text stringlengths 0 2.02k |
|---|
[48.44 --> 60.94] This episode is brought to you by our friends at Influx Data and their upcoming IoT application training. |
[60.94 --> 66.64] You'll build a fully functional sample application called IoT Center and build it on InfluxDB. |
[67.12 --> 71.64] With functions such as the new IoT device registration and IoT device emulation, |
[71.64 --> 76.02] this demo app is based on Node.js and React, uses Kafka to write measurements, |
[76.42 --> 79.78] and requires no dB installation with InfluxDB Cloud. |
[80.22 --> 86.60] Space is limited and it's free to join, so register now at influxdata.com slash changelog to save your spot. |
[87.02 --> 90.38] Again, that's influxdata.com slash changelog. |
[90.38 --> 110.62] Hello, we're here with Ryan Dahl, creator of Node, creator of Dino. |
[110.84 --> 114.08] Ryan, we're big fans of you, and we're so happy that you're here on the show. |
[114.58 --> 115.92] Hello. Glad to be here. |
[116.22 --> 117.12] I think this is a first. |
[117.76 --> 118.12] Yes. |
[118.12 --> 121.16] You've never been on the changelog before, like right, even back in the day. |
[121.50 --> 121.70] Nope. |
[122.10 --> 122.30] Never. |
[122.38 --> 125.10] We've only talked about Node, not about the creator of Node. |
[125.44 --> 128.88] Yeah, we've talked about Node, but we haven't talked about the creator of Node. |
[129.30 --> 130.72] What a shame. Here we are to fix that. |
[131.14 --> 133.26] Well, 10 years later, or ish. |
[134.10 --> 134.46] Welcome. |
[134.92 --> 135.24] Thank you. |
[135.64 --> 139.10] Yeah, I think this is only the second podcast ever I've done. |
[139.50 --> 139.82] Oh, wow. |
[140.10 --> 142.76] Is that right? Well, thank you for giving us part two of that. |
[143.04 --> 143.74] That's right. We're honored. |
[144.30 --> 147.04] So, you're not here to talk about Node. You're here to talk about Dino. |
[147.04 --> 151.10] You've rearranged the letters. You've been hard at work for a couple of years now. |
[152.02 --> 157.36] And I would like to go back to your JSConf EU talk, which has become famous now in 2018, |
[157.68 --> 162.94] the 10 things that you regret about Node.js, which turned out not being 10, it seems like. |
[163.30 --> 166.14] I don't know if you even named it that or they named it that on the YouTube file, |
[166.70 --> 171.72] because it was like 7, and then you introduced Dino, and I was going through it trying to find it. |
[171.72 --> 174.14] They named it that, yeah, it was not 10. |
[174.70 --> 176.72] Yeah, that's just the way they named it on YouTube. |
[176.98 --> 178.42] Like, I don't even think your slides said that. |
[178.42 --> 181.04] They're like 1, 2, 3, 4, 5, 6, 7. That's close enough to 10. We'll just call it 10. |
[181.04 --> 186.36] Yeah, round numbers and lists apparently work, you know, on the internet. So, somebody threw 10 on it. |
[186.46 --> 190.32] Yeah, if you go to the titles, if you like look at that talk and see the title slide, |
[190.32 --> 194.62] it has a completely different title. It's not 10 things I regret about Node. |
[195.30 --> 200.38] Well, good, because you had 7 of them. And it seems like that was a talk that was put together somewhat last minute, |
[200.46 --> 203.22] maybe because Dino wasn't quite as far along as you hoped it would be. |
[203.30 --> 205.76] Or what was the backstory on that? Because even in the beginning, you said, |
[205.88 --> 207.96] this wasn't my talk, but now it is my talk. |
[207.96 --> 212.58] Was that the thing, is Dino wasn't quite ready to give more of a fleshed out version of it? |
[213.26 --> 218.70] Yeah, I was working on a different project, and I had applied to give this talk on this other project. |
[218.70 --> 226.08] It was something very similar to TensorFlow.js. So, it was TensorFlow binding to Node, |
[226.24 --> 233.64] so that you could train machine learning models in Node. Plus, something that's like Python's |
[233.64 --> 240.28] Jupyter Notebook, or in JavaScript, there's this Observable company, like a website where you could |
[240.28 --> 247.42] kind of insert a bit of code and plot functions and stuff. And yeah, TensorFlow.js and Observable |
[247.42 --> 251.86] actually came out just before that talk. And a month before that, we decided, like, |
[251.94 --> 259.04] actually, these projects are kind of dead end, or we should do something else. And JSConf would |
[259.04 --> 263.90] insisted that I continue to give a talk on something else. So, I put together this Dino demo. |
[264.38 --> 268.86] Nice. So, there's a bunch of stuff about Node. Anytime you make design decisions, you know, |
[268.90 --> 274.48] in retrospect, it's easy to look back. And it's hard not to find flaws in things that we create. |
[274.48 --> 278.84] But we don't always get a second shot at it, you know, and Dino seems to be, to a certain degree, |
[278.92 --> 283.66] a second shot, or at least another round of something different, but similar. And I'm just |
[283.66 --> 288.88] curious of the seven or so things there that you listed, which if you don't remember, they're |
[288.88 --> 293.74] not sticking with promises, the security system, which I think you definitely are addressing with |
[293.74 --> 298.92] Dino, the build system, package JSON, Node modules. Those are a few things, just high level, |
[298.92 --> 303.66] that you mentioned, that were regrets. How many of those do you feel like you've |
[303.66 --> 307.50] alleviated with Dino? How many things did you fix? |
[308.40 --> 314.56] I mean, all of those things, I think. You know, generally, Dino is an effort to |
[314.56 --> 323.74] move the server-side JavaScript platform forward in kind of radical steps. Node has a lot of users, |
[323.74 --> 330.66] of course, and they try very hard to maintain backwards compatibility, which is laudable and |
[330.66 --> 338.44] great for all of the users of it. But that means that it's very slow in taking on new changes. And, |
[338.54 --> 343.30] you know, I think the major change in JavaScript over the last couple of years, maybe two major |
[343.30 --> 352.08] changes are async await syntax and the ES modules syntax. These are pretty radical changes to the |
[352.08 --> 357.28] language. And, you know, JavaScript is like this evolving system over time. The JavaScript of today |
[357.28 --> 364.48] is nothing like the JavaScript of 15 years ago. And yeah, it just felt like Node was not keeping |
[364.48 --> 368.44] pace with web browsers in terms of bringing on these new features. |
[369.14 --> 374.54] So now you're a few years into Dino. We've hit 1.0 last year. Lots going on, lots formalizing. I'm |
[374.54 --> 379.02] wondering now, since you've gotten into the weeds for a couple of years, you look back, |
[379.02 --> 384.22] have you made any mistakes so far? You had seven or so with Node that you regretted. Any regrets with |
[384.22 --> 384.64] Dino yet? |
[385.22 --> 391.18] Yeah, definitely. I mean, it's hard to make software and not make mistakes, of course. |
[391.30 --> 391.48] Totally. |
[392.00 --> 397.36] I mean, this is going to be very controversial, but the TypeScript aspect of Dino, Dino has |
[397.36 --> 403.38] TypeScript compiled into it, is super nice, very friendly, very nice to be able to just get up and |
[403.38 --> 411.22] started easily. But browsers do not support TypeScript. And Dino's overarching philosophy is that |
[411.22 --> 419.30] is bringing server-side JavaScript closer to browser JavaScript. And in this aspect of supporting |
[419.30 --> 427.30] TypeScript out of the box, we are kind of overstepping our goals. So, you know, we're supporting this |
[427.30 --> 433.68] extension to JavaScript. I mean, I don't know about you all, but I feel like TypeScript is eventually |
[433.68 --> 439.24] what JavaScript is going to be. I feel like it is kind of the next generation of JavaScript. And, |
[439.72 --> 443.22] you know, I'm very eager, and I think many people are, to kind of go in that direction. |
[443.74 --> 452.64] But it's a big complication to our system to support TypeScript. And I think weakens our argument that |
[452.64 --> 459.94] Dino is web-compatible. So, we're certainly not removing it now because Dino depends on it. |
[460.18 --> 462.54] Right. Go back and start over, would you? Yeah. |
[462.88 --> 467.80] But, yeah, I think it would have been easier for us to get started had we not supported that out of |
[467.80 --> 468.82] the box initially. |
[469.38 --> 474.20] That being said, it seems like it was also a big draw and a big differentiator at first. |
[474.30 --> 474.66] Absolutely. |
[475.14 --> 480.64] Let me just say that I was around when you announced Node back in 2009 or whenever it was. I remember it |
[480.64 --> 486.94] right around the iPhone time. And I remember the reaction to Node. Probably not as well as you do, |
[487.06 --> 490.90] because you were on the project. I was just an innocent bystander. And maybe we remember it |
[490.90 --> 496.18] differently. But to me, it was like immediate. It was a hit. Everybody was excited. We had Michael |
[496.18 --> 500.28] Rogers just on the show a little bit back, and he was just reminiscing a little bit how like nothing |
[500.28 --> 505.12] existed for Node. Like when Node came out, you announced it. There was no ecosystem. There was |
[505.12 --> 510.56] nothing. And then like everyone just piled in. Like people saw fertile ground. And you're looking at me a |
[510.56 --> 514.88] little weird. Like maybe that's not how you remember it. But it seemed like people came in |
[514.88 --> 520.56] droves to fill out that ecosystem. It just was like immediately a hit. And I think it filled a real |
[520.56 --> 524.66] need. Well, first of all, let me stop and say, did you have that same feeling or am I seeing it |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.