3 ms·
More Languages Won't Fix the Computing World
- ggm 4y agoA novel type of language might, but we're entering the odd place which is about the equivalence of types of language mapping to the architecture of instructions on a CPU. At this point, it's hard to say what specific attribute a new language brings which we don't already have choices for. Parallelism for instance is notoriously hard to express, without confusing the neophyte. Designing a language to make asynchronous parallelism easy helps neophytes but that isn't a state of permanency in anyone. FP is said to preserve opportunistic parallelism latent in a problem and permits overall optimisation in parallel ways, when possible, so as a {type of language, problem solved} tuple, it has benefits. It also is a giant mind warp, so introduces its own "can do it, can't explain it" problem set (if somebody posts the joke about monoids, my job here is done) I had high hopes for nocode. I continue to believe graphical presentation of problems has potential, with no evidence, only intuition. It saddens me nobody has taken UNIX piped commands, and instrumented them as cartoon sewerpipes bulging and emitting steam, as the IO chunks through.