Convopage
See the entire conversation
Interested in decentralization today? Learn what worked in the 2000s for p2p file sharing:
⚖️ Decentralization is a legal tactic used alongside activism
⌛ Decentralizing the wrong things is a waste of time
☠️ Decentralizing everything is suicide by bad UX
Thread! 👇
41 replies and sub-replies as of Sep 16 2018
Decentralization should shadow the law it is trying to avoid. Chronology of popular p2p file sharing apps:
1999: Napster
2002: Kazaa
2005: BitTorrent
Each app degraded its UX and altered its protocol for better legal protection.
Decentralization for legal protection at the expense of UX:
• Napster used p2p to avoid central hosting
• Kazaa decentralized file search, hurting results and quality control
• BitTorrent left search to others (The Pirate Bay). Terrible onboarding but better legal positioning
Decentralize your tech until activists can takeover:
• The Pirate Bay founders have gone to the ends of the earth to keep it online
• Tor depends on activist exit node operators
Activists *will* come! Coppersurfer runs a Tor relay, Bitcoin node, and a BitTorrent tracker!
What should we decentralize? Look for hacks you can formalize:
• Link-only mp3 sites → Napster skirting hosting
• User hosted Napster servers → eDonkey architecture
• Coin tumblers → privacy coins
• "No logs" VPNs → Tor
• Anonymous domain purchasing → hidden services
Tech vs. law starts w/ decentralized protocols, but is continued by activists, businesses, and OSS
• Pirate Bay was stronger w/ Sweden's copyright law
• Their servers are "raid-proof"
• Pirate Bay proxy sites help w/ ISP bans
• Protocols evolve via OSS communities
Recipe for a successful decentralized protocol:
• Enable a valuable, legally suppressed use case
• Company's products avoid sensitive use case. Let 3rd parties do that
• Tech should be useful outside sensitive use case
BitTorrent checked these boxes. @AugurProject might too!
p2p file sharing history is littered w/ predictions that a new **unstoppable and totally decentralized** protocol will takeover.
It never happened. Users always flocked to the least decentralized system possible.
Big lesson for today's blockchain and decentralization ecosystems
Decentralization is part of a larger legal playbook. More doesn't always help: BitTorrent was less decentralized than Kazaa but more legally resilient.
Their use of OSS, marketing, PR, and implicit reliance on 3rd parties (The Pirate Bay) mattered more
Decentralization is one strategy in a bigger picture, including:
• OSS (eMule harder to kill than eDonkey)
• Nonprofits (pure torrent trackers argue this)
• Diff in intl law (Sweden > US for anti-copyright)
• Reducing impact of takedowns (Pirate Bay hosts own DB archive)
BitTorrent won the legal war by being thorough:
• Website talked about hosting not downloading
• @bramcohen only talked about legal uses. Claimed no profit motive
• Prominent free speech defense on website
• Redundancy via OSS clients
• Outsourced liability to Pirate Bay
Decentralized protocols don't absolve the company of responsibility, it just shifts to the client:
• Limewire had a copyright filter, disabled by default. Court reaction: "What? Make it mandatory!"
• Kazaa had a NSFW filter. Court reaction: "So you could filter copyright too!"
Passing the buck to activists buys room to decentralize. While Hollywood fought Pirate Bay, the BitTorrent community added
• Multitracker support (takedown resilience)
• DHTs (decentralized torrent metadata storage)
• PEX (peer discovery w/o trackers)
Helping The Pirate Bay!
Some look at the file sharing movement and say "BitTorrent didn't win, Netflix+Spotify did!"
Decentralization exploits legal boundaries, exposing rent seeking in anti-competitive industries. BitTorrent created the conditions that allowed Netflix and Spotify to exist
We all won!
The political framing for decentralization is "censorship resistance." The market framing is "anti-competitive market resistance."
P2P file sharing corrected anti-consumer markets for music and movies. Imagine what we can unlock with Bitcoin and Ethereum!
The chronology of the "most popular file sharing app" is roughly:
Napster
Limewire
Kazaa
eDonkey
BitTorrent
Reflecting "minimum viable decentralization" over time. Limewire was early, Kazaa more usable. BitTorrent's UX was bad, but won legal long game
Minimum viable decentralization – John Backus – Medium
If you want people to use your product, decentralize the bare minimum. Lessons from Freenet, I2P, Darknet markets, and Reddit.
medium.com
When decentralizing, half measures bite you in the ass:
🙈 Aimster obscured searches and user IPs from its protocol
💪 BitTorrent disowned search and left IPs in the clear
Courts argued Aimster's conduct was "willful blindness." BitTorrent's choices made its position stronger
Boiling down decentralization movements, we build technology to resist governments installing filters:
💿 Copyright (BitTorrent)
💰 Anti-money laundering (Bitcoin)
💌 Content/communication (Tor)
Legal battles and tech evolution continues until its too hard to install a filter
Once legal pushback enters a decentralized product space, evolutionary pressures start filtering:
• Weaker willed torrent search engines closed
• BitTorrent's protocol evolved
• The Pirate Bay doubles down
Expect same when the US government takes on real crypto companies!
Decentralized technologies mixes legal and objectionable streams together:
• Can't separate darknet market traffic from anti-censorship on Tor
• BitTorrent protocol is unaware of content
You can't throw out the bath water without throwing out the baby, by design.
In a decentralized product market, recklessness buys early growth. Many companies grew by marketing to pirates.
I call BitTorrent an "ugly duckling." It didn't seem like a winner until late in the game.
Here's how to spot ugly duck investments:
Invest in the ugly duckling - decentralization, product/market fit, and the law
Investing in decentralized technology is different.
medium.com
What parts of our tech should we decentralize? How do we know when we're done?
I think a lot of people in crypto are struggling with these questions. I think history provides an answer. Find your "minimum viable decentralization"
Minimum viable decentralization – John Backus – Medium
If you want people to use your product, decentralize the bare minimum. Lessons from Freenet, I2P, Darknet markets, and Reddit.
medium.com
This thread focused on decentralization as a legal tactic to be used sparingly, pulling from the history of p2p file sharing.
This isn't my only deep dive into p2p history and crypto! Check out my thread on market dynamics, governance, and monetization:
Blockchain world can learn a lot from p2p file sharing era
😡 Incentive issues (most shared NOTHING, tons of fake files + viruses)
🤝 Lots of attempted market mechanisms (from game theory to tokens) to fix incentives
💰 Monetization lessons
🍴 Forking vs governance
Thread 👇
☝ If you enjoyed this thread, my blog post "Resistant protocols: How decentralization evolves" goes into more depth: medium.com/@jbackus/resis…
If you don't follow me yet, I tweet regularly about crypto, decentralization, and p2p. Plenty more to come, stay tuned!
Resistant protocols: How decentralization evolves – John Backus – Medium
Decentralization is tough: Too little? You're dead. Too much? Unusable. We can learn a lot from the evolution of p2p file sharing
medium.com
if government keeps cracking down enough on internet freedom, token-economy youtube or twitter - run by anonymous developers that don't even know each other's identies - might eventually actually become economically viable. anti-fragile.
I'm confused. Why would one want to resist anti-money laundering filters? What rent-seeking is being exposed?
In practice, the effect of a government filter is more than the literal filtering. In the US, you need money transmitter licenses (MTLs) in every state. Expensive and time consuming. Killed centralized currency e-gold in ‘09 taking 1mdc w/ it I believe en.wikipedia.org/wiki/E-gold#Ch…
Requiring a license to do business is a hallmark of rent-seeking, big and small. Sometimes it is pure special interest like dentists creating barriers around teeth whitening. Sometimes it is market paralysis: huge cost to enter, uncertainty that gov could cut you off at any time
Did BitTorrent make any key mistakes?
Well, AFAIK Bram wound up not being a fan of raising VC money
'Pave desire paths' in action