The focus on high-performance on a single machine resonates a lot with me. I feel there is way too much eagerness to throw in the complexity of distributed computing way too early. We have to learn to make more out of the clock cycles of the computers we have, to make real progress, (and save the planet, or whatnot :) ).
More, I find the choice of D interesting. I'm very much a Gopher these days, but I've been drawn to consider other languages for very math-heavy code solving complex algorithmic problems in Go ... especially those having more of functional properties, as that seems to map better to math-cetric modeling of problems. Been considering mostly OCaml and F# for that so far, but using D in this context is interesting, if it is the case that it has enough of functional programming constructs to make it productive for that.
D's syntax isn't quite as nice as, say, Haskell (which strongly depends on preference anyway), but it has all the necessary tools to write strongly pure, referentially-transparent functions. There's the `pure` and `immutable` keywords that let the compiler statically verify the above properties.
The focus on high-performance on a single machine resonates a lot with me. I feel there is way too much eagerness to throw in the complexity of distributed computing way too early. We have to learn to make more out of the clock cycles of the computers we have, to make real progress, (and save the planet, or whatnot :) ).
More, I find the choice of D interesting. I'm very much a Gopher these days, but I've been drawn to consider other languages for very math-heavy code solving complex algorithmic problems in Go ... especially those having more of functional properties, as that seems to map better to math-cetric modeling of problems. Been considering mostly OCaml and F# for that so far, but using D in this context is interesting, if it is the case that it has enough of functional programming constructs to make it productive for that.