Original ReactJS Core Team Manager on What It Was Like Building React & Web Dev Future (w/ JS Party)

progressive disclosure of complexity. This is a space that is very hot, a little weird and strange. All these panels and buttons and menus and all this stuff and it's like, do I need to know about all of this up front? Here's what I want to happen. The compiler takes care of all of that. Nobody wants to touch that. We never will find that button. Two reasons that React really took off. We don't build services to make money. We make money so we can keep building better services. Behold, this is the answer. We finally done it. You're doing the same thing over and

over across lots of different companies. It's very hard to change developer behavior. All right, we are here live at React Summit. I'm Kball from JS Party. I'm joined by my co-host here. I'm Nick Hoyo Hoy. Hoy hoy. And we have a special guest here today with us. Tom, why don't you introduce yourself? Hey everyone, I'm Tom Okino. Uh I was a part of the the founding team of React at Facebook Now Meta. Uh, and these days I'm the chief product officer at Verscell. Given that we are at React Summit, I'm sure we

have to go into the story. So tell us a little bit behind the scenes, you know, what was it like getting React started? Yeah. Well, how long do you want uh what version do you want? How long do you want me to speak for? I could go for 5 minutes or 5 hours. Well, I mean, you've told this story again and again. We got to start with it, but let's keep it tight and then we'll try to get into like what are some of the the behind thes scenes stuff that maybe hasn't come out before. Yeah, sounds good. So um in the early days I think we were exploring a number of different ways to build web applications and on the sort of ads side

of our business we had some more sophisticated applications than you know on the consumer side. You know the way that we were kind of building those at the time was the same way that everybody was building web apps was sort of a client side MVC. It was not you know dissimilar to something like Backbone or or Ember or Angular at the time. you know an engineer came along and he's like look this code is very hard to maintain as our team grows we're moving very slowly nobody wants to touch that thousandline model over there like you know only two people can touch that and so he thought there was a better way uh

he was inspired by a bunch of things that we had already been doing uh in other parts of the the business but from that a prototype of an early version of what would eventually become react was born at the time I think we didn't have the right home in the company inside of the ads business um to sort of like you know incubate that type of new technology. So Jordan came over and talked to some of us on the product infrastructure team where we built uh frameworks and technologies that enabled other developers to be more efficient and build higher quality stuff. We gave

it 5 minutes. We tried to to build some things with it. We knew there was something something important there. Um, and that's when we kind of started to double down and and you know, the rest you can you can hear about in the the documentary or in one of the other the other interviews. But yeah, it was very much a you know grew out of this organic sort of need this emergent need for a simpler way of thinking about and developing our apps. I'm curious uh in

ads before that uh were you using something that's familiar or was it more like an in-house solution? I'm just curious like what what drove the need to react? Yeah, it was something in-house, but it was not unfamiliar. I mean, one of the people that um co-created our in-house framework, which was called Bolt, was a core contributor at Backbone. Um, we also had uh, you know, folks from the the sort of like Dojo mobile team, like we were very much like in the soup of the the JavaScript

ecosystem at the time. And so, yeah, I would compare our our sort of in-house framework to a sort of like backbone with a a different way of doing the sort of view layer. Um, and it was good, very good. Um, and and we used it for a very long time. And it wasn't until our team started to get bigger and our app started to get sort of pathologically complex that we started to need a simpler model, no pun intended. Uh, we had these massive controllers and these massive models that like nobody really wanted to touch. So many inputs and um, you know, we couldn't make changes with