maybe just look at invocations that end in the same result. maybe it's just just a different expression of the same thing, and since we are programmatic, every piece of information expressed is precisely weight and marked, and so,
we know the inputs are same, we know the outputs are same, it's just the userland syntax that's different. it's like driving a car, everything else same, one in which the gear is on the steering wheel, and in the other, beside the driver.
maybe we can look at expressing a bigger program, like a webserver,
in both the styles. then might we get some feel of the ux of the syntax and could posit that this feels better than the other.
building bigger programs, in fact is where the divergence happens. it is much more elegant to express bigger programs via an IO monad, than otherwise.
maybe just look at invocations that end in the same result. maybe it's just just a different expression of the same thing, and since we are programmatic, every piece of information expressed is precisely weight and marked, and so,
we know the inputs are same, we know the outputs are same, it's just the userland syntax that's different. it's like driving a car, everything else same, one in which the gear is on the steering wheel, and in the other, beside the driver.
maybe we can look at expressing a bigger program, like a webserver, in both the styles. then might we get some feel of the ux of the syntax and could posit that this feels better than the other.
building bigger programs, in fact is where the divergence happens. it is much more elegant to express bigger programs via an IO monad, than otherwise.