thinking about how much energy was expended in the 80s trying to decide whether or not Forth should have a separate stack for floating point numbers... and eventually everyone decided that yes, it should, for reasons - but mainly because floats were always wider than ints, and kept in separate machine registers
so forth has retained the separate float stack... even now, when even Forth is transitioning into 64 bit models, and it makes no sense whatsoever to keep double precision floats separately from native ints
i mean they still end up in separate registers, which makes cacheing TOS tougher - but cacheing TOS is outside the Forth model anyway, and code generators like VFX do much more extensive optimisations that can easily encompass that one
The social network of the future: No ads, no corporate surveillance, ethical design, and decentralization! Own your data with Mastodon!