Hacker News Reader

Comment by icen | Hacker News Reader

Why would you expect a k.c? This is b. There's a b.c.

It's referenced in the Makefile.

On a possibly unhelpful tangent, I'm expecting it to look something like this: https://code.jsoftware.com/wiki/Essays/Incunabulum

It sure does: http://kparc.com/b/b.c

I'm so tempted to just respond with "wanker", though I know HN norms don't allow that. Either the code is for human consumption (in which case, fulfil that goal) or you're writing demoscene assembler which displays virtuosity without any positive impact on society.

I think the problem is, we intuitively estimate the readability of code based on the space it takes up, when we should take into account the information density.

That code is extremely readable - if you assume it will take just as long to understand as the entirety of the GCC compiler: https://github.com/gcc-mirror/gcc

Read "A programming language", where APL came from - https://www.amazon.com/Programming-Language-Kenneth-Iverson/... - you might get more understanding why it is the way it is.

It is the mathematical notation, put into, as Dijkstra expressed, "technologies of the past", completed (to be Turing-complete) and made executable.

At least, I think, this is the basic idea.

The code I linked to was written in one session (probably) and it really isn't comparable to GCC in complexity. It's an interpreter, but one where the author has made no attempt to make the process of interpretation legible. He's too busy "achieving" things with his code.

I guess there should be a niche for people like that, but the APL/post-APL community seems to think that Arthur has special computer science knowledge that can be best (or only) expressed in this form. We all know that's bullshit.

Companies pay large licensing fees to use this software. There must be some "positive impact" on their business.

k is probably omitted because it's not open-source.

Reply on news.ycombinator.com