please elaborate on “cutting across the protocol stack”? I think I know what you mean in this context, but I don't want to guess. A quick example or two would be helpful.
People often talk about the “protocol stack” of the internet, moving from transport (lowest) to applications/content (highest). And the idea is that lower layers are not supposed to discriminate against the higher ones. We've been used to the FCC regulating transport, but they're getting into the business of regulating applications — like VoIP. And companies that provide broadband access would like to be able to discriminate in favor of their own voice and video services, rather than letting all comers ride across their network. So both companies and regulators are cutting across the layers of the protocol stack.
Thanks for elaborating.
To counterpoint a little… from a developer's perspective, while the protocol stack model is a very useful conceptual tool, it's not necessarily the best recipe for actual coding. Message passing across abstract boundaries imposes costs. The point at which it becomes “too much” may be a judgement call about engineering tradeoffs without a “best” answer.
There may be analogues in regulatory policy.
Of course, I'm not denying that “too little” structure and abstraction generally results in buggy, unmaintainable spaghetti.
I don