This is 100% the key. I learned to code in my 30s, but I had been trying to learn for a dozen years before that and nothing stuck. Until I had a product I absolutely wanted (or had) to build, it was impossible for me to have the motivation to learn and have the concepts stick. But once I had that product in mind, everything fell into place.

I like this nice list for finding those sorts of tutorials https://github.com/tuvtran/project-based-learning

Yep. I can't even count how many times I've lost steam trying to learn something by doing tutorial projects where you're making lists of vegetables and otherwise building nonsense. It's like trying to learn guitar and being given "Mary had a Little Lamb" instead of some melody you actually like.

This.

I developed a .Net web application for a company 20 years that is still going strong (better than ever and now with multiple developers). I didn’t know how to code back then but bought two good books and their concepts are the basis for almost everything in the app.

Now I am trying to learn Python/Django. I have tried doing the Udemy courses and going through building a “blog” and “real estate” apps. They are fine but boring and I lost interest. Then I remembered what I did 20 years ago and am now building a real product based on what I am learning in a couple of good books. It couldn’t be going better.

This is my advice to anyone learning to code or learning a new tool. If you cannot build something that interests you, you're going to have a hard time building something for a paycheck.

I've started learning guitar more seriously, and I'm convinced this is the fastest way to learn. Set a semi-ambitious goal of something you can show others, and use deliberate practice to get there.

It'll take longer than you want, but you'll get better much faster than you think you can.

I agree. I learned to code in my 20s but I felt real confidence only after building multiple products that have real paying customers or products that other web developers use to build their products with paying customers. After building a dozen products, I believe that I can learn and build any feature within technically possible. This was not the case after coding parts of application or going through tutorials, but became the case after building entire applications from scratch for real customers.

EDIT: typos

> follow it TO BUILD A REAL USABLE PRODUCT

And what would be a "real usable product" for you in that context? It's a genuine question, since I'm someone who would like to build engaging online courses.

I do have a Python course online although it teaches concepts from small examples..

A real usable product consists of many things you need to do for it to be deliverable and usable.

Hobby projects often don't take much account into production usage, somethings we called it "best practices".

In short, imagine your code is in real world usage, which has real user, instead of just "small examples" or project to learn only.

It's about mindset here that settle us to be really serious at what we're coding and thinking.

No joke, but testing in production is one of the braviest ways to learn, too.

> In short, imagine your code is in real world usage, which has real user, instead of just "small examples" or project to learn only.

So, for instance, something like starting from an idea (a "webforum for all my friends", or other usable example), designing the database, implementing authentication, implementing the frontend (html, css, js, etc.), then adding those small things and finally deploying to a webserver?

The thing I think I'm asking is what kind of topics would be good to motivate people to follow through? For instance, the webforum above could work for some while others would have other ideas. It would be hard to design a course around a moving target like that..

I hope I'm not boring you with my questions above, I'm mostly trying to understand your point because I've been thinking about designing interesting courses for people, but I wouldn't like to do another basic introductory course with small chapters and exercises as other 1000 courses out there..

I got it.

The point is, not trying to add too much at a time. Try adding one feature at a time, but make it production ready, stable, reliable. It's what make progress !

That's probably the best advice. I'd add to it: make sure it's something your passionate about. Bonus if you can get real users.

Most of the time, you'll surely get the goal if you're passionate about it ;)

Out of curiosity, why do you believe that the learning process is more pronounced with making a real usable product rather than a fun project ?

Because when you make a real usable product, you're not doing just for fun. You're making product for others to use and to be happy. It's the mindset that make them different.