advice for junior devs

TLDR;

Build stuff. Build it bad, build it good, build it fast, build it slow. But build.

Intro

Not so long ago a junior dev (rare as they are) asked me "How would you approach learning including best practices?"

My first thought was it's a great question for a start and likely isn't to dissimilar to what I was asking when I was a junior and depending on who you ask you will get an entirely different answer. Especially now at the time of writing this (AI uncertainty and all that)

I will try my best to answer by explaining the way I learnt and I think it still works today, arguably it might be better with the chainsaws we have at our disposal today instead of the hand axes I was using. It's pretty simple really. Build things.

Best practice

Best practice is changing all the time, there's some basic givens like if you name things sensibly it will make it easier to read but aside from that and the legal implications of processing data now its basically a free for all. Which is both amazing and terrifying especially as a junior. Be wary of people that say this is the best way without at least entertaining the idea that there's perhaps another way. They tend to be zealots and you'll find yourself reciting scripture before long.

Build stuff

I would start by thinking of basic idea or something you might find cool to build, perhaps it's a website (that'd be my go to) an app (there's a bit of up front cost there so I'd go web first) but something that makes you wanna work on it and then just try and build it. Don't worry too much about if it's been done before or how viable it might be in a marketplace. The real value of building whatever you decide to build will be all the little hurdles you learn how to climb along the way.

Think less about the right way to do something, the wrong way will become clear quickly as you have to overcome different issues, and just figure out any way you can have it work and do that. It might mean "work" in spare time or using personal development time but I there's no quicker way to learn than by doing.

There's acres of free tiers of pretty much every kind of service you can imagine these days so I'd say hell with AWS unless that's something that really interests you because it's a mess of overcomplicated UIs and options built for companies managing many employees.

AI and mentors

Use AI sparingly, it may be tempting to get it to whip something up real quick but it will be much more beneficial long term to climb that hill by yourself. It can however be used to explain snippets understand the difference between two approaches etc it's very good at looking at what you have and finding rationale so use it like a documentation finder powered by 2000gb of RAM.

Apart from that, the finding someone to be a mentor, bare in mind you can even explicitly ask if they will be a mentor, is a great way of getting someone that will help you understand concepts and explain things in depth. It will likely benefit them as well since explaining something is a great way of ensuring you understand it.