Django Was Just the Excuse

This weekend I stood in front of a room of developers to give a talk that was advertised as being about backend development with Django.

The funny thing is, I planned on talking about something most beginner backend developers often overlook.

I begun by asking a rather unexpected question: "As an aspiring backend dev, what will you be doing in a real world scenario, what would your typical day in the life look like?" The answers I got was just as I expected, and that right there was what I wanted to bring to light in my talk.

Somewhere along the way, many of us start believing that learning backend development means memorizing a framework. We learn models, views, serializers, authentication, migrations, and enough commands to build an application that works.

But eventually you realize that none of those things answer the questions that actually matter. How do you decide where business logic belongs? What happens when two people try to buy the last item at exactly the same time? What if a payment succeeds but your server crashes before saving the transaction? What if a user keeps clicking the payment button because nothing seems to be happening? Those aren't Django questions. They're backend engineering questions. That's what I wanted my audience to leave with.

Throughout the session, we explored the idea that backend development is really about making decisions. Every endpoint represents a business process. Every database table represents something that exists in the real world. Every bug is usually an assumption you forgot to question.

One idea kept coming back throughout the talk:

*Frameworks help you write software. Thinking helps you build systems.
*
Preparing for the session also taught me something about myself. Building projects is one way to learn. Explaining those projects to other people is another level entirely. It forces you to strip away jargon and ask whether you actually understand the ideas underneath the code.

Walking off that stage, I was happy because the talk went well. I was also happy because it confirmed something I've been working toward over the past year.

I don't just want to write backend software.

I want to understand how great backend systems are designed.

And then help other developers think that way too.

Here's to more code, more architecture diagrams scribbled on whiteboards, and hopefully many more conversations about why the most important question in backend development isn't "How do I build this?" but "What happens when this breaks?"