Wanna learn LISP. Show me the way.

I heard a lot about the family of programming languages called "LISP". I also heard, that it can run just as fast as C or even faster. I love science, math and video games. And the lisp community seems to consist of like minded people. I wanted to learn steel bank common lisp and was thinking if it would be a good choice for implementing a pet game engine. It would be nice if you could share some wisdom with me. :D

>>5067 >I also heard, that it can run just as fast as C or even faster. Depends how the code is interpreted, compiled or executed. For pretty much any language. At the end your CPU is passed machine code.

If you are new to lisp maybe try getting into racket (http://racket-lang.org) instead of common lisp which is more difficult. There are many forms of lisp all with some slight differences, but most share a somewhat common setup.

>>5067 TLDR LISP-family languages are garbage, better learn haskell or erlang Only reason one should give LISP a try is to get better understanding why languages like C, c++, java are superior over LISP. LISPs have no practical usage and they do not solve any problems, they only create abstract problems(which could be easily avoided by using better language) and then solve them in most fuarrrked up way. Trough you CS career you will meet lots of people who will tell you to read SICP and learn LISP. You should probably just give it a try and play with it a bit, but nothing more. This languages family attracts people with obsession on logic, proving trivial things using induction. Don't fall for this trap.

>>5074 Learning basic concepts applicable to every language is never wrong. Just going to Java and C++ because they are popular will only spawn another generation of coders that have to open stackoverflow to code anything useful.

OP don't listen to this guy >>5074 Haskell isn't even homoiconic, there are infinite levels of metaprogramming you would be abandoning by not learning lisp. That said, it's not the best idea for a first language (but neither is haskell). Erlang is a very good language but if you don't get your head around the message passing model and OTP it will be very frustrating. It comes easier with some programming experience. Common lisp is going to be the most portable, but it's also going to have the most 1950s-era cruft (like prog, which was added to attract fortran programmers to lisp). If you have some CS experience you might think about looking into SICP (Structure and Interpretation of Computer Programs/the wizard book). I haven't read it personally, but supposedly it has very good exercises. It was the textbook for CS 6.001 at MIT for some time. It's written in scheme, which is also a lisp, with a few more modern shortcuts and slightly less cruft. It doesn't really matter what lisp you learn, the main concepts are very transferrable. ecl (embeddable common lisp) and elisp (emacs lisp) are also popular. If you do decide to learn lisp, don't stop until you have your head completely wrapped around macros and how they work. It will make you a better programmer no matter what language you move onto next. Last note: avoid Racket. It's not a real lisp (not homoiconic), and its hygenic macros are heresy.

Don't mind me, but I see a few fellow LISP and schemers were around.

>>5074 You sound like a babby that just started his first Haskell tutorial. You'll end up with more nuanced opinions when you mature a bit.

>>5075 > Learning basic concepts applicable to every language is never wrong. Just what I've said > Just going to Java and C++... I haven't said that, but I agree with you >>5077 What is so good about language being homoiconic? I know that my c++ programs are transformen into AST under the hood and then transformed into the ternary code and then after optimizations to machine language. So what? Are you trying to tell me that I should leave all the great tools that makes my work simpler and faster and write programs in AST-like flavor? Is there any cons of such behavior? >It was the textbook for CS 6.001 at MIT for some time. AFAIR it is no longer a textbook for this class. And If that's the case I bet that you know why. > It will make you a better programmer no matter what language you move onto next. Why? I would suggest OP to better use his time, you can for example read Dragon Book (or any other topic which interests you) >>5083 You sound like frustrated CS looser who can't think for himself, I have never stated that Haskell is great. All I have said is that it is better than LISP

>>5087 if the language is homoiconic you can write your own code to modify the ast in that language. If you want to modify the AST in a language like c++, if you want to do something not supported by the language you have to start writing a gcc plugin. In lisp there are no limitations

>>5074 >This languages family attracts people with obsession on logic, proving trivial things using induction. Don't fall for this trap. Well, I have nothing against proves by induction. It makes me feel comfortable about my code. I always try to make sure that my program won't be able to do something stupid. >>5077 >If you have some CS experience you might think about looking into SICP (Structure and Interpretation of Computer Programs/the wizard book). I haven't read it personally, but supposedly it has very good exercises. It was the textbook for CS 6.001 at MIT for some time. I'm comfortable with C and Python and proficient in Java. But they all seem like a hassle. (and java is so fuarrrking verbose jesus christ.) Also some very basics of emacs-lisp. xD I will work through that book on the side, now that it was mentioned yet again. I WILL CREATE A GLORIOUS 2D ENGINE. (probably not) P.S. sorry for posting this on /sec. It was an accident.

>>5067 >I also heard, that it can run just as fast as C or even faster. Not really. For the past few decades vendors have been optimizing their chips for running imperative style C code. Alternative languages and styles of programming are not fighting on an even playing field. >I love science, math and video games. And the lisp community seems to consist of like minded people. This doesn't have anything to do with lisp. Lisp was just a popular language in the 1980s. >I wanted to learn steel bank common lisp and was thinking if it would be a good choice for implementing a pet game engine. sbcl is just a compiler. Common Lisp is the language. Land of Lisp is an introductory book which deals with game programming if you are interested. As others said though, learning common lisp in 2018 is not a particularly worthwhile endeavor. Common lisp is a dead language. Nobody uses it. No libraries. No projects to get involved with. No jobs. There is not a lot of reason to learn it besides emotional/hipster/attention-seeking reasons. >>5077 >Haskell isn't even homoiconic, Most languages aren't. You probably haven't even heard of the non-lisp homoiconic languages so it's really not a valid argument against Haskell. All it means is the code you write it the AST itself so compilers/interpreters are simpler and it's easier to modify your code with code (metaprogramming) which is neat but it's hardly worth building a whole programming language with crippled syntax around this one idea. In the 80s people thought that this might be the key to artificial intelligence. Boy were they wrong(look up AI Winter). >>5092 >If you want to modify the AST in a language like c++, if you want to do something not supported by the language This is disingenuous because modifying the AST is not an end in itself, it's just one metaprogramming technique. C++ offers other features to achieve similar results such as templates and the preprocessor.

>>5098 >Not really. For the past few decades vendors have been optimizing their chips for running imperative style C code. Alternative languages and styles of programming are not fighting on an even playing field. Don't trust this hack. Compilers are what is being optimized. Machine code did not much much. There are only a few new instructions like SSE4.x aka Streaming SIMD Extensions, which is Single Instruction, Multiple Data to run similar calculations parallel. This guy can't possibly a CS student, else he would be aware this literally first semester knowledge.

>>5098 >Common lisp is a dead language. Nobody uses it. No libraries. No projects to get involved with. No jobs. There is not a lot of reason to learn it besides emotional/hipster/attention-seeking reasons. Pretty much everything in this is a lie. I'm guessing you just made a snap judgment about CL and never bothered to look too deep. There are plenty of libraries. Funny how you say no one uses it when someone just wrote an article about how CL is becoming popular again thanks to the AI/ML boom, just the other day... Lisp has been around for 60 years and yet is still around. That has to mean something. Just because something is popular at the moment doesn't mean everything else becomes irrelevant. If that were the case then Haskell shouldn't be suggested either. But, I am kind of annoyed with all these threads getting turned into, "X language is better threads". They are languages. It makes as much sense to debate about why someone should or shouldn't learn Latin. If they are interested in it, it shouldn't matter what your opinion on it is. OP, use whatever language you want. You can create a game engine in a lot of different languages. Go look at game engines already written in the language you want to learn, read some tutorials for your language, and have some fun with it. Here are some written in CL: https://github.com/CodyReichert/awesome-cl#game-development

>>5102 >lisp is not dead Name 1 user facing application written in Common Lisp. No Googling. Just 1. Go. >If that were the case then Haskell shouldn't be suggested either. Absolutely not. Haskell is very much alive and kicking. Besides being actively developed, having a vast online community and presence, being used everywhere throughout academia. Besides all that, I can go to almost any bank and get a job writing Haskell. Name 1 company hiring Common Lisp. I'll let you Google that one. >Here are some written in CL: >Game Development >Xelf Toy and unmaintained. >cl-mpg123 Wrapper around a C library. >Trial Fair enough. Same author as the one above though. >trivial-gamekit Another toy. That's it. >Graphics Most of these are just wrappers around C libraries and the ones that aren't haven't seen a commit in years. I apologize if you took my works too literally. I meant it more like you will have a hard time finding libraries that are actually useful and actively maintained. >I'm guessing you just made a snap judgment about CL and never bothered to look too deep. I looked into it quite a lot when I had my functional programming/lisp phase a few years ago. I ended up writing a fair bit of Racket and Clojure which are much more alive than Common Lisp but still struggling for relevance. My intention is not to soykaf on your precious language. I'm just trying to be realistic.

>>5101 No I'm not a CS student, I graduated years ago. Presumably you are? Fine. Tell me what is the fundamental data structure lisp is based on. Tell me what the algorithmic complexity of traversing that structure is. Which memory segment will it be stored in? What is a stack frame again? Now go and look up what a lisp machine was. Hopefully this will help you understand why modern hardware is biased towards C. Don't assume everyone who disagrees with you is an idiot and good luck with your studies kid.

>>5104 >do this >do that Unlike a complete tool of a person my mental parser doesn't need to do what your input says. So you don't assume I need to understand anything from a person lacking basic CS knowledge. Amen.

>>5101 ever heard about out of order execution, speculation and all the other stuff which have happened under the processors hood after 2000? >>5104 lisp machines are very interesting >>5105 oh yeaaah, classical lisp acolyte buffoon. Have you secured you scholarship for doing all (and nothing more) that your professors are telling you to do? Do you have your grades good enough? Have your for at least one fuarrrking time created any useful software?

>>5103 Realistic about what? OP wants to do a hobby project. This was never a discussion about which language is more relevant to mainstream use. I wouldn't recommend lisp or Haskell for that. I'm trying to help OP accomplish the goal they set out. What reason is there in trying to prevent someone from accomplishing something they set out to do? We should be supporting others instead of feeding them propaganda. If OP wanted to create a machine learning library in brainfuck I would still try to point them in the right direction in order to succeed.

>>5103 CL is going through something of a renaissance at the moment, there's quite a lot of community activity and even a few jobs going around. CLIM is being revived, there are graphics engines now, and that grammar analyser thing you see advertised around the place is written in CL, for example. I don't know if it's still the case, but CL was the language of choice for quantum computing research a few years back too. The fact of the matter is that Lisps have been around for over 60 years, yet people are still interested in using them, they still have some of the best literature available, and so they're still worth learning.

>>5105 >Unlike a complete tool of a person my mental parser doesn't need to do what your input says. If you cannot answer any of these questions then it's not surprising you don't understand what >>5098 is talking about. I would strongly encourage you to educate yourself. The path has literally been laid out for you. >>5107 >Realistic about what? Realistic about the state of Common Lisp in 2018. >What reason is there in trying to prevent someone from accomplishing something they set out to do? The purpose is to warn OP about the reality of Lisp because he seems to think it is something it's not. >We should be supporting others instead of feeding them propaganda Like I said I am trying to be truthful. Pretending Common Lisp is still alive in 2018 is the propaganda. >If OP wanted to create a machine learning library in brainfuck I would still try to point them in the right direction in order to succeed. Absolutely, but you wouldn't indulge OP's illusions about bf being faster than C or having an active community or other such nonesense going around in this thread. >>5108 >CL is going through something of a renaissance at the moment, there's quite a lot of community activity and even a few jobs going around. I don't doubt there hipsters and gray beards still on comp.lang.lisp but what evidence do you have that the community is growing? >The fact of the matter is that Lisps have been around for over 60 years, And yet you can't name a single company who is running common lisp in production.

OP here. I love this place. It's the perfect mix of information, jokes, irony and a bit of toxicity. :^D I want to learn common lisp, because the concepts are intriguing. I don't want to write c, because it slows down my thinking process. >Python is cool and all, but IT"S JUST A WRAPPER AROUND C. <irony>Python is useless, because it's just a wrapper around C. Just program c.</irony> I'm here to learn. I'm not studying computer science for the money or because I want get hired by google. I'm learning lisp, because it has a completely different paradigm of thinking. If something sounds interesting to you, you should explore, or you will never know. It could turn out to be the most awesome thing you ever discovered. Only because something isn't popular, there's no reason to not to start learning it. If you don't like it you can stop at any point. Regarding: >I also heard, that it can run just as fast as C or even faster. I just wanted to know what lisp compiler creates the "best" bytecode. (directly compiling it to machine code would probably not allow for runtime changes) Have a wonderful evening everyone. ^^

>>5129 If you want to learn lisp then Racket/Scheme or Clojure would be much better choices because they actually have active communities and a bit of industry traction. Realm of Racket is the Racket equivalent of Land of Lisp I mentioned earlier. I think Little Schemer is an essential read for any Lisp developer, even if you don't care about Scheme specifically. The Joy of Clojure is an excellent read if you want to go down that route. Again, I can't in good conscious recommend Common Lisp.

>>5129 ><irony>Python is useless, because it's just a wrapper around C. Just program c.</irony> Python supports some functional programming language features that c does not, therefor it is closer to lisp than c. Modern lisp languages like typed racket for example are multi paradigm so you can use them like other languages too. The lisp and the c world that once were very different have slightly merged. No less i think a true lisp is worth learning.

>>5129 >>5074 anti LISP flamer here I just have to react when someone is persuading that LISP is the best thing ever created (which have happened in this thread). If you want to learn lisp then start with Racket, it has lots of cool features, has nice documentation and was intended to be used by beginners. Sorry if I was too toxic, but I fuarrrking hate lisp-mania. I hope you will find your "most awesome thing you ever discovered." Peace

>>5130 This pdf was the main reason I caught interest in common lisp.

This thread has not at all gone in a security direction. There is already a Lisp thread in /lambda/: >>>/lambda/4793 This thread is locked.