Short excerpt from a YouTube conversation (Dennis H. Lewis and another speaker) comparing no-code, low-code, and pro-code approaches. Key points: no-code aims to build without writing code (example: Bubble's visual engine where generated JavaScript is hidden), low-code sits in the middle where tools may generate editable code (including some AI tools), and editable low-code can let developers tinker but also risk breaking things.
No. Code, on the other hand, is essentially trying to get as much of your development done without actually writing a single line of code. Right.
That's what we do using Bubble's visual development engine.
If we get to the middle ground and we find low code, what do people typically mean with low code?
Well, to me, I think a lot of those AI tools fall into the low code basket because you're not actually developing, but it is creating actual code that you can later adjust and change. So I think that's kind of the middle ground. With Bubble, you do not there while it is creating JavaScript. That JavaScript you never see, it's never under your, you're never touching it, you can't actually, you know, it's part of the Bubble infrastructure. So with, with.
Low code, you can actually meddle with the code. You can actually see the code go in, make changes.
Which could be good or could be bad.
I was going to say it sounds good to be able to have that option, but it's also the way typically people end up.
Screwing stuff up.
Screwing stuff up. Building Frankensteins that they can't fix when they do screw up stuff.
Exactly. And one of the bad parts of that low code model is that once you've made changes, you often can't go back to the automatic part because now the code base is its own, it's been changed.