zlacker

> Roman concrete that we can't reproduce.

That claim seems to date to a particular article written in 2017 that wasn't well sourced. Roman concrete is interesting stuff and has useful properties, but humans have since created concrete mixtures that are far superior. But they're expensive, so it's not too surprising we don't see them getting used in buildings that compare less than favorably to a temple built a couple thousand years ago. Survivorship bias taken to the extreme.

>>rootus+(OP)
I asked my dad about this a while back - he spent his entire career in civil engineering and he said almost exactly what you just did. Basically; modern concrete is far better, but often not built that way.

>>geoffm+08
I haven't seen that theory tested. I can't find any tests or studies anywhere of replicated Roman concrete. And no tests of genuine Roman concrete, either. We don't know how it actually performs. So to make a comparison to modern concrete is specious.

Modern concretes can do a whole lot of stuff Roman concrete can't, because there are so many formulations of it. But if you want to stick a building literally in the ocean and have it never ever disappear, nobody has shown that we can actually do it today.

There's a whole lot of theory and talk by experts, about how we don't need to make it, but if we wanted to, boy would it be easy, but don't worry, modern concrete is just so amazing, you should just use that, for modern use cases, and oh by the way, it would be too expensive to make, even though we haven't actually made it or tried to bring the price down.

There's a world of practical experience needed to claim for a fact that modern concrete is legitimately better, much less that we can actually make it and that it would hold up as we expect. I'm still waiting for concrete evidence.