Ask HN: Which language(s) should I learn to do one-man projects and earn money? | Hacker News Reader

Ask HN: Which language(s) should I learn to do one-man projects and earn money?

31 pointsbehnamoh75 comments

1) No CS/programming background, but interested in functional programming (clojure, Haskell, CL...)

2) I don't intend to work for other startups/companies, just want to build my own projects and monetize them.

3) I like iOS/android programming, but I'd need to learn Swift+Java for both, and cross-platform solutions don't seem alright.

4) I also like web-apps, but I hate JS, and frankly, don't like any imperative lang for backend (python, ruby, ...)

?) Besides web and mobile dev, is there any other field where I can do solo? (of course, w/ money)

View on HN

Personally I use JS extensively for these kind of things. I hate JS the way most people use it, but JS is actually a crazy powerful functional programming language in disguise. And you can use it on server or client almost transparently. I mostly use only a subset of JS myself: I write const fat arrows that either return functions or immutable objects. No or very little state. I don't use "this". I use react + redux on the front-end. I maybe use 4 different language constructs total, both on the server and on the client, and my life as a programmer has become very peaceful, I can focus on delivering features. The nice thing about JS is, like the language or not, the tooling and ecosystem are awesome. For one man projects you can really make the language your own and achieve crazy productivity. YMMV.

Here's an example of what I mean, a Promises/A+ implementation in less than 100 LOC - just for the fun of it: https://github.com/djfm/promesse/blob/master/lib/promise.js

I think it kinda feels like Haskell, yet browsers understand it :)

> JS is actually a crazy powerful functional programming language in disguise.

I feel the same way. And since you can write a cruddy jQuery mess that fetches data and spits out to backend-as-a-service providers, Javascript is hands down the easiest way to go from nothing -> project others can use.

The ecosystem is in a bunch of growing pains right now, but if you don't care about maintainability and scaling up to a large project, just throw up an HTML page with some styling and javascript that makes it interactive.

I have a feeling the ecosystem is becoming, err, slightly less worse :) My react redux babel boilerplate has not changed in about a year and I strive to stay up to date. Once you can write a webpack.config.js without looking at the doc you're all set :)

100% agreed. Webpack 2 is going to be a fairly sizable change, but for the better. I've been following the webpack meeting notes, pretty excited to see where it's all headed.

It's hard to compare JS to Haskell without mentioning the JS's complete lack of a type system...

That said, if you want something more ML-like in your JS, Flow is an increasingly-viable option: https://flowtype.org/

I think as a beginner, I need to start w/ "standards" first. I mean, shouldn't I first know JS to understand Flow? Wouldn't learning FlowType put me in a corner where I don't have all the support of JS community?

Starting with standards is probably good yes. Something I found very helpful in JS is doing hard-core TDD at least at first (again, the tooling is awesome, see mocha, chai) - then after having done the same thing a hundred times you can become more relaxed. Good unit tests make up for a lot of missing types.

Yeah, I've been experimenting with it for a few days and I find it very promising. But the lack of types in JS is not such a pain point IMHO. In JS you can define a monad that works only with numbers, which is sometimes useful, and which you cannot do in Haskell without black magic. I think Flow's idea is that most JS programs are actually naturally well typed, and it often verifies in practice. Just because you don't have a compiler yelling at you doesn't mean your program is not well typed :)

That being said I cannot wait for the time when Haskell becomes truly mainstream in web development.

> In JS you can define a monad that works only with numbers, which is sometimes useful, and which you cannot do in Haskell without black magic.

What does this mysterious assertion mean?

I don't know a lot o javascript, using clojurescript instead, but you've piqued my interest, is there any learning resource for that kind of javascript use?

Have a look at redux's doc for instance, http://redux.js.org/docs/introduction/, and do read the code (it's about 300 lines total), they're both masterpieces of modern JS.

If you really know cljs then you are already doing it in a less kludgy and more elegant/sane way ;)

I have worked with Redux as well, and write functional JS pretty much exclusively, and agree that it's actually very pleasant to work with these days, especially if you're also in an environment where you can take advantage newer language features like async/await.

My #1 pain point with functional JS though, is that using a persistent data structure library like ImmutableJS or Mori removes the ability to use some of the awesome new syntactic sugar introduced in ES6 like the spread operator, destructuring, etc, for making code less verbose. So you have to make that tradeoff between being able to use a succinct syntax for common operations vs having the optimal performance in terms of memory efficiency and garbage collection pressure.

I'd really love to see native persistent data structures as a first class citizen in JS so they can make use of all new language features. Until then, I'll probably continue to choose ClojureScript for my own projects.

You're making things too difficult.

You are looking for the one right answer. No such thing. Whichever you pick will have its short comings.

Remember programming languages are tools, they exist because they fill a need.

Also you have a lot of points here. Get started with one, then move on to the others.

You said you are interested in functional. Pick one. The easy way is to pick the most popular. Do a google search. Pick the most popular one, at least you'll be able to ask for help and have someone that's been in a similar situation.

For mobile, what do you have available? Pick that platform. iOS takes a bit more money. If you have both pick the one you use the most. I hear iOS brings in the most money but unless you have a great idea it's hard to standout from the millions of apps.

For web, look at what's popular so you'll have help when needed. People will tell you to use this or that but once you get started and have a few projects you'll understand what's best for you.

You can ask and ask but you will never know until you get going. Define a project, pick what language you find interesting at random and focus on only that until you finish a project.

You'll find out what you like and dislike and you'll be able to pick what language is best for you.

Those are really good points, thanks!

I'll try to just get going and pick a lang for that. Web-dev intimidates me, cause I should know HTML/CSS+JS for front-end, and SQL+{ruby|python|clojure|haskell|php|...} for back-end. That's why I was thinking about going w/ sth like clojure/ruby, cause that takes care of front-end as well. But what do I know?

Take a look at hackathon starter kit, you can do javascript in the front and backend, I haven't used it but I think it looks interesting. https://github.com/sahat/hackathon-starter

Dude this is basically saying "I want to build something, but I hate programming".

If you don't want to learn js, then forget about web programming.

If you don't want to learn any of swift or java, then forget about mobile programming.

If you don't want to learn java/python/ruby, then forget about server programming.

That leaves you with nothing, maybe learn some MS excel to become an accountant? That qualifies as one man project that will earn you money.

Obvious troll is obvious.

Obviously OP is too picky, but there are plenty of server-side and client-side options that fit the requirements. Eg Clojure / Clojurescript, Haskell, Scala, Typescript

Reminds me of that Ricky Gervais bit about fat people. "I'll do anything to lose weight. Not jogging..."

>Obviously OP is too picky...

Yes, unfortunately. I know it, but just can't help it. Maybe I'm an idealist, or maybe just not that familiar w/ the programming world.

you sound like a fat person who is trying to get in shape but is extremely picky about which 'routine' he wants to follow. he has a reason to hate each one so, they're all not good enough for him. he's looking for the perfect routine, man. one that will guarantee success.

oh and also that annoying bit about eating right, well that sounds like a pain, so let's not do that, either.

all this, while sitting on the couch, daydreaming of 6-pack abs. in reality, all you have to do is walk around the damn block and you're 1/4 mile ahead of where you were before.

Oy, quit stealing my analogy!

I actually spat my coffee out laughing

> If you don't want to learn java/python/ruby, then forget about server programming.

Guess I'll have to inform the .NET, PHP, JavaScript, Clojure, Common Lisp, Haskell, Erlang and Elixir[1] guys that they are, in fact, _wrong_.

[1] ...and C, and any programming language in existence...

I think that the parent's point isn't about languages or ecosystems.

No disrespect, but why are you trying to optimize your choices so severely before you've even learned anything? You seem to know 'about' programming, but once you learn the fundamentals of programming, you can learn any language or platform you want.

I think you should just choose whatever lets you build the project you want to start today, forget trying to cross platform / future-proof your decision.

Thank you.

> why are you trying to optimize your choices so severely before you've even learned anything?

Maybe because I can't become a full-time developer, so choosing the right language matters. I actually envy "real" programmers who have time to learn more than just one lang...

>I think you should just choose whatever lets you build the project you want to start today...

I like mobile/web apps, but I don't seem to find any other field where I can do on my own. Do you have any suggestions?

Most developers learn in one major language first, then adding others as they expand. Since you're looking to make money by selling software, you're probably going to work on 1) a website or 2) a mobile app or 3) desktop app (basically similar to a mobile app; compiled software probably distributed through an app store).

Do you have an Android or an iPhone? Download the IDE for your respective platform, get Hello World running on your phone. There, you're now a developer :D. Next step, follow some tutorials. Don't paralyze yourself with these decisions that seem important, they're really not.

Yeah, I think I have made this way too difficult for myself... I just wish I could have all the good features of different languages altogether, but that's not gonna happen, welcome to the world of programming.

I like Haskell and LISP for the challenge, but I wish I could do mobile/web dev in them easily (like RoR).

I really appreciate your answers, I think I should just get going then, and try to figure out the rest next.

GHCJS and Purescript work with react.js and directly with the DOM. So there are options on the client side as well if you wanted FP. You could also consider learning clojure, and closurescript. A few backend teams I've met retooling all their RoR code to Haskell. Sounds like you are on the right track with your thinking, but with no CS background be sure to read books that help you understand the pros and cons of different languages.

Specialised line-of-business apps for small biz work great. The downside is you need to know their line of business well yourself. The best markets are the untapped ones where, if you tell someone what you do, they had no idea that job even existed.

Many of these have succeeded while being written in Visual Basic.

1) Some background is necessary for building and monetizing projects. Not necessarily university education, but you're not just going to pick up a book on CS and start monetizing projects that month. What's your time frame for this?

2) "Just". This is not a trivial exercise. I recommend reading your way through patio11's extensive writings on bootstrapping software companies if you're serious about this

3) OK.

4) You can't really build web apps without JS, at least in some form. Which part about web apps do you like, if it's not JS or "python/ruby/...". This is probably the easiest way to monetize software projects these days though, so maybe reconsider. You can still do some pretty nifty things if you find a niche market and put up a public facing CRUD application with a few lines of Django/Ruby boilerplate.

?) I think all that's really left after your exclusions is to try pick up on the dying markets like Wordpress/CMS development. This can be done through plugins or managing entire sites for other people, or building your own ecommerce business. For the latter, look into ready-made stuff like Magento. This is really how people made money a few decades ago though -- I know a few people who have tried recently and failed.

Last suggestion is to look at more business oriented stuff. VBA is still exceptionally powerful in terms of integrating into tools (Excel) that non-tech people use. A lot of managers etc spend an inordinate amount of time doing repetitive work on spreadsheets -- you can save them hours every day with a few lines of VBA dev. Not fun, but might be what you're looking for.

Good points!

About that VBA thing, I was actually thinking about automation. Bash, Python, ruby, even JS, are really great for that. But then we have VBA. I guess most things you can do with VBA in Excel, you can do with python as well.

I agree, I need some background building before jumping into programming. I just figured maybe a good programming book will cover the necessary CS parts, too. "Learning Python" is over 1500 pages, and really is intimidating. "Haskell Programming" is also +1000 pages. Do you think picking up huge books like these, will help me build the required background?

Yes, everything you can do in VBA you can do in another language, and probably more easily. But then you have to persuade the users to use your app - if they're already comfortable with spreadsheets, VBA can make it easier to get them using your product.

If you can work your way through an intro book, it'll be super beneficial! Everything in either of those books will probably be useful at some point or another. No way I'd be able to do that in your position though. Personally the only way I can learn new technology is to have a goal that I find motivating (often, 'build a system that can solve some pain point in my life') and then identify the minimum useful step I can take towards that goal, and then mess around with books, forums, documentation, etc until I've achieved that step.

Generally it works the other way round: you have an idea for making money and need the best (or easiest) tool (or language) for execution. Starting from nowhere, your best shot is learning to customise e-commerce templates and set up stores online for you or for prospective clients.

If you're just getting into things Elm might be a good bet for front-end. Maybe you'll get fewer contract gigs but the current state of JS is a mess, might as well skip it (for now at least).

If you don't have any experience and programming background you should start with C/C++ to learn basics, data structures, algorithms etc. Then you may continue with C++ or move to different language - for example, C# or even F# if you like functional. With C# you can write mobile apps for both platforms using Xamarin. Please don't start with JS, actually try to avoid JS as much as possible because that's not programming language - but that's my opinion.

Yikes. Dis-recommend. I started my career in C, which was my first programming language, back in the mid-90s. It's been advantageous for me in my particular field (vulnerability researchers should know C), but I look back and wince about all the energy and time I wasted shipping code in C and C++.

It's also, for this particular question, the family of programming languages most hostile to functional programming.

ES5 JS is a clunky language, but people have built amazing things in it. ES6 JS is a credible language on its own terms. I wouldn't be wary of someone saying "Javascript sucks" (I would mostly agree), but I would be wary of someone who said "Javascript isn't a programming language".

OK, as I said, that's my opinion but I also appreciate yours :)

It's an awfully hard question to answer without some idea of what you are trying to do to make money.

For instance you really can make web apps with very little Javascript and, for many applications, this is the way to go.

For instance, if you take the "path less taken" and not load your site up with 15 MB of Javascript, tracking pixels, third-party scripts, another 100 MB of junk from a CDN and instead move a minimal amount of data through http/2 you can make a site that loads like a bullet being shot out of a gun on a computer, phone, tablet, whatever.

This site

https://legalentityidentifier.info/lei/get/5493000HEH4YCM7L6...

has competitors that take 2 minutes to load the first page because the developers throw in everything but the kitchen sink so they can pad their resumes. It's hard to get people to see the difference, however, without actually sending them to a competitor.

Frankly I think CDNs are a scam like DOIs. DNS lookups are one of the worst things when it comes to long tail latency and if you hit enough servers the odds approach 100% that the browser is going to freeze up for 15 sec during loading because of a DNS hang.

It depends a lot on your priorities between mastering a language/programming and having a successful product/service (ie. making money). In any case you should figure out first what you want to build as the choice of tools will depend on it.

Also keep in mind that while functional programming may be fun, it may be simpler and faster to just slap some php onto some random webhost (even more so when validating an idea and getting feedback) :)

You will not find good answers to poor questions. You must first decide to build a house, then pick up the hammer.

Build me an iPhone app, written in Swift, that will let me sell my recycling (cans, bottles) to someone else so I don't have to drive to the recycling centre to collect the refund.

If you research your market, your audience, and verify your idea through testing and prototyping, you can get your business started without ever writing a line of code.

If you want to build things with interfaces that users use, I implore you to reconsider JavaScript. First of all, it's a powerful functional language (https://www.youtube.com/watch?v=ya4UHuXNygM).

React and Redux are heavily inspired by functional languages, and the two can be used on the server-side as well (which, if you're serious about web development, is going to be important).

With React Native (https://facebook.github.io/react-native/), sharing code becomes even easier for mobile development without really sacrificing the quality of the end-result (no more WebViews).

If you want JS to have a type-system, consider TypeScript (https://www.typescriptlang.org/), or if you want a more powerful ML-like type-system consider using Flow (https://flowtype.org/).

I can completely relate to this mindset.

I was once similarly minded in that I had read enough about programming to know what was out there and what it was used for.

I then formed impressions about what was 'hard' and what was 'easy' and what I would like. It took me ~2.5 years from the start of my desire to learn, to actually write my first line of code. Picking a language, then picking a learning methodology took me that long.

Then, 2 years of learning/programming on the side, I got my first paying gig for writing code (I'm was never a full time software engineer, just learning on the side, like you). This was 2 weeks ago.

Perhaps this is a common path, or we're a special breed. What helped me stay motivated was trying to solve my own problems with code. First project was a script that took data from a google spreadsheet and texted it to me at relevant times. Took me 3-4 months to complete. Could write it in 3-4 hours now.

Find a (small) problem, write code to solve it. Monetization will come, but probably slower than you expect. Good luck.

Man, this is so like me!

I guess my concern is that I want to do quality work (with well-designed languages) AND earn money for it. I'm beginning to realize that (sadly) it's not really how things are done in the programming world. As long as you can deliver a fancy web-app in short time, nobody cares if you have used JS or ClojureScript.

It's true no one cares for the most part. But clients do want maintainability, so a more mainstream language makes your skill-set more attractive. Of course I speak as a freelancer, not from a product perspective.

I picked Python, because:

1. I thought I would give up if the language was too hard.

2. I didn't know what I want to build eventually, and Python seemed to be used everywhere.

Getting into the mindset of spotting problems, defining it's scope and writing code to solve that SPECIFIC scope has been highly useful. It's also something to continually improve at. Good places to start are complaints from friends/family about commonly used software.

Elm. http://elm-lang.org/

Have a look at Meteor (https://www.meteor.com/).

- JavaScript is basically a functional programming language, too.

- Meteor has Cordova integration for mobile development. However I never tried out Cordova, so I can't say whether it's good or not.

- Meteor is JavaScript, but as others have pointed out, your list of conditions doesn't make sense, so I ignored your fourth point. Or perhaps program in CoffeeScript, TypeScript or a different language which Meteor can compile to JavaScript.

- For the last point, there's a lot you could do. Depends on your experience of life. One example is emigrating to a country with low living costs and provide your knowledge from there remotely. This way you can sell for a lower price.

Hadn't heard of it before. Sure, will have a look at it. Thanks!

Given you'd like to focus on functional languages and you'd also like to be productive enough to build and monetize apps as an individual, you might like and want to look into Elixir and Elm.

Elm is a nice functional frontend language that compiles to JS, which you'll need for web apps.

Elixir is a very good and solid functional backend language that is good for general purpose programming, as well as web app programming (via Phoenix) and embedded (via Nerves).

Both of those languages are functional and focused on developer productivity and have some very good tooling around them.

Thanks. I had just heard their names before, but now I'm thinking maybe the combination of Elm+Elixir would suit my needs. Some other people have also suggested Elm here, didn't know why.

I use Haxe for a lot of side projects. http://haxe.org/ It's a cross-compiling language, so it can output javascript, java, python, c#, c++, php, and lua. It covers every platform I'm currently interested in working with.

There's communities within Haxe that focus on things like game development for iOs/Android, or react-style web programming.

You can't expect Haxe to completely smooth out all of a given target's problems and quirks, so expect to sort out some target specific quirks if you're really trying to push performance (e.g. mobile gaming). However, having most of the core business logic in one language is incredibly gratifying. And, unless you're using Scala or Haskell, I bet Haxe ends up being the most expressive and powerful language that you use day to day.

Even if you hate it, the most pragmatic choice would be Javascript, it covers the browser, the server and mobile apps but you should spend at least one semester doing an introductory programming course, there are great courses online.

How are you planning to make money? Making money building things on your own has very little to do with which language you use. It sounds like you're just stalling.

I think language of choice decides how I can earn money, see:

Swift -> iOS

Java -> android

JS HTML CSS -> Front-end

RoR SQL -> Back-end

C# -> Desktop software

...

Personally I pick clojure by default. Clojure is a wonderful LISP that runs on the JVM which means there is all the libraries you need.

If I can't pick clojure, I gonna pick C because C works everywhere too and has all the libraries you need too. C is not functional but it is imperative and pretty damn simple.

I'd start with what you already know. Depending on the project, different languages will be a better fit.

If you are just starting out and are focusing on web development, I'd recommend JavaScript since all you need to get going is a browser and a text editor.

I hear all this hype about clojure/haskell being a better fit for web apps. Esp. ClojureScript which compiles to JS.

Is it sane to ditch JS (and node.js, angular.js, bootstrap.js, ...) and go with clojure? (maybe even ruby?)

If you want webapps, it's good to understand Javascript. I love clojure to death so much I write it in my spare time as a developer working on other stuff. But the reality is that its surrounding community is tiny, and that makes the hurdles faced when starting much, much larger.

Clojure/clojurescript will get you a single language on front-end and back-end, but I think that from a monetization standpoint you really want a Javascript front-end and Java or Ruby back-end.

Keep in mind that if you use "more standard" technologies, you'll also have a more likely chance of cashing out on whatever you build if you decide to sell because the buyer will be able to hire people to fix it. If you choose something obscure without a developed community, any price offered to you will be much, much lower or even just never come.

I'd personally skip the node.js hype for now.

Thanks, so I guess clojure is not suitable for a beginner now.

On one hand, I like using something that makes me think and also gets the job done (clojure). On the other hand, it feels good to be a member of a huge community (JS, ruby, python developers) and know what many others know.

I had not thought about your points about selling a software in a "less standard" lang. I think you're right.

Agree with everything you said about Clojure[script], but isn't nodejs past the hype already? sure, articles are written for it everyday but is a technology that is used heavily in the industry, that has a large community and library ecosystem.

It's true I haven't used NodeJS for actual projects, but writing entire apps in Javascript where other languages are available does not appeal to my personal aesthetics or technical sense on many levels. Simple namespacing in JS is a burden. A lot (not all) of what I hear about NodeJS being a great tool is from people that both accept callback hell as a way-of-life, and who decline to understand anything past the headlines of great performance.

Guess how that great performance comes about - C++ and non-blocking I/O! There's also the aspect of the community that makes packages for single functions and can't develop a useable package manager (last time I checked). I'm not against it as pure infrastructure for running clojurescript, though =P.

Depending what you're doing then you can do lots of languages to some level with a browser, eg https://www.google.co.uk/search?q=repl.

Ruby, rails and intercoolerjs is the best combo for what you are looking for.

It will still be a logic work.

What is your previous programming experience?

Just a couple of courses on C++ in college. Although I did work with MATLAB for my projects, if that counts.

Brutally honest: Well if you need money and hope to earn them by programming: go for something with low barrier of entry and high demand.

That used to be PHP + Javascript. You should still be prepared to put in some work in personal projects etc to convice people that you know what you claim.

If you have all time in the world and want to be perfectly positioned for some perfect future opportunity, I don't know, I have no idea.

Personally however, if I am the one to hire in the future I'll try to go with people who get stuff done.

How can you hate JS if you have zero programming experience? Try it first.

Maybe the OP saw some code samples involving jQuery

then scheme is made for you. learn racket-lang.org and give a try to "lambdanative.org" to build native mobile apps.

Elm, JavaScript Swift, Java Python

Do it in APL

English?

Reply on news.ycombinator.com