I think the problems is essentially that we had most of our worst minds working on crypto for a long time. Folks though there were incentives to do X, so almost

I think the problems is essentially that we had most of our worst minds working on crypto for a long time. Folks though there were incentives to do X, so almost everyone went out & got busy making cryptocoins. We saw people put their bitshovel in the bitground and mine up bitgold. And suddenly unsurprisingly everyone wants to go bitmining. Now that the bad incentive are much ameliorated, we can begin to see what real enduring uses might be. The bad systems aren't stealing all the oxygen. I'd also say that I think platform evolves at a slower timebase than languages. Posix has been around for omg long now. We were almost a on SysV or very alike until very recent mass migrations off. Http3 is fancy as heck but it does very very nearly what http 0.9 did. Programming languages come and go much faster than these sort of things change. And the next jumps we keep looking to make, most involve pretty significant increases in scope (one counter example might be the Zircon microkernel, https://fuchsia.dev/fuchsia-src/concepts/kernel https://fuchsia.dev/fuchsia-src/concepts/kernel, but it has a sizable platform atop it again).