Today is the second anniversary of Celestia's mainnet. Two years' worth of iterations, learnings, ups and downs, and thousands upon thousands of Git commits. When you're building something new, you can get some things right, but you also get a lot wrong. We certainly did. I've been working on Celestia for 6 years, and have been in crypto for 15 years. I've seen the industry evolve and mature over time, from the early cypherpunk days in 2010, to the ICO craze in 2017 where everyone believed that Web3 will entirely replace Web2 and we're going to put everything on chain, to today: a more matured perspective on where blockchains actually have product-market fit. Similarly, Celestia also launched with a "build whatever" mindset, trying to be something to everyone, without a clear focus on what kind of chains people were actually going to build on Celestia's high throughput. We burned many cycles going down unproductive rabbit holes. But taking this approach led us somewhere more powerful than we could've planned: we found a hot spot of product activity. Over the past year, numerous high-performance onchain trading products have begun choosing Celestia. Why? Today, Celestia is the *only* way to ship the lowest latency onchain trading venues by any metric. You get: • 2ms latency with rollup frameworks like Sovereign SDK, • 6 second finality on Celestia (100x faster than other rollup base layers), and • ~200k TPS worth of bandwidth. Celestia is the only product that offers this combination, with projects like Bullet, Hibachi, VEX, RISE, Ethereal, and XO Market already pioneering this frontier with Celestia underneath. And while building around the needs of these projects... we now have a short-cut for doubling down in this direction :) More soon.