The Python programming language is almost finished with a long-term transition from version 2 to version 3. New programmers often have questions about which version they should learn. It can be confusing to hear that Python 3, which was originally released in 2008, is still not the default installation on some operating systems such as macOS. However, that situation is rapidly changing as the final version 2 release, Python 2.7, is approaching its end-of-life that is definitively scheduled for January 1, 2020.
The simple answer right now is: learn Python 3, specifically the latest version which as of May 2018 is Python 3.6.5. If for some reason you absolutely have to learn Python 2, for example because your employer is working on a bunch of legacy code, you will be able to transfer the majority of your knowledge from Python 2 right into Python 3. Likewise, you will still be able to read and write Python 2 code if you start with Python 3.
There are enough great resources out there that will teach you to code in version 3 without any prior version 2 experience. Python 3 is the future and you will not regret starting with the latest version of the programming language.
One caveat to the recommendation to go full-on Python 3. If you are interested in DevOps-type work with configuration management tools such as Fabric or Ansible, then you may have to sometimes use both Python 2 and 3 because parts of these libraries are lagging in full Python 3 support.
Since upgrading from Python 2 to 3 has been such a huge undertaking within the community, many projects have sprung up to make the transition easier.
six is a 2/3 compatibility library that is a dependency for many popular Python projects to make it easier to support both Python 2 and 3 at the same time.
Python 3 Readiness is a visualization of which most popular 360 libraries (by downloads) are ready to be used with Python 3.
The Python clock counts down the time until Python 2.x is no longer maintained. While Python 2's retirement may still seem a long time away, it can take a lot of time and effort to migrate existing application to the modified syntax in 3.x.
Moving an existing codebase to Python 3 from 2 can be a daunting task, These resources were created by fellow developers who've previously gone through the process and have advice for making it less painful.
Python 3 Porting is an entire book with details for how to upgrade your existing projects and libraries to Python 3.x.
Moving from Python 2 to Python 3 is a PDF cheatsheet for porting your Python code.
The official porting code to Python 3 page links to resources on porting Python code as well as underlying C implementations. There is also a quick reference for writting code with Python 2 and 3 compatibility.
Django and Python 3 How to Setup pyenv for Multiple Pythons is a screencast showing how to run both Python 2 and 3 for different projects using pyenv.
Upgrading to Python 3 with Zero Downtime supplies advice on transitioning a large existing Python 2 web application to Python 3. Their process involved upgrading dependencies, testing and deploying the new version before going back to clean up unnecessary code created by the transition.
The only problem with Python 3's str is that you don't grok it specifically covers strings and some of the misconceptions around how strings are handled in version 2 versus version 3.
Migrating to Python 3 with pleasure is a porting guide that focuses on code that data scientists would typically use in their programs.
Instagram Makes a Smooth Move to Python 3 explains their upgrade process for getting all of their code over to Python 3 compatibility over a period of about a year.
Practical steps for moving to Python 3 is a podcast episode that goes over migrating a large existing application's codebase to Python 3 from Python 2.
The following resources will give you more context on how the community feels the transition from Python 2 to 3 is going, as well as why you should upgrade as soon as possible.
Why should I use Python 3? is a detailed FAQ on important topics such as unicode support, iteration improvements and async upgrades provided by 3.x. There is also a great follow up post by the author titled A Rebuttal For Python 3 that counters some arguments made by other community members who are unhappy about various features in Python 3.
Want to know all of the advantages and what's changed in Python 3 compared to Python 2? There's an official guide to Python 3 changes you'll want to read.
Python 3 is winning presents data and graphs from PyPI to show that at the current rate, by mid-2016 overall Python 3 library support will overtake Python 2 support.
Python 3 Retrospective from the Benevolent Dictator for Life is a talk by Guido van Rossum on what is working, not working and still needs to be done before the changover can be considered complete.
The stages of the Python 3 transition provides perspective from a core Python developer on how the transition from Python 2 to 3 is going as of the end of 2015.
How Dropbox rolled out one of the largest Python 3 migrations ever explains how their transition began in 2015 and was successfully completed in 2018.
Why Python 3? randomly outputs valid reasons to use Python 3 over 2.x.
Rules for Radicals: Changing the Culture of Python at Facebook is a fascinating look at how Facebook moved from primarily Python 2 up to Python 3 due to the efforts of a small passionate group of developers within the company. Definitely worth watching to understand how to shift a large organization with an established codebase.
Porting to Python 3 is like eating your vegetables explains that there are treats in Python 3 that are worth porting for and has some tips on making the transition easier.
Scrapy on the road to Python 3 support explains from the perspective of a widely used Python project what their plan is for supporting Python 3 and why it has taken so long to make it happen.
All major scientific Python libraries have pledged to drop Python 2 support no later than 2020, when Python 2's maintenance life is over. The pledge strongly encourages Python 3 adoption by publicly stating their intentions.
There isn’t really a Python 2 vs Python 3 Problem tries to cut past the concern about the Python 2 vs 3 divide and explain that it's really not a major concern as a Python programmer. The recommendation to new programmers is to use Python 3 and for experienced developers to keep your explanation simple: Python 3 is where to begin.
10 awesome features of Python that you can't use because you refuse to upgrade to Python 3
is a great slideshow with code snippets that show useful new features
of Python 3 that are not available in 2.x such as keyword-only
arguments, chained exceptions and the
yield from keyword.