So I was sent this video by @QuaiNetwork (a project I know nothing about) that alleges to compare Quai with $kas. I don't have time to watch a 45 minutes "comparison" video, so I decided I would just watch it up to the first major mistake which arrived (surprisingly?) about 5 minutes in. The guy who calls himself "Dr K" claims that GHOSTDAG reorgs are easy "because Kaspa is a UTXO chain" and that it would be much more difficult for "smart contracts or an account system", he then assesses that for that reason "Kaspa will never have programmability". We've heard this kind rubbish since Kaspa launched. Any layer that works above the consensus level (be it UTXO, AAM, EVM, etc.) is agnostic to how the chain is obtained. Any computations that are made, are made with respect to the chain returned by GHOSTDAG. If there is some reorg that changes say the last 10 blocks in the chain, the complexity of recomputing is exactly the same as it would have been in a depth 10 reorg in a regular chain. The (proven and peer reviewed) security of GHOSTDAG assures that only a small suffix of the chain (that is, the ordering of the entire DAG output by the consensus layer) is ever changed. This is exactly the same as the security properties of say Bitcoin or pre-merge Eth. assuring only shallow reorgs. Beyond that point, the complexity of recomputing is exactly the same. Arguing that recomputing is more difficult than DAGs is, essentially, not understanding the blockDAG paradigm at all. I didn't watch the rest of the video, so if a point that's actually salient is made feel free to bring it to my attention. But always remember that the "DAGs can't support smart-contracts" attack vector is futile, and people who use it are either misleading or just haven't properly done their research. When it comes to Kaspa its the same story every time: 1. people tell us what we are aiming for can't be done (often not as an honest observation but in attempt to step over us to promote their own project), 2. we do it anyway. Smart contracts and other forms of programmability will be no different.