Recently, Simon Morris, a long-time BitTorrent exec, wrote a provocative series of posts on the nature of decentralization, in the wake of BitTorrent Inc's acquisition by TRON.  They are relatively short and a good read:

1. Why BitTorrent Mattered — Bittorrent Lessons for Crypto

2. If you’re not Breaking Rules you’re Doing it Wrong

3. Intent, Complexity and the Governance Paradox 

4. Decentralized Disruption — Who Dares Wins?

There are decades' worth of experience here, which are absolutely relevant for anyone and everyone working in the area of cryptocurrencies, cryptonetworks, and decentralized computing today.

In the second post in the series, Simon makes the argument that the killer feature of decentralized systems is rule-breaking:

"While a decentralized architecture can be effective at routing around a variety of different failures in a network, the type of decentralization that was achieved by Bittorrent (and by Bitcoin for that matter) has enabled routing around rules."

While there is undoubtedly a strong dose of truth here, I think it is a dangerous place to stop.  There is already a narrative that cryptocurrencies and decentralized systems are for pirates and criminals, but if we focus on that alone, we risk missing the more important characteristics and properties of decentralized systems.  It's a little bit like saying the original internet is only good for porn and copyright infringement, and stopping there.

For today, let's focus on one key aspect of decentralized systems -- a characteristic that was fundamental to the creation of the original internet protocols: resilience.

I like this definition of resilience: "an ability to recover from or adjust easily to misfortune or change".

For example: decentralized mesh networking is resilient to centralized telecommunications going offline in the case of a disaster (as happened in NYC during Superstorm Sandy).  USV portfolio company goTenna was founded out of the Sandy experience, and now serves a wide customer base of first responders, law enforcement and military who desperately need communications that are resilient to traditional network failure.

Or, decentralized HTTP/DNS (e.g., IPFS) which is resilient to infrastructure failure and censorship, as demonstrated by IPFS's republishing of wikipedia in Turkey during internet censorship there.  IPFS, generally, is a major improvement to content addressing on the web, adding substantial resilience by detaching physical location from the logical address of content.

Or, a simple example that Joel typically uses: the Bitcoin network has had 100% uptime for 10 years.

These are real, important properties.  Remember, the original internet protocols were designed so that the network could withstand nuclear and other major attacks.  Many centralized systems trade convenience for fragility, and resilience is a real, valuable property.

Coming up, I'll look at other important properties of decentralized systems: platform risk, security, and innovation.