Ask HN: What should I learn as my first programming language? | Hacker News Reader

Ask HN: What should I learn as my first programming language?

26 pointsFalconNine93 comments

If you have absolutely no experience programming, start with Python. There are three things to learn when programming: 1) language syntax, 2) concepts (conditionals, logic gates, loops, etc.), and 3) how to solve problems.

What makes python a good choice is that you don't have to spend too much overhead learning language syntax in order to implement concepts, and you'll easily be able to keep the learning of those two items separate.

Once you have a strong foundation in concepts, you can much more easily pick up a different language. The biggest downfall there is that if you decide to later learn Ruby or JavaScript, you might approach a solution as, "This is how I would do it in Python." Don't fall into that trap. Learn the features each language provides and how to use them.

Since you mentioned games, there's a couple of books on the topic that are free: http://inventwithpython.com/

I always recommend Lua as a first language. It's about as straightforward to learn as Python, but it's smaller and simpler, with fewer "gotchas". Once you've learned Lua, moving to Python is about as easy as moving to Javascript.

I don't recommend Python or Javascript because while having a rich set of third party modules is great for actual development, it's a detriment and distraction for beginners. When you're learning the nuts and bolts of programming, it's easy to be misled by how easy some common task is to perform via API call and think you understand it when you haven't really been exposed to it. This becomes a big problem when moving to some other language, where library functions work in a different but equally straightforward way. When it comes to things outside the core standard library, you should really learn to implement them yourself before becoming accustomed to using third party libraries.

Python is a great second language, but the way you write things for a beginner programmer is very different from the way you write things when you're trying to be clever and squeeze a lot of behavior out of a few characters or a few cycles, and python has too much of a mixture: searching for information on how to solve a problem in python will get you equal numbers of people who don't know how a for loop works and people who prefer to do everything by passing lambdas around.

Lua is not a bad choice: the docs are good and once you get the basics down https://love2d.org/ is great for writing Tetris and having having something you can point to and go "I did that"

I don't know that hiding complex ideas is necessarily helpful for a beginner. Admitting "by the way we're writing 'public static void' every time for a reason, but we won't worry about it yet" or "yes technically you could complete this assignment on one line, but trying to learn that right now is not useful" don't seem like bad ways of confronting the idea that programmers don't always know everything and that's okay. Keeping them completely insulated from those kinds of things could make them seem scarier than they really are.

Of course learning is complex and the "best" way totally depends on who is learning and their goals.

I disagree with some of these points. Specifically, programmers should necessarily be left unable to understand everything they use. Packaging systems are excellent for providing you with functionality without the mental cost of having to understand how that functionality works. They're not a distraction, they're how people get things done.

And Lua is not a great first language. If you know how to use it, it's a great language; but its 1-indexing, tiny standard library, inability to distinguish between hashmaps and arrays, confusing "metatables", and lack of external libraries really make it tough for me to recommend as a first lang.

This comes from someone who used LuaJIT often with Love2D to prototype games.

Anything that's trivial to write (think "left pad" and 90% of what's in NPM & PIP) shouldn't be used by people who can't conceive of how those things might be written, if their goal is to learn to write things. I'm not suggesting that beginners should start with writing bootloaders in machine code or anything; I'm suggesting that beginners should master things like loops and functions and variables before they start plugging together big chunks of other people's code, lest they overestimate their abilities and take a major blow to their self-esteem when they realize that they are completely incapable of doing simple things in another environment because they never really learned how to do things. Some of the biggest and most important hurdles to clear when learning to code are related to learning to master things you've never considered; convenient API calls are convenient to avoid considering. In this way, a tiny standard library is useful: all of the things in Lua's standard library are things that a beginner can afford to put off learning for a little while.

Lua has a handful of strange attributes that distinguish it from most programming languages -- 1-indexing, lack of a distinction between integer and fractional numeric types, lack of a distinction between integer-keyed and non-integer-keyed arrays (and lack of distinction between arrays and objects), and the use of tilde rather than exclamation point in comparison. The number of idiosyncracies is comparable to those in python and javascript. I don't think that distinguishing between arrays and objects is useful for a beginner, nor do I think that the fact that in most languages integer-keyed arrays are special for optimization reasons is terribly important for a beginner. And, metatables are an advanced feature -- they are similar to mechanisms in python and C++ and aren't all that weird, but they won't even come up until long after our hypothetical student is ready for their second and third language.

A first language is for learning. Whether or not it's useful for real work is totally irrelevant; someone who doesn't know ten languages inside-out isn't qualified for real work in any of them. Lua is a good candidate because although it can be used for real work, it's not optimized for that. Instead, it's a simple language that combines ideas that have historically been separated for reasons related to ease-of-implementation and separates ideas that have historically been combined to make ugly hacks possible. In other words, it's a fairly internally coherent and conceptually pure language that is easy to pick up because it lacks "gotchas" (except to people who come from other languages, and are offended by Lua's improvements).

90% of what's in npm or pip is not trivial to write. Calling database adapters or HTTP libraries "distractions" is silly. And in the same sense, restricting yourself to a subset of a language that makes it easily ported and contained within any application seems... silly.

Furthermore, plugging together pieces of other people's code can very much help in learning how to express concepts in the language. You can look at existing examples of how it's done, and do it yourself. You make small changes, then big changes, then write your own. Easy peasy.

Lua is a weird language for many reasons. Not that being weird is a bad thing, but that makes it difficult for people to learn other languages. If you learn a C-style language, you'll be more able to read other C-style languages.

And no, metatables are integral to Lua. You can't use prototypical inheritance without them, and they're really difficult to wrap your mind around. Furthermore, prototypical inheritance is very useful. If you've actually learned Lua, you've got to learn about that aspect of metatables.

I love Lua for real work. Honestly, actually do. But I don't love Lua for learning.

You need metatables for rolling your own OO system. But, object orientation (particularly when, as in Lua, you need to roll your own) should not be part of a beginner curriculum. Like I said, by the time you need to worry about metatables, you should have already moved on to your second and third languages -- you can learn about OO in the context of Python or JavaScript and then later apply that to Lua.

Don't learn a programming language. Learn how to accomplish projects.

Pick a project, whether it's a blog or an automated tea brewer or 3D model viewer or something... and then pick a language that suits your needs.

So many beginners fall into this trap of learning just for the sake of learning; I've found that people with projects and with clear goals make the most progress.

If you want to learn a programming language, read any of the other comments. (I agree with Python being first on the list.) But if you want to learn to program, don't fit yourself into a category just yet. Be pragmatic and use whatever's best for the job.

This is really great advice that jibes with my own experience. A small thing I would add is that when you do this, the following main goals are possible:

1) To finish the project.

2) To learn a language.

3) To finish the project, but also learn a language.

4) To learn a language, but also finish the project.

My $0.02 is that: 1) tends to lead to short term hacks; 2) and 4) tend to lead to dwindling motivation; 3) is ideal. And if it's your first language, do the project while you read a good generic programming book like 'code complete'.

C, on a Unix system like Linux or BSD. It'll make you learn how your computer works, which is an essential skill to have and will help you understand everything else you learn.

That's how it is done in universities. That's how I started. My professor even did not allow compilers which can specify errors. We were supposed to find out errors on our own. And you know, I had to learn vim first to be able to write anything. He just wanted to make it impossible to have any aid. The recommended book was by Ritchie and Kernighan.

http://www.ime.usp.br/~pf/Kernighan-Ritchie/C-Programming-Eb...

I enjoy those days and happy that I scored A in his class. It has given me confidence for sure.

I learned C when I was 11, an uncle gave me a book as dense as a dictionary covering all the aspects of the language, as well as algorithms. He called it "The Bible".

For 1 year I woke up at 5:00am every saturday and sunday, to use the only computer we had at home before other people were up so I could do every single exercise of the book (there was about 20 per chapter, 10 first trivial, and the rest were small projects of increasing difficulty).

I consider that this experience is what made me the best programming student even at college, and an excellent developer when in the workplace. The lessons I learned from C are still priceless even though I mainly develop in managed and OOP languages today.

This will teach you basics that will never fade, there're certain Google searches you will never make again.

The important thing when you start is to know how a computer works, and how lines of code are interpreted and executed by the computer, as well as how to express and solve real world problems with computer instructions. You don't need to know every detail, but the less magic it feels, the better.

It might be worth pointing out that your ability to do this (learn C from a book at 11 years old) is at least uncommon, if not extremely rare. This may have worked very well for you but is likely to be difficult, discouraging, or down right impossible for most other people.

Forgot the most important part :

>The important thing when you start is to know how a computer works, and how lines of code are interpreted and executed by the computer, as well as how to express and solve real world problems with computer instructions.

And C is both low-level enough and high-level enough to allow you to do all of this at the same time.

Someone show Ruby some love:)
Learn Ruby.
Its a beautiful language well worth the effort and it has a HUGE ecosystem behind it. You'll never lack for resources. It is powerful and I found it doable to learn it all on my own.
It will aid in the transition to making dynamic websites because it is used in Rails.
In the early part of my learning I found that the break in compatibility between python 2 and 3 was irritating and confusing to me. Some of the good books were on python 2 and I had been informed the world was moving on to 3. Ruby has major versions (1.9 vs 2) but no break.(I started out by trying to learn python but switched midway to ruby)
The only caveat is if you want to do math heavy stuff or machine learning - python excels at that.
Python is frequently recommended by people with a CS / computational background. I did not have that.
Some people do not approve and I understand. But this was my experience. YMMV.

Hmm. I kind of want to go for Ruby...but how would it do if I wanted to make simple games? Nothing too fancy but simple pixel games or clicker types.

> ....but how would it do if I wanted to make simple games?
Yes ruby can be used for games. https://www.sitepoint.com/is-ruby-good-for-game-development/

Ruby is a terrible first language. Too many ways to do things and looking at other peoples example code will only confuse people new to programming. Also too easy to fall into bad habits.

Imagine you're an alien who just arrived to Earth. You are trying to decide what language to learn.
One hipster tells you "You need to learn Esperanto! That's totally cross platform and the most ideal language ever!"
Another hipster tells you "Learn latin! That's the most legit language ever. All other western languages are based on it!"
But all you want to do is live in the US and make some friends. Will you go learn Esperanto or Latin, instead of learning English? Don't listen to these hipsters.
I agree with another guy on this thread saying don't learn a language for the sake of learning it. Because you will probably become a mediocre dabbler that way. Only learn if you want to accomplish something because only by using it in a practical context you become proficient at it. Otherwise you can maybe tell your friends "Hey I know Esperanto!" but does that really matter? There are thousands of other more valuable things you can do with your time than learning Esperanto, especially if you're looking for the "first language to learn"

Java has no Shakespeare.

Python is a good choice. I would argue that C, followed by C++, might have more applicability later on; a lot of languages are modeled on one of those two. They also are lower level and so give you a more in-depth understanding of how memory allocation works.

There are a lot of great "visual" programming tools you can play with, often right in your browser. Typically designed for kids (Aged 6 to 106)!
MIT Scratch: Create shareable stories, games and animations
https://scratch.mit.edu/
Minecraft: Hour of Code
https://code.org/mc
Processing: flexible software sketchbook for learning how to code within the context of the visual arts
https://processing.org/
Google Blocky: the library for building visual programming editors
https://developers.google.com/blockly/

+1 for Processing

Mods, please add the "Ask HN" prefix to this one. I thought this was an article on some other site.

Julia. http://julialang.org
It's got everything you need to succeed in the future.

Javascript.
It's simple enough, completely ubiquitous and has an incredible amount of online resources. To start, just open developer tools in your favorite browser and type console.log("Hello, World!");. Congratulations! You've just written your first computer "program". Now go try one of the thousands of great javascript tutorials for absolute beginners.
Once you learn the basics you're now able to code anything from a basic website to an embedded machine. Your code will work on any operating system and can be distributed on a platform (the web) that makes it trivial to show off your work to your friends.
But even if you don't choose Javascript, please avoid some people's suggestions of learning systems languages like C, C++, Rust or a paradigm-specific language like Java, C#, Haskell. Yes, these are all fine languages in their own right, but you'll spend a month writing boilerplate you don't understand, which is both confusing and unhelpful. If you have any friends that can program and are willing to help you, following what they know may also be a good idea.

* Want to make webpages ? Learn JavaScript, HTML and CSS.
* Android apps ? Learn Java.
* iOS apps ? Learn Swift.
* No idea what you want to do, but strong desire to learn to code ? Learn Python. You could use Python for general purpose programming, small games and so on ...

How much easier will new languages be after I've learned Python, assuming that's what I go with?

Once you learn a programming language, the second one is usually much easier to learn.

1. Learn an "easy" general purpose scripting language (JavaScript because you can just type F12 in your browser now and start writing code, or Python if you want something just as simple but more "logical" and "sane", Lua ok too) until you feel you "grok" the basics: once you can write a simple text-based questions and answers game that uses a loop, and you know how to write a simple function that returns another function (called "higher order function" but do not be scared), it means you "grok the basics.
2. Pic the platform/domain of your first project and use the easiest language that is still a first class citizen of that platform. Easy guide:
- iOS -> Swift (don't pick Objective-C if you haven't touched C or C++ before) - Android -> Java (don't pick JavaScript first, unleast you have serious *web development* experience, the zillions of all-slightly-wrong options and choosing a framework madness will make your brain explode even if the language is simpler!) - web browser -> JavaScript - web server -> anything *except* Nodejs (JavaScript) or PHP, you don't want to start by diving head first in callback hell and a zillion ways to do everything and you also don't want to learn all horrible worst practices accepted in PHP-land that you'll have to unlearn later: *Python* or *Ruby* or *Go* are ok here but avoid large frameworks (choose something like Flask for Python or like Sinatra for Ruby first) - linux -> C - desktop/console game using Unreal engine -> C++ & UnrealScript - dektop/console/mobile game using Unity engine -> JavaScript and C# are both ok here - windows desktop -> C# or JavaScript depending on your purpose - macos desktop -> start with Swift, got to Objective-C if you can't get to your goal using Swift - hardware hacking -> "Arduino C" as suggested by someone else or Python if the platform you're hacking supports it - machine learning / AI -> Python

I like very much http://www.codeskulptor.org/ because of the small entry barrier: easy syntax, small ecosystem, no installation, good documentation, possibility to write small games.
If you are young, you can also motivate friends, emulation, ...
And for you second language you will probably know better why you want to leave codeskulptor: - full python to have bigger ecosystem - java if you want an IDE (eclipse or netbean) - javascript, html, css if you want to build a "startup" - haskell or scala if language is your interest - C, unix if system interests you - C++ if you are masochist - Excell VBA - ...

https://www.pyret.org then Haskell.
The other advice might allow you to check off marketable skills faster, but these will make you a better programmer.

How to Design Programs: http://www.ccs.neu.edu/home/matthias/HtDP2e/

This looks interesting. Which programming language is this geared towards? (PS: am not a programmer).
Edit: never mind, just read the preface.

I would say Python - it's a great general purpose language of which many interesting things can be built.

The language doesn't really matter much at the end of the day. I think I started with "high level" COSMAC VIP assembly (and I'm not that old... I needed something that would drive 5V electronics.)
First pick a project. Most people pick video games, or light electronics (simple robot, nest replacement, etc). Then, pick something with a strong community and lots of documentation geared at newbies. I strongly recommend a Raspberry Pi (but buy the ~$60 official kit with power supply and SD card -- they have crazy compatibility issues with SD cards and USB chargers). The Pi comes preloaded with beginner-oriented things like PyGame, a computer music language, and software to drive cheap breadboards for electronics projects. There are tons of tutorials online.
Alternatively, maybe download Unity, and watch the YouTube 2D game tutorials (the open source suggesions in other replies are also fine, and most of the language suggestions make sense for some project or another).
By all means, if you have an experienced programmer that is willing to help you, tell him/her about your project idea, and ask for help picking a language and setting up a development environment. A suboptimal language choice will be more than made up for by in-person tech support.

Like many others I don´t consider HTML and CSS programming. To some PHP isen´t even a programming language it´s a scripting language. But I ended up programming/scripting PHP. I like it, I feel at home in it, and it does what I want to accomplish. Make stuff for the web. 14 years ago I started learning HTML, I wanted to make websites, after that I had delusions thanks to movies like Operation Takedown, Hackers and Antitrust about being a hacker. I learned the basic concepts in C++, C# and Java, but I diden´t know what I wanted to do, so I froze, and stopped learning in those languages. There is a term PHP is forgiving and it is, some of what is natural to PHP might bite you in the ass if you go on to C# for instance or Java. But you´ve still got a bunch of programming concepts down that you can reuse. If you can find a drive, something you want to accomplish in the language, be it a web or desktop project. Then you will learn. I hope you find this useful. Whatever you decide to do good luck on your journey to becoming a programmer.

Honestly, Follow any of the mainstream languages you will be alright.
When I began programming I had this exact question myself, and deep down the actual question was "what language is the easiest to learn and is there a language that will not make me quit because its too frustrating?".
I think becoming a programmer means you have to be able to handle a lot of disappointment, and you should be able to deal with the fact that you have to constantly learn (you asking this question on HN means you are going on the right path, it means have that explorer inside you). You will have to deal with frustrations and to do something over and over again until you have perfected it. (this is why I think a lot of us see programming as an art)
So learn any programming language, build the smallest thing you can imagine, start with a hello world, make a calculator, make a text based game, etc etc and keep going until you are fed up of programming and want to take up farming instead.
And one more thing, please don't think memorizing syntax is going to make you a great programmer, great programmers program, they make things and they make more things. Make things and never stop.
All the best.

I agree with this completely. I started with assembler, moved on to C, Lisp, Fortran, BASIC, Pascal, and C++, and then many more from there. It never mattered. Sure, I sometimes used the wrong tool for the job and made life harder, but those experiences shaped me.
You can make mistakes and you will learn from mistakes. You can build a house out of cardboard and it might not be a good house, but if it provides you shelter and you call it a house, it is a house. Don't get caught up in the cult of programming language. The good thing about this cult is when you meet people who say, "I am an <insert language> guy," you can turn on your skepticism detector to an extreme level or pretty much ignore them by default because they are either stupid, inexperienced, naive, or living in some bubble that probably won't be the same for you. The tools analogy really does work rather well - you can pound in a nail with a set of plyers, but hey, why would you and why wouldn't you invest in buying more tools when you can afford it?
Getting things done, failing, learning, and problem solving are what you want to be doing among many other things that have little to do with language choice. Don't be too ambitious, just stick with it and get things done.
While I wish more people were better at programming, the truth is you don't even have to be good to succeed on some level. I wish it weren't true, but if you saw the source to most games, you'd cry. When you go up against time, budgets, other people, technical limitations, and more, the end product suffers, but to most people, as long as it succeeds doing whatever they deem valuable, then you won.
Parent really gives good advice to start small. You want your gratification loop to be as constant as you can make it. Trying to build Minecraft, Skyrim, Call of Duty, whatever as your first game might be an option for a few people, but for most it isn't. Don't try to do it, just get things done. If after awhile you see that maybe you can't seem to learn other languages, which does happen with a lot of people, I'd say that maybe it's time to look elsewhere for a career.

I think it depends on you, to some extent, so without knowing you it's harder to say. The big question is "why?".
1) Are you learning to program because you have something specific you're dying to build? Depending on what it is you dream of building, there may be a language that's more appropriate than others. If you want to build a mobile app, you pretty much have to learn either swift (iOS) or Java (android). If you want to build a web app you could reasonably choose Ruby, Python, Javascript, Elixir, or probably a dozen others I'm forgetting. If you have a lot of data you want to analyze, you might want to learn R. If you want to write games, then Lua is a great choice. etc.
2) But maybe you don't have a project in mind, you're just interested in how it all works. In that case, you should know that it works a lot of different ways, and different languages will teach you different things. Learning C will give you a great understanding of how computers actually work at a low level. Learning Ruby will teach you a lot about object oriented programming, which is currently the most popular paradigm. Learning Haskell, Elm, Elixir, or Erlang will teach you about functional programming, which is quite a different paradigm but one that's getting a lot more popular lately. If this is what's driving you then I'd say you should learn 2-3 languages that are sufficiently different: something low level, something high level and object oriented, something functional.
3) But maybe you want to learn to program because you hope to get a well paying job asap. I'm not sure that's such a great idea, but if that's the plan, then I suggest something where there are a lot of jobs right now. Javascript, Ruby, or Swift, most likely. Learn one of them sufficiently to be useful, and get yourself in a position where you're getting paid to do this. Then, if you want to have a happier life, follow the advice in section 2.

4) If you're looking for a career as a programmer, I'd search jobs in your area to see what most companies are hiring for and focus on learning that language.

HTML, CSS and JavaScript was what I first learnt almost 20 years ago and I still think it is a great path to beginners

I don't mean to sound negative - but the HTML stack is the worst thing on earth to start leaning to program in.
HTML, CSS are not 'programming', and both they and javascript are a convoluted, dysfunctional mess.
Yes - they are easy - but they are all super quirky and those quirks form bad habits.
Python is a great place to learn basic algorithms and syntax. Java is a great way to learn types and basic OO. C is a good place to learn how to deal with lower-level things, but it can be challenging for beginners.
Javascript should be a great place for beginners - but it's full of so many oddities, that I think starters will be misled. Every JS environment is different, and it's hard to tell what works where. It's also managed by consensus and varies a lot.

Yep, I don't disagree with you. I gave a practical approach, which worked pretty well for me, I think. I also think that the native web stack (HTML, CSS and javascript) will keep growing until it becomes ubiquitous. But as it is a personal bet, I don't always use it as an argument.

'web stack' is a natural entry point for newcomers because 'they can do stuff', i.e. make web-pages that do things they can touch and feel. So it makes sense on some levels ...
But for programming proper ... no way. I wish that were not the case.

You'll learn the most if you have projects you want to do, and different platforms are closely tied to particular programming languages. Choose the platform you are most interested in, and try to be conscious of what you are missing - whatever language you choose.
Cross Platform Games - Mono/C# w/ Unity
Web Apps - JavaScript (HTML/CSS too)
iOS Apps - Swift
Android Apps - Java
Scientific computing/data analysis - Python or R
Robotics - C++, though Python and Node (server side JavaScript) frameworks exist w/ lower barriers to entry
Once you learn one and have built a few things with it, try to learn another that takes a different approach. For example learn a compiled language after Python, or an interpreted language after C++. Try out a functional language like Haskell/Clojure/Elixir. You'll start to see how the language you chose helps and constrains your programming, and where it falls short or offers new hurdles.

As you can see there are many answers provided, and the reasons all seem to conflict with each other. I hope you are able to disregard any advice that blindly suggests starting with language X, without knowing anything about your background or interests. Currently the 1st and 2nd highest voted answer are C and Lua, which provide completely opposite reasons for each.
There is no unanimous cry of a single language "JAVA! Definitely start with Java!" because there is no right answer. There are more similarities between every programming language than we tend to admit, and the right answer is the one that will keep your interest and solve your needs.
Do a little research on what programming language is often used to work on topics your interested in, but don't sweat it too much. Pick one and start making mistakes

C# of course, easy and nice!

My wife started some time ago to slowly learn python as her first programming language. REPL is very handy. But I am not sure that it was a good choice. She will not see results visually. Maybe better choice would be JS with HTML and CSS?
I'm mainly C and C++ developer and I worry about dynamic nature of python. I am not sure that seeing failure in run-time is all that helpful. But that's just static vs dynamic typing dispute.
What I really have to do is to sit with her and talk. What would she like to achieve? Is she bottom-up or top-down learner?
I was also thinking about awk. It is small and nice language with immediate results. Learning it requires learning regular expressions and I wonder is it a right time.

Depends on what you want to do and what you got.
If you have a Mac or an iPad I would recommend learning Swift. Playgrounds are a very neat way to start hacking and seeing results immediately and you will end up learning a language that is very modern and has a good future.
If you want to learn not only programming but also how computers work (and are really motivated) then C is a good starter as well. It requires more commitment but you will also learn how computer memory works.
Finally if you want a simple language I can recommend Scheme. This was the first language we learned at engineering school and although I have had quite some experience with programming before going there I found it very refreshing.

If you want a language somewhat close to math, take OCaml. If you have an even more mathematical mind, take Haskell. If not so much, choose a language that has static typing, so you learn what a type is right from the start. This will save you many headaches later on, especially in languages with dynamic typing. If you're going with C, because you want to know how software works from the ground up, fasten your seat belt and wear a helmet. And don't panic, the worst that can happen is your screen exploding (aside from your head every once in a while).

I've long thought that learning an imperative language first hindered my comfort level with FP. I felt like learning a functional language first would prepare me better for learning anything else.
So, maybe Elm or Racket?

It really depends on your ambitions. I like the idea of lua as a simple/pure langauge; the problem is mainly the support and reception of it doesn't lend itself to growth so much; since lua is mostly used, in practice, as part of systems more complex than itself.
I am grunching this thread, but I think the obvious choices are python or javascript; unless you know exactly what it is you want to build, and it is an 'App'; in which case you might as well start just learning the native language necessary to build on your desired platform.

What do you want to do? If you want to do stuff with the web then learning JS is obviously a necessity. Anything else is pretty much open to your choice. Like everyone else has said focus on the goal of what you want to do first since major languages are supported for a wide variety of use cases ranging from PCs to embedded electronics to even robotics. Once you tackle the most basic thing you want to do the rest comes easy or easier (depends on how complex the thing you want to do).

your comments say you want to learn to make small games as a hobby. Look for resources that teach you how to make small games and don't worry about learning a programming language.
You will learn programming as you go, and gradually find a language or framework you like.
If you want want to make games, but you pick up Python 101, you may get bored and move on, because its not what you are interested in. I recommend physical books for beginners. Look for books like "Programming [X] Games for Beginners."

I've been teaching programming to someone with zero computer experience. This person had difficulty with mutating variables in Python, so I switched to teaching them Haskell. This person has found the equational reasoning/substitution of Haskell much easier to mentally model.
I'd say about two thirds of the people I teach have no difficulties learning Python, and one third do much better learning Haskell because then variables don't vary!

I'd say LISP family via SICP book. Almost zero syntax, multiparadigm (i.e. ain't force you to do something in a strict way), lots of high level stuff, extendable. Once you'll learn Lisp - you won't care in which programming language to code at all, since it covers all of the ground.
Another variant is Rebol and Red, with them you could start tinkering with GUIs in no time, in other aspects it's similar to Lisp, Forth and Ruby.

Lots of people answering with a great first language - here's a second, third and fourth (assuming the first is an OO orientated language):
Lisp (Clojure) - Learn how to think functionally (Go deeper with Haskel)
C - Learn how to think imperatively and work with low level OS concepts (Go deeper with Rust, C++, Go, etc)
ASM - Learn how the computer actually thinks (Go deeper by creating a toy OS)
And finally, create your own language with a parser and compiler.

Techical writing (in your native tongue).
Write the instructions for creating a peanut butter & jelly sandwich. Have someone try to follow your instructions. Did you get a sandwich?
If yes, you're now a programmer.
When picking a computer programming language, most important is having a buddy / mentor who can help you climb the learning curve. If you're looking for local resources and communities, maybe start with meetup.com.

Replace "Have someone try to follow your instructions" with "Have someone who hates you follow your instructions" for more accurate results.

What do you think you will enjoy doing? Web pages? Android apps? iOS apps? Pick something fun then choose the right language.

I'm not doing this for work or school or anything, just for fun. I want to make small games as little pet projects, not aiming to make money or anything, just to have a hobby.

Learn Python. It is a general purpose language. Since you are specifically looking to start with games or for little pet projects, I would recommend:
https://inventwithpython.com/

Well, unless you really want to make a project that needs them, do not start with C, C++, Java, C#, or swift.
Those languages are complex. There's a lot to learn just to master syntax, a lot of restrictions you have to be aware, and a lot of learning material pointed at big projects that will be harmful to somebody just learning how to program.

Which language is better then? BASIC? I see a suggestion for Lua too, which seems attractive.

Python, Ruby, Lua are all great. If you want to make it serious and will go into lower level languages later, Pascal is great.

Learn about the trade-offs, you'll be more productive in C# or Java and have a more pleasant debugging experience over C, but C gives you better understanding of hardware and opens doors for embedded devices, OSes, IOS.

C# :)

python 3 and to help you visualize some concepts, use thonny http://thonny.cs.ut.ee/ which i find quite unique. Later for real projects you want to use some better IDE. Hands on learning: https://www.hackerrank.com/domains/python/py-introduction

I'd be tempted to say Smalltalk or Python.
Real BASIC was awesome, but I don't know anywhere which would have a traditional, old-fashioned BASIC. Maybe an Apple II emulator?

Erlang

> Erlang
Genuinely curious. Why Erlang?

I think a case could be made based on the relatively straightforward syntax (very few gotchas).
I mean, most functions will look like:
f(X) when is_list(X) -> firstExpression, % commas separate a sequence of expressions secondExpression; % semicolons tell us an alternate sequence follows f(X) when is_integer(X) -> thirdExpression. % periods terminate the block
Case expressions (similar to switch/case in other languages, cond in various lisps):
case Expression of Case1 when SomeGuard -> Expression1; Case2 when SomeOtherGuard -> Expression2; _ -> DefaultExpression end.
Receive expressions:
receive Case1 when SomeGuard -> Expression1; Case2 when SomeOtherGuard -> Expression2; Case 3 -> Expression3 after 1000 TimeoutExpression % handle timing out end.
Note how they're all following the same form (this is also a selling point of python, its syntax is pretty consistent, especially for the features used by beginners).
The conflation of strings with lists and characters with integers can be problematic, as sometimes a list of integers will display as a string if they fall in the printable character range (and this isn't what you want). This creates confusion, but it's surmountable.
Variables don't vary in erlang, but if a new user isn't accustomed to doing:
x = y x = x * x x = x + 4 x = sqrt(x)
then they won't have too big a problem doing:
X1 = Y, X2 = X1 * X1, x3 = X2 + 4, X4 = sqrt(X4).
It is more cumbersome, but it's also likely that those intermediate variables could have more descriptive names.
The concurrency model is fantastic, and will, IMO, be greatly beneficial to anyone new to programming. Not because it's available in every other language (without libraries it likely isn't), but because it gives a really good mental model for concurrency that can be mapped to other existent concurrency implementations in other languages. It also helps to encourage a design based on loose coupling and high cohesion, which many (most?) of us believe are good design principles.
Syntactic sugar around list comprehensions, maps, binary comprehensions make a lot of algorithms very concise (though perhaps not performant, but this is effective for a first pass implementation when you're aiming for correctness and not speed).
The library ecosystem was less comprehensive last I checked (several years ago now), which made integrating some useful libraries very challenging. But I believe this has changed, and the Elixir project has certainly helped a great deal here as well.
The performance of numerical code hasn't been good. And a lot of people mistake concurrency for parallelism expecting performance boosts that just won't happen. Concurrency is largely a design/communication principle, not a performance one (though it can and does aid performance). So those are areas for concern.
There's probably more pros and cons, but I've got to go now.

Great case you make there! I'm off to look into Erlang :)

Swift. Create mobile apps instead of websites.

Why mobile over websites? Just curious, because I'm also a newcomer that is debating web vs. mobile.

Arduino, it's C but cleaned up a little bit and you can go under the hood a little bit too.

GoLang and JavaScript are pretty important areas to know these days and into the future. Node is still big and will continue to grow on embedded devices I'd hazard. I would also recommend learning Python and if you have time swift (native apps!).

i'd try Go and Elm, although you'd encounter some JS and HTML a little bit with Elm.

You could go old school and start with Pascal. It's a great learning experience. And isolated from the real world to some extent, which at the start is a good idea (there isn't 300 frameworks staring back at you).

Agreed. Pascal has several advantages as a first language. The original Pascal (oP) language is small and quick to learn. ISO Pascal (iP) adds objects to oP, so the transition from oP to iP invites a graceful transition from procedural to object oriented thinking.
As a type safe language, Pascal also introduces concepts of data placement and memory management early on, and supports dynamic allocation and pointer access clearly and painlessly.
Learning C, C++, or Python after Pascal should be quick and easy, since they're just variants and generalizations on the primitives learned from oP and iP.

I'm going to be blunt, almost all the comments thus far suggesting a specific language are wrong. This happens everywhere, every time someone asks this question. I would instead pay much more attention to anyone speaking generally about problem solving and learning. I've learned countless programming languages and eventually you'll get to the point where learning a new one is just a matter of spending some time with it on a project. You're not married to any language ever unless you're a terrible programmer, so you can always switch to something else and at most, you've lost time.
My suggestion would be to pick a problem, analyze the problem deeper to see if it's even worth solving, then proceed from there. That alone is a huge subject, but it's really what you need to be comfortable with if your goal is to be a programmer. If you simply want to have a language or two as tools (ex: scientists), then sure, dive right in, but you will forever be stuck at a certain level if you approach things that way unless you are really lucky or really brilliant (most of us are not). Otherwise, take the time on the problem you want to solve, first and foremost, every time. The problem generally (obviously some exceptions) has little to do with language, computers, or anything technical, even if the problem is actually focused on a computing topic. If you still want to solve the problem after doing things like figuring out who/what/where/when/why, then try to pick the language(s) and tools that best fit the problem. Unfortunately it is hard to do this without having programmed in a language yet, but you can at least do some basic research, reading, and so on to find the best tool for the job.
You will fail in this your first time, and probably your first ten or even 20 times. Most programmers fail in this their whole careers, so do not get discouraged. The best you can do is learn from your mistakes when trying to solve problems with technology, and try to do better the next time. Even if you fail to pick the "best" tool for a job, making a mistake will teach you so much and you'll build on this going forward. This also means you can always change languages if you think it won't work for your next project or you just want to learn something new. Getting things done, learning patience, solving problems, critical thinking, balancing everything, and countless other non-technical things are really what it's all about and learning them will keep you going. Language choice can save you lots of time, lead to better results, and much more, but it often is just a distraction for the average problem.
You mentioned you wanted to make a game. I would say 99% of programmers I have ever met say this and very few of them have gone on to make games at all, and even less, professionally. Saying you want to program a game first is like saying your first race will be an ultra-marathon with no training. I am not trying to discourage you, rather I am trying to be real and help you think about what goals you want to set for yourself. For reference, yes, my first program was a game, but this was a long time ago and I'd say it's a miracle I continued.
I have worked on games and game engines professionally and would be happy to discuss more things related to the learning process. To avoid being "that guy" who never gives any specifics and is just a contrarian, I will tell you the honest truth as it has been for a long time in game programming -- If you want to be at all decent at game programming, you must learn C and C++, and eventually at least a decent understanding of assembler would really help you. Anyone who tells you otherwise is wasting your time, lying, and/or has never written a game (why this is would be a huge post). You can be productive and write good games in any language, but if you are anything like most of us, one day you will want to write a game that is not just a toy, text-based, or a glorified web page. There's nothing wrong with those games if that's your thing, but I've rarely met anyone who has the long-term goal of writing those games except for spammers, crazy people, or unscrupulous people (ex: Zynga). Again, that's not to say your game has to be written in C or C++ (you can write amazing games in C# using Unity for instance), but having command over those languages still teaches you so much of how a computer works, how to optimize games, and will save you when eventually you'll want to communicate with things written in those languages and not treat them like a magic black-box. For the goal of being a game programmer, it's a good use of time even if you never write anything professionally in C or C++. It's less about the mechanics of C or C++ and more about learning how things work at lower-levels and expanding the size of your "safe" zone mentally. Likewise, you'll want to learn quite a bit of math for most game programming, be in matrices, algebra, geometry, calculus, diff eq, etc. So you may not even want to learn C or C++ as your first language, and that's fine, just know that long-term if you are serious about games, you will have to do that. The message is that you need to focus on computer science skills, not language specifics in order to work towards that goal, even when working in other languages. In the process of doing this, you will learn a lot more about what I said earlier - picking the right tool for the job, which means not just languages, but libraries, editors, formats, algorithms, etc.
Building a game, at least remotely well, requires quite a lot of skills that are far beyond a new programmer's abilities and it is nearly impossible to master them all in a lifetime. You need to learn the basics first and to understand your limitations and resources - you won't be building a AAA or even a nice looking indie game for a long time unless you are truly exceptional, lucky, or have a team of people doing it mostly for you. At best, you'll produce something that on the surface works, but your programming ability will suffer and if you eventually do work with competent people, they will think you are terrible or hate you. That said, building a game will require you to learn so many areas and thus teach you more than almost anything the best way - hands on. It's a balance of getting things done and doing them right, and knowing where, when, and how to cross that line is vital. When you first start, you won't even likely have any ability to filter out all the information out there that is wrong, will waste your time, is bad design/BS/awful. Some programmers never learn this, but this too is hugely important.
I strongly advise you to see if programming is something you even want to do first, figure out what you like to do, where are your deficiencies, and work from there to figure out what your role would be in game dev, where you need help, how to judge others, etc. A game can be a motivator to learn more and teach you things along the way, but it can also drive you away, overwhelm you, teach you bad habits that you will never shake, and generally lead you astray. As such, just writing simpler things that solve real problems might get you there faster and sharpen your skills better along the way, while ensuring you don't throw up your hands. You might be different, but that's the pattern I've seen in most people. Pick a reasonable, small problem you care about, match the tools, go from there, repeat, potentially with new languages after you get comfortable programming for awhile.

pascal or haskell

P.H.P

Don't.

Don´t? Don´t what? learn programming at all?

Stop trolling the OP. How would you feel if he believed you and actually picked up PHP?

Stop trolling PHP developers. How would you feel if your darling language received this kind of review?
Honestly, PHP is a fine language to begin learning, it's very easy to get a local development server going for someone who barely knows the CLI, it's integrated tightly with HTML (for better or worse) which make things simple before you move to patterns like MVC, etc.
You can build some pretty cool, performant, beautiful software with PHP - and there are always well paying jobs out there (especially if you're not in the valley where Ruby & Python are disproportionately more popular than elsewhere.)
PHP 7 with a new, (loose) type system and a lot of the old, insecure cruft removed isn't actually a bad language, especially if your build process involves Composer and you're following the standards laid out by the PHP FIG/PSR.

Well of course I played to the negative stereotype as a joke : ) Obviously reality is not that simple, but to briefly sum up my personal take on it, yes good PHP may be good, even beatiful, but in practice it's rare to come by. One of the reasons being the prevailing culture of duct tape mentality that seems to surround it. I'm speaking from my subjective experience, YMMV.
> How would you feel if your darling language received this kind of review?
I'm not that attached to my favourite languages, but the thing is - they just don't receive it for some reason : ) It's probably correlated with why they're my favourites. I don't have overly fancy preferences, I like C# (it is getting its share of hate, but hardly ever as a language in and of itself, it's typically directed at Microsoft and that's another story) and Kotlin (sort of like Swift but for Android).

Modern PHP's not actually that bad; most of the language's reputation comes from much older (and much worse) versions.
And it is really good at providing server-side programming to a web site with the utter minimum of work --- after all, that's why it's used so much! No compilation, no deployment, not even any need to have two files open at once; just stick a couple of lines of PHP into an HTML file, hit reload, and you're there.
I wouldn't call it a particularly interesting language, but if the OP reckons they'd be motivated by building web apps, it's a perfectly reasonable choice.

It has not improved much. It has added some superficial things. But I am afraid that the worst part of the language is still very much alive.
>No compilation, no deployment, not even any need to have two files open at once; just stick a couple of lines of PHP into an HTML file, hit reload, and you're there.
>And it is really good at providing server-side programming to a web site with the utter minimum of work
This is exactly the problem with the language. It lets you cuts corners when those corners are there for a reason. For example, look at the ease of deployment that you have mentioned. For a novice, this will be appealing. But for someone with experience with web applications, it might look appalling that you can map urls directly to execute arbitrary files in a directory.
This theme is found throughout the language. In an attempt to make work easier for novices and to make it easy to work with data send from html forms, the language has been packed with so much weird behaviors that you need to learn before you can build stuff in php with any sort of confidence.
And that is just the tip of the iceberg....

> just stick a couple of lines of PHP into an HTML file, hit reload, and you're there
This ease of use and low entry threshold is a double-edged sword though

Reply on news.ycombinator.com