This is Iran's third total internet shutdown, but the methodology has evolved into something far more surgical. They didn't just block IP addresses; they severed BGP routes, killed mobile data, and effectively jammed Starlink signals into a dead zone thanks to Russian imports. When the signal itself is murdered, your Tor bridges and VPNs become expensive paperweights.
As builders, we are being out-engineered. We have grown complacent, assuming the "always-on" cloud is a fundamental constant of the universe. But if your software requires a remote handshake to function, it is a liability, not a tool, in a crisis zone. Every application built with heavy reliance on centralized APIs vaporizes the moment the backbone is cut.
We must stop designing for the "connected" illusion and start building for the darkness.
This is my plea to the HN community: stop treating "offline-first" as a niche feature and start treating it as a human right. We need robust, decentralized mesh networks that bypass state-controlled gateways entirely. We need isolated documentation tools and local-first databases that can sync via Bluetooth or physical handoffs.
Build for the 212 regions that went dark last year so that the next time a state pulls the plug, the people aren't left helpless.
a throwaway account for obvious reasons (they have also Chinese tech to track); make your code work when the world goes quiet.
I'm not saying that nobody should ever consider "the state cuts off the internet" as a criteria when deciding what to do, but making that a foundational requirement is like starting out with "handle google-scale" as a requirement when you have zero reason to believe you will.
There are plenty of good reasons for local first apps, but "build for darkness" is pretty far down the list for me.
The sad thing about continuing development of existing technologies is that all reliability, robustness, and multi-purpose capabilities get optimized away over time. In the ideal world, companies wouldn't even sell you hardware or software, they'd just charge for magically doing the one thing you want at the moment, with no generality and no agency on your end.
It's a miracle we still have electric outlets in homes, and not just bunch of hard-wired appliances plugged in by vendor subcontractors.
As opposed to what? Everyone pays the overhead and price of apps designed for things like local-first Bluetooth sync?
This is a situation where the market will prevail and people would go toward (and therefore pay for) apps designed to fit their needs, not apps designed around rare and unusual scenarios.
Build specific tools for specific situations. You won’t get anywhere trying to get all general purpose apps to focus on niche requirements.