I've been stuck thinking about the right way of writing code for 5 years now, still stuck in the beginner phase.

I really really needed to hear this. Thank you.

Should we learn some other things besides coding ? like robotics or control system, math perhaps,... to be useful to humanity ?

You are the best man keep producing

For a 2 min video, you give the answer in first frame and your advices have been excellent Thank you!

One thing I only learnt recently after a 13 year engineering career, 10 of those coding, was to forget everything else and just write your idea into code. It sounds so obvious, but it's also easy to fall into the trap of trying to write "professional" code on the first go. Get it working, and then refactor.

This. Is. Amazing. I literally put off my project for a month because I decided it was just 'sphagetti code' and wanted to quit. But how else am I gonna get my first major project done in general? 💀 Genuinely kind of inspired me.

Thanks, I needed this

I love this feeling, when the code is the most horrible piece of anything created by mankind, but it works and the result is actually beautiful. By doing that you gain immeasurable amount of knowledge. Humility is so important, being okay with not knowing everything.

You don’t like Rick & Morty?

this guy really is bigbox

You sir, are a gem. I definitely needed to hear this.

This is very vague

Undertale dialog system is a 10k+ line switch statement, there's people who follow best practices and then there's people who ship

Some time ago I tried to comply with different code conventions, even tried to follow all the SOLID rules. But trust me guys, if you have a short deadline to make a solution you'll probably forget about 90% of that stuff. As the author of the video said, don't be ashamed to write an ugly code, but also don't forget to rethink it afterwards

nvim, emacs, and especially vscode are all bloated soyware. real programmers use '$ cat > file'

Sounds like the red-green-refactor talk, go from failing test to a full suit of passing tests as fast as possible, refactor the shitty code and don't touch the tests a second time.

No way, did almost the same thing, but instead of experiments, I call them concepts

Tell me you're bad at coding without telling me you're bad at coding.

Yep. Always write your code at least twice.

wait, why saying `bigbok` in the end?

I've been doing this for like ~1.5 years, I call them PoCs but it's the same idea. 10/10 would recommend to a friend.

Agree with your advice. Use git as your power tool. You want to validate your idea? no worries just create a new branch and start working on it. You can break things without worrying about it too much always knowing you have working code in a different branch. Once I had 6 different rails projects for the same app because I was constantly changing db schemas and it was too much work to revert changes and roll back. Instead I created diff projects based on elements name from periodic table. Aluminum, Carbon etc..

Your best advice so far. That was my biggest problem

thx, but i am not sure yet. I will watch another tutorial about this topic so in realy understand how to escape

Thank god I just know that my code is absolute dogshit no matter what because I'll look back at my old code from a year ago, shake my head and call myself an absolute dumbass, then move on, knowing the code I write now will be judged the same way in a year by myself

Don't bother me at all because I'm a data scientist working on separate tasks (like prepare data or train a model, or test it) and we don't code review because if metrics are good, their are good, of they're not, well, it's your problem :) Except for shared tools but I didn't like a data prep tool we wrote so I partly rewrote it myself.

Create an environment conducive to your objective. Minimize external distractions, like sound and interruptions. Minimize internal disruptions like positional pain and fatigue; manage your water and bowel movements through regularity. If you have a bad back, you are going to want a reclining setup with multiple monitors, Master your tools. Limit your research related tangents to less than 2 minutes or file them as optimization projects, block out some time and create tangible products like code snippets that can be used across projects. Understand the intent of the code. Master secure coding for your language. Master design patterns. Master architectural patterns. Improve your typing speed with particular emphasis on navigating the file system, adding comments, cutting and pasting, using code completion and shortcuts. Practice system administration basics: cyber hygiene, system backups, configuration backups, cybersecurity through system isolation. Choose tools that are likely to be around for a long time. Use templates and automation wherever possible.

big bog

top_p = 0.99 Do_sample=True

I wanted to start game development using godot, I was following Brackey's new tutorial about it and other tutorials with no results, but now I just started a new project and started cooking... I am at the end of this day already made a player and enemy patrolling without any tutorials

I kinda needed that. thank you

Bro literally described my life

So,.. You realy chose to use Ubuntu Font for a really good video!?

just get things done first

i think bro is indian/pakistani

I'm currently on a startup business, and since I'm on my own for everything, the "experiment branch" method is what got me pass my "block" on code, I'm amazed you described it exactly as I do it too. Cheers!

Bro is Big Bo

Well i guess they didn't sponsor this video😢

This is also called Proof Of Concept in tech world, very useful when you are solving a unique and challenging problem. Pro tip: when you are done with POC - add some unit/integration tests on top before refactoring - in this way you can always be sure that you did not break anything important during your refactoring and you can go crazy.

not starting/doing it because i want to start/do it in the best way, and i didn't find it yet what an evil trap

That is EXACTLY what I was doing for the past 1,5 year. I always thought it's me that is the problem, I'm not good enough, I'm writing terrible code, I don't know how to structure this, what method would be the best, what database would be the best for my use case, should I go serverless or docker, REST or tRPC, node or bun. I will FINALLY try to write. some. code.

so it's like prototyping?...

Good idea mate

Mate, is easy for me to write code that writing some scientific writing. What is wrong with me? I can't even make a scientifict product by my experiment yet 😭

Remember folks, "Premature optimization is the root of all evil"

also stop using vscode

I liked this video first try.

After 2 decades of programming I have a simple motto : « Code first, think and refactor after » And don't be afraid to trash all previous code when refactoring. You're not writing the 9th symphony, it just has to pass the same unit tests suite

thanks, it was helpfull