My Latest Work In the Wild
In 2020, I'm building a
browser-based IDE.
I would have never expected to add "developed an IDE" to my resume.
The problem space has been vast and incredibly interesting, as well as challenging.
Is it presumptuous for us to build our own IDE?
Isn't this a problem that's been solved many times over?
Do I want to solve this problem at all???
I don't think so; yes; I don't know.
Here's what I do know.
I yearn to learn, I love solving problems, and I prefer to do both with a team of talented people.
And I've got that.
So, here's to the IDE and dbt Cloud.
Four years ago, I needed to store React component composition in a database.
The resulting source code was short – gist-able.
But there was more than source code!
I created a sandbox, wrote tests, and shared the code within my organization in confidence.
I figured it'd be valuable to other developers, so it was open-sourced.
A few issues came in, suggesting to expand its functionality.
I supported these ideas.
We could add more mappings by resolving strings to expressions.
We could add JSON validation to prevent parsing from breaking.
We could support middleware for manipulating JSON and components during parsing.
We could make the process of constructing React components from JSON better!
I rewrote react-json-schema.
Twice.
I never pushed the code.
I left issues and pull requests open for a year, frusterated with not having a "right" answer.
An answer that I was willing to share with my digital peers.
It took me a long time to realize react-json-schema did not need to be improved.
Validation and manipulation could happen outside of react-json-schema.
What react-json-schema needs is more of why it was open-sourced and not just posted as a gist.
So, what would I like to see for react-json-schema?
Better documentation, including FAQs.
Improved testing to make experimentation accessible.
Automation to make maintenance less of a chore.
Continue to serve as insipration.
Not more features.
