Blog author from Keybase here. Always game for a Hacker News discussion!<p>There's a subtle point I cut from my post for simplicity reasons, but which feels perfect for HN. I've been convinced by Mazières and the Stellar team that the classic "blockchain" works great for native tokens but is extremely dangerous for anything with counterparty redemption. For example, imagine the shitshow after a truly contentious fork, if there are tokens which are supposed to be redeemable with a counterparty.<p>Let's say Deutsche Bank had put €1 billion into colored coins on Bitcoin. Suddenly, after a fork (e.g. bitcoin vs. bitcoin cash), there would be €2 billion IOU's in the wild. The people on each side of that fork would not roll over and die, and it's not simple to say "Oh, whoever Deutsche picks wins." Or even "Whoever has the strongest chain wins." I have a hard time imagining a company would ever take that risk. I worry big companies would never dare to put anything real-world redeemable directly onto, say, Bitcoin or Ethereum, for this reason. They'd just get sued over and over again.<p>The Stellar federated consensus story (HN debates about SCP below [1][2]) has Deutsche Bank as an actual player on the network. If you want DB redemptions then you would include them in your trust lines / quorum slices, and if Stellar fell apart and became partitioned, you would stay on DB's side. All said, it seems significantly faster and more stable for cryptocurrency-to-real-world mappings, both for the consumer and counterparty.<p>Fun discussions:<p>[1] <a href="https://news.ycombinator.com/item?id=9341687" rel="nofollow">https://news.ycombinator.com/item?id=9341687</a> -- of particular note, because it has David Mazières, Vitalik Buterin, and Greg Maxwell all weighing in.<p>[2] <a href="https://news.ycombinator.com/item?id=16125920" rel="nofollow">https://news.ycombinator.com/item?id=16125920</a>