Funny how the dinosaurs in this thread are arguing whether a < T+2 settlement is possible as fiat transactions are not atomic. In forth semester of computer sc

Funny how the dinosaurs in this thread are arguing whether a < T+2 settlement is possible as fiat transactions are not atomic. In forth semester of computer science, I toyed around with Bitcoin (2014?). A client had asked me to build a crypto exchange and so I faced the problem of when to credit a user Bitcoin that they sent to exchange-generated addresses. Turns out the whitepaper specifically addressed this problem and by waiting 6 blocks, Bitcoin was designed to have such low probability of a chain reorganization that I was safely settled (finality in 60mins). That's how I built it, and it's also the way that 99% of centralized exchanges operate today.