What is the best programming language to learn for backend developers?

Question

32

Options
Considered

819

User
Recs.

Aug 2, 2020

Last
Updated

Here’s the Deal

26 Options Considered

Best programming language to learn for backend developers	Price	Price	Site
88 Rust	Free	Free	rust-lang.org
86 Elixir	-	Open source (Free)	https://elixir-lang.org
85 Python	Free	-	-
82 Golang	Free	-	-
-- Clojure	-	Free, Open Source	www.clojure.org

See Full List

score 49 · Answer 1

Rust

Top Pro

Compiler prevents memory management and concurency mistakes

Top Con

Lifetimes can be tricky to learn

Specs

Price:Free

Site:rust-lang.org

Current stable version:1.4.0

Find the best product
instantly.

Stop wasting time searching endlessly. Lustre recommends the best products at their lowest prices – right on Amazon.

4.7 star rating

Add to Chrome - It's Free Try now - it's free

score 74 · Answer 2

Elixir

Top Pro

Fault-tolerant

Elixir applications are composed of lightweight processes, mainly divided into workers and supervisors. When such a process crashes it can be automatically restarted in a number of ways so that there is minimal downtime. See More

Top Con

Lack of static typing

Should be mentioned that Elixir also has its own way to check type safety : pattern matching, guards, typespecs and dialyzer. See More

Hasnat Babur Raju's Experience

Still requires to know Erlang See More

Specs

Redis -> ETS MySQL -> Mnesia See More

See All

score 95 · Answer 3

Python

Top Pro

Easy to learn

Python has simple to learn syntax which helps beginners to utilize this programming language and has support guide available. See More

Top Con

Slower than alternative languages

Compared to alternatives such as C++, C#, F#, Elixir/Erlang or Go, Python can be sluggish and is often not a good choice for applications where performance is a significant concern. See More

Elias Van Ootegem's Experience

I've tried my hand at python, and I have to say: I'm still baffled at its popularity. It's slow, the tooling isn't great, and the syntax is horrible (I know: "You'll get used to it", and "after a while, the significant whitespace becomes a plus"... spare me the cliches, it's not for me). For quick prototyping, I suppose Python allows you to get started quickly. It's well supported and a great number of people can write it. Those who can't, can at least read it. True, but the performance is pitiful. Because so many people write it, a lot of the code out in the wild really shouldn't exist. The documentation is painful to navigate, the dependency management tool is poor (I mean: PHP has a better tool in composer!). Most of the people who write Python use frameworks like Django, and don't really understand what it does under the bonnet... I've found myself explaining to Python devs why golang yields much better performance. I started off by saying that you don't need something like nginx for a start. Your binary is the server. They seemed confused by this concept. To use python on the back end means you have too much money to burn on CPU power and RAM. Same holds true for any interpreted language in fact. The pythonians (or whatever they call themselves) can say what they want about something not being very pythonic (or "unpythonic"). The fact remains: the language allows you to do things the wrong way. PHP gets a lot of hatred for allowing people to query a database in a template, and rightfully so. Why don't we hold Python to the same standards? PHP is often criticised because it has no real deprecation process. It just ships with one or more replacement modules next to the old one, then at some point it's announced the old module will start issuing E_DEPRECATED notices, and further down the line, the extension will be removed. That is indeed messy. But let's look at Python. How many places still rely on Python 2.x? Youtube (owned by google, who employed Guido Van Rossum, creator of Python) is one. some popular modules only support Python2. How long until the community makes the switch? It's been years now... The language itself feels messy. Passing through the implicit object variable self isn't a huge problem, but it makes it harder to distinguish between static and non-static methods. The "private method by convention" rule is a joke. Python, to me, feels more like a hackers tool. An alternative to the mess that is bash, and more readable than something like Perl. Given the choice, though, I'd rather write a production grade application in anything other than python. Python is a confused language. It doesn't quite know what niche to focus on: is it a language for the web? Django seems to think that it is. Is it for data analysis, machine learning, and science? Others claim that it is. Is it a scripting language that takes away the pain of writing portable Bash, or spending days wading through someone else's Perl scripts? Perhaps. Why not add GUI's to the mix while we're at it... oh, someone did (Pygame, PyGTK, ...). I wouldn't at all be surprised if one of these days we'll see a kickstarter to build a python machine like the good old lisp machines from the days of yesteryear. Some specific oddities about the language that really irked me: Only recently type-hints were added (just checked - about bloody time) No switch-case - a wide-spread construct. This issue is commonly mitigated using dictionaries. If you need a fallthrough case, God help you... Try-catch is again a well-established paradigm, but for some reason, python chose to replace the catch keyword with except. Grammatically, this doesn't even make sense. I read "Try X except Y" to mean: try doing X, except in case Y, implying an exception is checked before the try. I know I'm nit-picking here, but things like this really bug me. No bloody explicit interfaces. Without these interfaces, you have to rely on the hack that is the pass keyword. The pass keyword. I can declare a function, class, or start a block (if, else, except, ... whatever) and tell the interpreter to just not expect anything to be there... just skip it. Save for a few very rare cases, I don't see the point. Even in those rare cases, I'd argue most of those might actually indicate you're doing something wrong to begin with. Like using a language that doesn't have interfaces, for example. Pythonians will say python is strongly typed, and it is. Internally, however, everything is a PyObject. This enables Python to be dynamically typed, but for most use-cases for backend development, you know what types you're dealing with. A statically typed language, like it or not, is safer. Python values are typed, variables are not. I can reuse the same variable for different types. It should be obvious how this makes code more error prone as time goes on. The threading issues are bizarre. Depending on the implementation, you might have things like GIL to contend with. Either way, giving threads to inexperienced developers (like many python devs are) is like taking a group of toddlers on a trip, first to a brewery, then a distillery, to end up at a firing range. It's going to get messy. I mentioned it before, but the syntax... yuck. Significant whitespace is all well and good, but leaving blank lines to make code more airy is important. brackets make it more clear when a block ends (editors even highlight the closing bracket for that reason). No such luck for python devs. See More

Specs

Current stable version:3.8.1 and 2.7.16

Python's syntax supports optional type annotations for use with a third-party static type checker, which can catch a certain class of bugs at compile time. This also makes it easier for beginners to gradually transition to statically typed languages instead of wrestling with the compiler from the start. See More

Top Pro

Includes pygame library

Want to start game development? No problem! Using pygame open-source library you can fast begin creating games without worrying about pointers or undefined behaviors which they exists in C/C++. See More

Top Pro

Multiparadigm language

Python supports three 'styles' of programming: -Procedural programming. -Object orientated programming. -Functional programming. All three styles can be seamlessly interchanged and can be learnt in harmony in Python rather than being forced into one point of view, which is helpful for easing confusion over the debate amongst programmers over which programming paradigm is best, as developers will get the chance to try all of them. See More

See All

score 74 · Answer 4

Golang

Top Pro

It is supported on all the mainline operating systems and supports cross-compilation

You can compile an executable for linux, windows, OSX, FreeBSD, NetBSD etc. You can easily cross-compile accross platforms. See More

Top Con

No user defined generics

This means that you can utilize strong typing only as long as you do not need a generic container like for example list. Only ways to deal with it is either code duplication or using interface{} which means that you lose strong typing. See More

Roman Scharkov's Experience

Go provides a very simple mental modal to solve a relatively complex computer programming problem, that is concurrency. Goroutines make asynchronous, non-blocking and highly concurrent code much easier to write and read. Unlike in many other languages running on top of an event loop, like JavaScript for example, where you have to write non-blocking async code properly to not shoot yourself in the foot blocking the single execution thread - in Go writing blocking synchronous code is just fine, because the builtin goroutine scheduler will automagically swap any blocked goroutine for another to not waste any CPU time. What makes goroutines special, is that they are very leightweight compared to regular OS threads we know from Java or C++ and similar, you can spawn millions of them and Go's scheduler will use up all available CPU cores to run as many of them at once as possible. The Go compiler compiles the source code into final machine code which is executed directly by the CPU bypassing any interpreters and virtual machines eliminating the runtime costs of such abstractions as well as any runtime dependencies. Although the compiler isn't probably as smart as a regular modern C or C++ compiler yet, it's compile times are very low and the generated machine code is very efficient making Go blazingly fast, almost as fast as C/C++ (not quite, but still). The built-in concurrent mark-and-sweep garbage collector has it's tradeoffs (like any another GC) but is optimized for latency to minimize stop-the-world delays and does it's job pretty well. The compiler also tries to optimize memory efficiency by trying to allocate as much on the stack (which is also special in Go, but that's a whole other topic) as possible through escape analysis and similar techniques. Besides that, Go offers a strong static type system, supports reflection and encourages duck typing, interfaces instead of classes and composition instead of inheritance. The language grammar is very simple, it makes working in a team of people with different levels of skill a piece of cake. Go has basic support for generics. The builtin map, array and slice containers are fully generic. Documentation is built into the language. Source code comments are automatically transformed into documentation. godoc.org constantly scans for Go repositories and generates public documentation for them on-the-fly. Modular design is achieved through packages, every folder represents a package containing exported and unexported functions, types , constants and singleton variables. All that together with a rich standard library and a high quality toolset make Go a simple, elegant and highly efficient platform for application and systems development. See More

Specs

Current stable version:1.12.1 / February 25, 2019

Platforms:Windows, Linux, Mac etc.

Clojure

Top Pro

Simplicity

Top Con

Requires the JVM

InterestingAm-heh's Experience

Truly the best general-purpose language, it has great language design and all the functionality you could wish for with back-end development. For free you get front-end thrown in too, thanks to Clojurescript an d Garden CSS. See More

Specs

Price:Free, Open Source

Site:www.clojure.org

Current stable version:1.10.1

Platforms:Windows, Linux, Mac

Java

Top Pro

Lots of libraries

Top Con

Bloated

For the sake of your code's correctness, you had better write (or have your IDE generate) getters and setters for data fields you expose. And for the sake of safety you should definitely add null-assertions to all non-null method parameters. You also have to catch or rethrow/redeclare all checked exceptions. This, among other things, makes the average Java code extremely bloated and difficult to read. See More

Abioye Bankole's Experience

I have used Java for more than 18 years and its concept of object-oriented programming makes code management a worthwhile activity. See More

Specs

Price:Open source (free)

Site:https://www.java.com

Current stable version:10

Top Pro

Mature corporate multi-platform language

Top Con

Can be hard to choose right stack (frameworks) for beginner

Top Pro

Novel ideas

See All

score 23 · Answer 7

C#

Top Pro

.NET is a great toolbox

C# runs on top of the .NET framework, which provides many libraries containing classes used for common tasks such as connecting to the Internet, displaying a window or editing files. Unlike many other languages, you don't have to pick between a handful of libraries for every small task you want to do. See More

Top Con

Lacks standard-library support for immutable data structures

Specs

Price:Free, Open Source

Current stable version:8.0

Platforms:Windows, Linux, Mac, Android and IOS(via Xamarin and Unity3D)

License:MIT, Open Source

JavaScript

JS code is usually quite small compared to other languages, especially with ES6 syntax. See More

See All

score 14 · Answer 9

Haskell

Top Pro

Safe

The strong type system makes Haskell a safe language to use See More

Top Con

Symbols everywhere

Haskell allows users to define their own infix operators, even with their own precedence. The result is that some code is filled with foreign looking operators that are assumed to be special-case syntax. Even for programmers who know they're just functions, operators that change infix precedence can potentially break expectations of how an expression is evaluated, if not used with care. See More

HarmoniousPonos's Experience

Haskell server frameworks are expressive and the APIs reduce the contact surface. Low boilerplate, high level, and as much a joy to use as the language. See More

Specs

Site:https://www.haskell.org/

Paradigm:functional

Type System:static

Top Pro

High level

Haskell is a very high level language, having tons of abstractions out of the box, making it perfect for backend development. See More

Top Con

Package manager is unstable & lacking features

Cabal (There are other choices but this is the most popular) can not uninstall a package. Also working at a few locations it is difficult to have the same environment for each one be the same. See More

monk's Experience

TypeScript

Top Con

Type checking not enforced by default

You have to use compiler flags to make sure it catches flaws like usage of implicit any, etc. See More

Top Pro

Optional static typing

Typescript has optional static typing with support for interfaces and generics, and intelligent type inference. It makes refactoring large codebases a breeze, and provides many more safeguards for creating stable code. See More

Ryan O’Hara's Experience

Reasons to use JavaScript: no compile step What TypeScript adds to JavaScript: a compile step See More

Specs

As a Microsoft developed project, it has first party Visual Studio support that's on par with its C# support with features like syntax sensitive statement completion. See More

Top Pro

Built and supported by Microsoft

Being built by Microsoft, TypeScript is much more likely than most other similar open-source projects to receive continued long-term support, good documentation, and a steady stream of development. See More

Top Pro

Has a repository of high quality TypeScript type definitions for popular libraries

There are many ready to use and high quality TypeScript definitions for popular libraries including jquery, angular, bootstrap, d3, lodash and many-many more. See More

See All

score 41 · Answer 11

PHP

Top Pro

Widely used for websites

PHP is the quasi-standard for backends for web applications and used by most of all websites. Many popular CMS are built with PHP, e.g. Wordpress, Drupal and Joomla. See More

Top Con

Very slow compared to alternatives

Compared to alternatives like Exilir and Go-lang, which provide comparable functionality, even the latest versions of PHP are vastly slower than their competitors. The difference is so significant that it has the potential to bottleneck even modestly sized applications and lead to highly significant increases in hardware costs, even with well-written code. PHP's syntax is also highly permissive of non-optimized code which can further exacerbate this issue. See More

Joao Gilberto Magalhaes (JG)'s Experience

A lot of tools for CI/CD environment See More

Specs

Price:Open Source (free)

Current stable version:7.3

Top Pro

A lot of good frameworks for it

Like Laravel, Zend, Symfony... See More

Top Con

Lack of static type safety

The lack of strong static type checking means that simple problems that could have been easily detected at compile time are detected at run time. Clearly that's a disastrous problem for back-end software development. See More

Luiz Barni's Experience

Runs JavaScript trought PHPV8 See More

Top Pro

Huge community of practice

Whatever challenge you face, you can Google it and find existing code that solves it. Well supported packages, objects, and methods exist for data manipulation, graphic generation, file handling and connecting with diverse APIs. See More

Top Con

Poorly designed language

Top Pro

Tons of built in functionality that don't require external libraries like most other languages.

Top Pro

Runs JavaScript through PHPV8

Yes, PHP can run a React app directly from itself. See More

See All

Answer 12

C

Add Video or Image</

What is the best programming language to learn for backend developers?

26 Options Considered

Rust

My Recommendation for Rust

My Recommendation for Rust

Compiler prevents memory management and concurency mistakes

Lifetimes can be tricky to learn

Catch errors at compile-time

Support for macros

Threads without data races

Compiles to machine code allowing for extra efficiency

Lifetime semantics have benefits in other languages

Generics support

Built-in concurrency

Supports cross compilation for different architectures

Makes developers write optimal code

Easy to write understandable code

Zero-cost futures or Async IO

Most popular programming language 2017

Official package manager and build tool

Functional Programming

Big community

Elixir

My Recommendation for Elixir

My Recommendation for Elixir

Fault-tolerant

Lack of static typing

Great concurrency and scaling

Takes some time to wrap your head around around OTP supervisors trees

Pleasant Ruby-like syntax

Requires some knowledge of Erlang

Built-in support for hot-upgrades and hot-downgrades

Still requires to know Erlang

Great for concurrency

Bad for CPU bound processing

Easy interoperability with Erlang libraries

Some design choices may seem strange

Great documentation

Immature backend runtime environments

Easy to download libraries

Doesn't have wide adoption in industry

Great getting started tutorials

Deployment is still not as easy as it should be

Powerful metaprogramming

Scalability

Full access to Erlang functions

Distributed mode

Functional programming

Syntax is similar to Ruby, making it familiar for people used to OOP

Types don't get in the way

Soft real-time

Regular expression-based method overloading

Built in Redis and MySQL

Python

My Recommendation for Python

My Recommendation for Python

Easy to learn

Slower than alternative languages

Lots of tutorials

Slow to run

Easy to get started

Speed can matter

Clear syntax

Not type-safe

Active and helpful community

Unstable python enviroments

Very similar to pseudo-code

Not good for mobile development

Comes with extensive libraries

Version Confusion with V2.x and V3.x

It's really simple

Hard to debug other people's code

Good documentation

Language fragmentation

Can be used in many domains

Might not be very future-proof

Cross-platform

Limited support for functional programming

Has many libraries for scientific computing, data mining and machine learning

Does not teach you about data types