☕ 5 min read
I use VS Code every day. I use it exclusively since February 2018, both personally and professionally. I like it a lot.
But I am frustrated.
I am frustrated because I miss a great tool to perform automated refactorings for me. And neither VS Code, neither existing extensions provide the experience I’m looking for!
Thus, 2 months ago, I decided to follow my own advice and hack my own problem. I started creating the VS Code extension I was looking for. This extension analyzes my code, suggests refactorings and executes them with the minimal amount of keystrokes needed.
It helps me focus on my intention instead of manually following the scripted mechanics of a refactoring.
It’s true that VS Code ships with basic refactoring operations. So what’s wrong?
Well, speaking from a personal point of view, few things aren’t great:
Before using VS Code exclusively, I used Webstorm professionally for a few years. And I prefer VS Code: it’s free, feels lighter to use and has a very active community.
But there is one thing Webstorm nailed that VS Code doesn’t provide: many, intuitive, automated refactorings.
This made me so productive back in the days. I relied on my editor to move fast 🏇
And I think it matters. If refactoring code takes 2 keystrokes I’m very likely to do it. If I have to find the correct combination of 5 keystrokes to make it work, I’m less likely. If I have to do it by hand… You get the idea.
Great automated refactorings are a productivity boost for every developer.
Yes! And I used it for some time.
I appreciate the good efforts that were put in it. At least someone worked to provide more refactorings, shared that for free and maintained it 👏👏
But at this point, it was not exactly what I needed, for a few reasons:
At that point, the best move for me would have been:
OK, maybe I can contribute and improve it!
Heading to the repository, I was really happy to see it seemed well tested 👍
But I felt the architecture was not quite what I’d have in mind. I’d have done abstractions differently…
Now, I’m not saying this is bad and I know better. Actually, I think the code is quite clean and well tested. But I started imagining how I’d build such solution, and I had this desire to make it happen. In fact, it would be a great opportunity for me to work on something where I could try things, fast, to scratch my own itch.
Sure, building another extension is not the best idea, community wise. For the greater good, I should have reached the author of the existing extension and see how could I help. And, at some point, I’ll probably ping him to see how I could help, now that I’ve a clearer idea of how all of this could work.
But for the moment, I wanted to build something on my own. Not to replace the existing extension, but to try something different. Something aligned with what I needed. Without having to deal with existing code and users.
That’s why, 2 months ago, I started to build this extension.
My goal is essentially to build the tool I’m missing.
There is a very cool side-effect of this: the learning part. Digging into Abstract Syntax Trees manipulations is super interesting. Also, I’m discovering how to build a VS Code extension for the first time. Finally, it’s a playground for me to test architectural decisions and practices, like documenting architectural decisions through ADRs.
I implemented a couple of refactorings across these 2 months. I published the extension a few days ago on the Marketplace.
Now I’m waiting for feedback to improve these first refactorings, fix edge cases and improve the UX to match what I’d expect from a great extension!
Then, if you want to help me, you can:
That’s it. I’m really excited about this and happy to start spreading the word.
Now, I feel better.