[ prog / sol / mona ]

prog


Lisp advantages over Python/Ruby/Perl/Js/PHP .. other dynamic langs

1 2024-10-17 13:50

I kno I'm risking to get 403'd for this thread, from textboarder Scheme tsar, but YOLO, I've been 403'd so many times I have no hope to stay here & might leave this place (so I may just leave & become friends with feds), so this could be my last thread here, I really want to know

So I just noticed this board focused primarily on Lithp derrived family of langs

I mainly been writing in dynamic langs whole my life, and I've seen your infiltration psy-op squadron on Hackernews & everywhere all over the internet brainwhashing me to learn Lisp, with Ritch Hickey & other more mature 'programming uncles' prophets, evangelists earning gazillions of USD from Oracle.

Is there any practical serious benefit for me ($INSERTDYNLANGNAME)ist to not just learn but actually USE Lisp in my day to day projects, life???

I've tried Dereck Banas yt tutorial & a bit of SICP, made Hanoi Towers, and seen some of MIT video lectures Sus man & Abelson ... It has not done miracles to me idk why, I can't grasp Lithp magic or what's so lovely about it that I'll feel the need that'll must use it everyday everythere.

Lisp kind of may be cool in comparison to C C++ Java 1.4 ... but for person using ruby/python/js I don't see benefits of ditching them for Lisp.

The most oversold overhyped biggest selling feature is Metaprogramming with macro. Well, maybe it does feel slightly more natural do do that in Lisp, write DSLs, define own syntax, or write code that generates code, since everything is a List. But I can do SAME, ALL of THIS & everything in Python AST. I can parse it, I can make DSL, define new syntax, new operators, generate code, eval it on the fly.
So why you keep telling me in my face & insisting me that Lisp is better & I have to use it?

Is it me being stupid here, not grasping some aspects of Lisp, or you trying to justify 50k annual spend on your Comp Sci degree by telling everyone what to do?

2 2024-10-17 19:12

Try writing a couple dozen compilers and you will start to understand what Lisp is all about. If you can't appreciate it yet from what you already know of it, you can only expand your understanding through experience.

3 2024-10-18 08:48

>>2
Elitist wizard.

4 2024-10-18 10:09

is 99,9% of programming jobs in market about building compilers/DSLs?

5 2024-10-18 14:25

>>4
Only if you work at Microshaft and have Alyssa P. Hacker, Ben Bitdiddle, Eva Lu Ator, etc. as your colleagues.

6 2024-10-20 06:19 *

>>3
Thing is, I won't insist you have to use it. You started this thread. You asked. If you don't actually care, I don't care either.

Lisp is a better tool for certain jobs. Hard jobs. Stuff outside most people's comfort zone. If all you're writing is glue code to make one API talk to another API, and you don't need it to run on any particular CPU, and nobody will care if you burn a million cycles per arithmetic operation, you can write whatever in Python and that's fine. In fact, you can get paid loads of money for it and it's none of my damn business.

For some other things, having a more intimate knowledge of Lisp will make your life easier and free you to concentrate on the problem at hand, but only if you've already invested the time and energy to develop those skills. If you run into such a problem, it probably won't be worth it to take a detour into building your Lisp muscles first. That's just yak shaving. Go deal with the actual problem.

I don't have anything profound to say. You can develop your Lisp skills by goofing off and doing shit that doesn't matter, with a small chance it might pay off someday as a productivity multiplier. If you can't afford to goof off and try learning random stuff, then I guess you just can't. Is that sad, or is it just reality? I don't really care. It's not my call to make. Perhaps you will learn to be at peace with yourself.

7 2024-10-24 12:38

>>6
the >>3 have not started this thread.

I know who started, and we care. Your post/reply >>6, is so vague and abstract without comparison of code side by side, makes me wonder if you posted gpt reply, or ai actually writes & understands subject better than you.

oh, and again, I'm not >>3 and not the one who wrote >>3 `s reply/post.

If all you're writing is glue code to make one API talk to another API, and you don't need it to run on any particular CPU

I want to speak for 99,99% of generic general programmers out there. Most of them are just writing glue code, mostly for web apps, which could be split further to backend & frontend. that being run on x86-64 mostly. And ARM/M1 lately.

watta heck you talking about man, you speaking like your lisp code involves some specific assembly instructions, like r u trollin with 'sed s/C/lisp' ?

lisp/scheme/clojure is high level, GC'ed lang, it's not a low level perfomance oriented with manual memory management, like you painting it to be.

And compilers you say... LLVM, GCC is not coded in Lisp family dialects. Some other langs (& rust) made use of OcaML for compiler.

OP is just saying that aggressive functional push for Lisps, backed by corporate money & shills, is not completely justified to make Ruby/perl/php/python/js developers switch to Lisp family dialects. e.g. it's not a better option for them (aka most of today developers) to spent valuable time/money/calories/resources to switch.
And OP is pretty clear , said how is main selling features of Lisp : DSLs, metaprogramming, introspection reflection - implemented just fine in other said langs. So aggressive religious cult around Lisp dialects are not economically/rationally justified.

8 2024-10-24 12:53 *

I know about Carp, and some low level dialects of Lisp/Scheme designed to run on microcontrollers.

But that is such a micro portion of marketshare.

9 2024-10-24 17:21 *

>>7

Your post/reply >>6, is so vague and abstract without comparison of code side by side, makes me wonder if you posted gpt reply, or ai actually writes & understands subject better than you.

I was phoneposting whilst intoxicated. Sorry for not posting code examples. I still won't post any but I'll reply again later to clear up some vagueness.

10 2024-11-01 10:40

>>4
Refresh your memory of Abelson's foreword to EOPL.

11 2024-11-13 22:15

Shit, colleagues, do you eat shit?
The other colleague is right, and I'll tell you something else: 99% of the code written out there could just as well be written in server-side-includes, which is Turing-complete.

12 2024-11-14 12:21 *

>>11 Now I want to see an OS written in server-side includes

13


VIP:

do not edit these