@ioa @phoe I read through most of the recent discussion on the Github issue. It starts civil enough, and then it just starts going downhill.

@loke @ioa Yes. And then it hits the bottom of the hill, and bursts right through it to keep going downwards.

@ioa @loke Well. I'm sorry that you have gone through it. But I guess that this is life.

I will publish a blogpost regarding this encounter in a few hours.

@pfm @ioa @loke as per the provided example.

@mdallastella You are welcome to try and figure it out yourself.

https://github.com/cl-library-docs/common-lisp-libraries/issues/3

@phoe @mdallastella I took a look at some of his projects, which wasn't very hard, given that they all seem to be single-file utility libraries.

I know I'm not a highly regarded expert with decades of Common Lisp experience, which is probably why I have absolutely no idea what this library actually does, even though I'm reading the documentation: https://www.hexstreamsoft.com/libraries/its/

@phoe Yeah, I'm reading it right now... the complains against thread arrows are quite stupid, the main point of Lisp is to be a meta-programming language, allowing to create your own syntax for the problems you're trying to solve.

That guy has been using CL for 14 years, but he has not grasp its core yet.

@mdallastella @phoe Personally, I don't like the threading macros either. I could make a better argument for that if I wanted, but it would boil down to personal preference in the end.

Which ironically is precisely what his argument is. The difference is that he doesn't admit it and instead asserts that his opinion is the only valid one. Which makes sense, given that he even admits that he never reads other people's code.

@loke @mdallastella ...which in turn makes him quite a poor expert since he doesn't, and can't, read other people's code. At least in the way I understand it.

@phoe @loke I didn't know this guy (probably because I'm not into CL too much). Anyway he's fascinating, I spent about an hour reading his rants all over the internet... 😆

I just feel sorry for the Lisp/CL community.

@mdallastella @phoe I read this comment from him:

"If arrows only provide some small value some of the time (and even this I doubt), then I don't think they justify all the overhead (especially cognitive overhead) associated with their use (such as learning to use them and importing them) and especially deciding when to use them or not and how and why."

Now, take that sentence, and apply it to his own micro libraries, which is essentially his entire output.

The irony 🙂

@loke @mdallastella @phoe micro libraries, oof

@mdallastella @phoe Typically it's just better not to invent your own syntax unless you really get some benefit from it. Some CL guys also hate format or loop exactly for these reasons.

As an CL oldtimer, I was initially also quite sceptical of the threading macros in Clojure, but in practice I find them often more clear than n-level deep nesting. Of course, you can use let to achieve similar visual clarity but to me this feels sometimes now like an imperative way of programming. I.e., sometimes this looks like "assign value to this var, use it later randomly whenever you need it" instead of clear composition.

So, to me it's all about preferences and options. Would I introduce the arrow macros into an existing CL codebase? Probably not. Would I consider using them in a new project? Sure, why not.

@schaueho @mdallastella @phoe To join you off on this tangent :-)

Like most abstractions, they can be used to aid or harm readability. In ML, I moved from using |> to increasingly giving names to intermediate values. And made the same procession in Lisp, where I made a threading macro in like 2001 or so, but then increasingly used let* instead. Giving names to things is generally good.

@phoe @mdallastella

(loop for years from 1 below 15
 collecting 'arrogance)

@phoe Oh, no! You will be debunked too! 😹​