Hacker News Reader

No experiment is worth doing without your production system at stake

Typically move fast and break things refers to improving a product, not the next move in battleship.

I think it's open to interpretation.

You don't actually know if something works until you've tested in prod.

Adding some community tags to BGP announcements is really trivial and mostly harmless.

Considering the modern role of BGP it's probably a bit like playing battleship on the controls of two nuclear power plants using the power grid.

As I understand it, generator turbines are synchronized in a region and therefore one can exchange info by modulating the frequency (on the opposite side, on a small scale, there is probably a product on the consumer side: network-over-a-power-plug).

https://preview.entsoe.eu/news/2018/03/06/press-release-cont...

https://computer.howstuffworks.com/power-network.htm

If your production systems break when some third-party plays battleship on them, you might want to invest in better production systems.

Indeed. The responsible way to do this would've been using either off-network routers with non-real AS numbers, or even virtual ones in something like Cisco Packet Tracer.