gitonthescene

In my experience code reviews are generally cursory and the emphasis on “readability” is more about a culture that seeks to treat programmers as fungible. Also complaints about APL’s lack of readability are never about its lack of structure. So I took “readability” to mean something else as should be clear from my previous comment. Is your complaint about its lack of structure? If so would you mind elaborating?

Serious question: Why is readability so important? For me consistency is far more important than anything as subjective as readability. I’d rather be able to reason about a code in its own logic than feel comfortable browsing code without much consistency. In the end all code needs to be understood for its internal logic and notation is secondary.

Are you counting reading code while you’re developing it? That would sound strange to me. If you’re talking maintenance by someone other than the original developer the pattern I’ve seen is minimal engagement with the code and an almost immediate press to replace it. I’d guess you’re talking about an idealized world.

Just to double down here I took a code base written in this style (not exactly atw but inspired by him) and spent about a day expanding it to this point: https://codeberg.org/growler/k/src/branch/expand/a.c My guess is it would only take a week to get it to what people here are calling “acceptable”.

Hard disagree on code being read more often than it is written. I’ve never seen any study of this and in my experience it is not true. Most code reviews end up being perfunctory. Even calling that “reading” is generous.

Who said code written in this style is lazy?

All good decisions are a product of the particular circumstances in which they arise. This post seemed to be about generalizing that process which I would guess comes out of a supposition of fungiblity.

As much as one can use a given style for a personal project so can one for a professional one so long as it fills the given need. Too often (in my view) fungibility is seen as a preeminent requirement and layers and layers of self justifying processes are built on top of that. I’m only saying that’s a choice and the costs and benefits are not as obvious as most suppose.

Also you can minimize risks with redundancy but most presume those costs to be too high. But again this quickly becomes about politics.

Best guess is that your analysis is missing some detail. People not tools write programs. Also any serious discussion here ends up in politics. If you design your software so that the programmers are fungible then the software suffers regardless of your choices.

Are you saying most employers are smart by default??

I think _all_ programming is about finding an appropriate DSL for the problem at hand. First you need to understand the “language” of the problem then you develop a “lingo”.

I’d bet that a lot of the work done with AI assistants is decidedly _not_ clever.

Losing control of a project is likely more due to the programmers on it than the tools they use. IMHO _anything_ done consistently can be reasoned about and if necessary undone.

The zoo is a collection of web interfaces to a number of array languages written by several people including ATW.

Perhaps a more precise question is whether you can write programs as performant as those written in C or Fortran and the answer is it depends on the program (and more likely the programmer). The languages tend to do memory management for you which means giving up some control. Most use “immutable” data structures which force more contraints.

But for the loss of control you get stuff like fancy SIMD implementation for nothing.

All and all there’s a cost/benefit calculation but that ratio can get quite low.

I know you! We worked together for a MF at GS.

You could say you enjoy toodling on the piano but don’t understand why reading music has to be so hard. It doesn’t _stay_ hard and it opens up new avenues.

I may be reading too much into this but it sounds like you’re searching for templates to stimulate ideas similar to how there are examples for smaller puzzle type problems.

I think most sizable stuff is proprietary. I implemented an lsp in an open source K which uses json/rpc. But the open source K is probably best considered a hobby project.

https://github.com/gitonthescene/ngnk-lsp/blob/kpath/k/lsp.k

You might consider joining one of the APL forums if you haven’t already.

Readable is probably in the eye of the beholder but here’s a partially expanded version of ngn/k https://codeberg.org/growler/k/src/branch/expand/a.c

The expansion is mechanical and thus not really at attempt at readability.

It wasn’t Unicode but it wasn’t ASCII either. I think here unicode is probably shorthand for not ASCII.

I don't know. If you're writing Java you may not be working on the same types of problems.

This comment would seem to address the point of that joke

I had a similar-ish project a while ago. I enjoy doing the "Spelling Bee" game in The NY Times Games section. In the comments someone worried that there weren't enough arrangements to keep the game going very long. I used an open source dictionary to generate all possible puzzles restricted by some basic heuristics like never using the letter S, having the total number of possible words in some reasonable range, etc. I found about 23,000 possible puzzles. My next idea was to use google's n-gram statistics to add some sort of "commonly known" heuristic, but my energy for the project petered out.

In any event these languages are great for exploring data in projects like these.

Oh, sorry. I misunderstood the question.

The readme has links to various resources including tutorials: https://codeberg.org/ngn/k

Chet Baker - Let’s get lost Gerry Mulligan - Night Lights

Also check out Bossa Nova by João Gilberto for a similar vibe but a different style.

I appreciate it's probably not possible to share too many details but I wouldn't be surprised if the choice of Java wasn't simply preference. It may have been a problem with the pipeline rather than with K. I.e. a fix might have been available using K but it can be easier (and harder) to just use some out-of-the box solution. I agree that columnar and time series dbs may have caught up with K over the years, but most of the complaints I've heard about K aren't technical.

Any detail whatsoever would make this a more credible claim. I haven’t met many people, including those skeptical of the performance claims, who have called K _slow_. Maybe for particular domains but I’d doubt that includes the kind of quant work that gets done at Millennium.

It’s not at all clear where you read pretentiousness. Is it the mere fact of its existence?

Ngn/k is GPL and thus more restrictive. https://codeberg.org/ngn/k/src/branch/master/LICENSE

Don’t. Feel inspired. Nothing worth learning comes easy.