As per Vishnu Som, the Su-57 has suboptimal stealth comparable to the Rafale, and argues the Rafale is the better choice. In contrast, Air Marshal Anil Chopra has argued on India Today that India should directly import two squadrons of the Su-57.

gMum! Today I spent some time thinking about Optimum(@get_optimum) from a validator's point of view. At the end of the day, validators care about numbers. How many slots did I miss? How much reward did I earn? Did my APY improve? The technology behind it is important, but the real question is much simpler. "How does this affect my returns?" That is where Optimum started to make more sense to me. There are many reasons a validator can miss a slot. But some of them come down to the network. Sometimes data arrives too late. Sometimes it never arrives. Sometimes it arrives, but not in time to be useful. Blocks, transactions, blobs, and attestations all need to reach validators quickly. If they do not, a validator can miss an opportunity even if everything else is working correctly. When that happens once, it may not seem like a big deal. But over time, those small misses add up. Two validators can run the same hardware on the same chain and still end up with different results simply because one receives information a little faster. That is exactly the problem Optimum is trying to improve. Instead of sending full messages, mump2p spreads RLNC-encoded shards across the network. Intermediate nodes do not wait for the complete message. They can recode the shards they already have and immediately forward new ones. In simple terms, the goal is to reduce the number of times data arrives too late or does not arrive at all. What stood out to me is that this is not only about improving average speed. It is also about reducing tail latency those rare moments when an important message arrives much later than expected. For validators, those moments can be surprisingly expensive. Most of the time everything works fine. But if a critical block or attestation arrives late at exactly the wrong moment, it can mean a missed slot. One missed slot may not matter much. Over hundreds or thousands of slots, however, those small losses can slowly reduce APY. For large validators, even a small percentage difference can become meaningful over time. I also started thinking about MEV. The sooner a validator receives transactions and bundles, the more options it has when building a block. Better information arriving earlier can mean better block construction, not just faster block construction. That means network propagation can affect more than basic rewards. It can also affect MEV opportunities. Another thing I like about this approach is that it is not based on throwing more hardware or bandwidth at the problem. Instead, mump2p tries to deliver more useful information using the same network resources. The idea is not: "Spend more money to get better performance." It is closer to: "Use the same infrastructure more efficiently." The more I think about it, the more it feels like Optimum is making a simple argument to validators. Maybe part of the revenue you are losing is not caused by your hardware. Maybe it is caused by the way information moves through the network. If the propagation layer becomes faster and more reliable, fewer opportunities are lost. Over time, that can show up in the numbers that validators care about most: Higher APY. Fewer missed slots. Better MEV opportunities. Behind all the RLNC, shards, and recoding, the message feels surprisingly simple. The same infrastructure. The same risk. Better results! :3