text stringlengths 0 1.36k |
|---|
[682.96 --> 683.52] Yeah. |
[684.78 --> 687.32] Are we, are we done talking about web components? |
[688.46 --> 692.74] So there was a question, uh, in the chat, uh, did Google start this? |
[693.70 --> 702.56] Um, I think that while the, the Google people, uh, I think were the ones, this is all came out of the web manifesto. |
[702.64 --> 706.76] Um, I, I believe, uh, so that was a large amount of Google people. |
[706.76 --> 712.66] And I think the core authors of the specification, uh, started Google, but it is not a Google only thing. |
[712.72 --> 719.38] It is a specification in the W3C that has passed and is real and, uh, is in, uh, multiple browsers and things. |
[720.04 --> 724.14] So I think it is of Google, but not solely by Google. |
[725.08 --> 725.32] Right, right. |
[725.36 --> 732.34] I think you, you, you address the polymer thing where people tend to conflate this with polymer and polymer is a Google thing, uh, like very directly. |
[732.34 --> 734.10] Um, but this is much larger. |
[734.94 --> 744.36] I think one of the core problems there specifically was in the beginning, no browser implemented web components, but you could effectively use them if you used polymer. |
[744.56 --> 748.62] And so for a while, the only way to use web components was with polymer. |
[748.62 --> 756.92] And, and I think that kind of history caused this conflation versus, uh, other similar situations that that didn't happen in. |
[757.78 --> 757.80] Yeah. |
[757.90 --> 766.40] That, that's kind of a funny thing though, when you think about it, because one of the big benefits that they continue to talk about is that you don't need a bunch of JavaScript in order to do this. |
[766.40 --> 768.02] Like you don't need this giant library. |
[768.14 --> 769.46] That's, that's the benefit. |
[769.46 --> 773.78] And then people are conflating it with this giant library to do it before it was in the spec. |
[773.78 --> 774.06] Yeah. |
[774.86 --> 777.34] I think it is pretty fundamentally different though. |
[777.54 --> 783.92] Um, a, the size of polymer is pretty different than the size of, uh, say Amber or, or similar things. |
[783.92 --> 791.02] Uh, but also you can get initial renders and like working things before you have like full polymer execution. |
[791.02 --> 796.64] Like you can, you can see the page because it's CSS and HTML and the JavaScript hasn't executed yet. |
[796.76 --> 801.62] And, and that is, I think pretty fundamentally different thing than the other stuff. |
[801.62 --> 809.54] Well, do you think then that, like we talked a bit about react and Ember kind of eventually moving their implementations towards using web components. |
[809.54 --> 814.42] Do you think that this is going to be something that we, we just changed the way that we build our tools on top of them? |
[814.42 --> 817.70] Or is it actually a new enough model that it's going to change the tools that we build? |
[817.70 --> 821.64] Like, are we going to build very different tools and react and all of that in order to take advantage of this? |
[822.52 --> 822.58] Yeah. |
[822.66 --> 831.48] I definitely don't think that you could just take the current react and then like throw web components over the top of it, but you could take a very similar, like react. |
[831.62 --> 834.70] Uh, whatever we're on, uh, 16 or whatever. |
[834.88 --> 836.32] No, they have 16th react. |
[836.74 --> 837.68] Call it just 20. |
[837.80 --> 846.70] Um, could like, they could theoretically change a bunch of the API and then be outputting different things at the build stuff, but it would be a pretty huge leap. |
[846.70 --> 848.42] I wouldn't actually expect it to really happen. |
[848.42 --> 857.78] It would be more like some new person says, okay, the initial renders for web components are insane, but I don't like writing raw web components. |
[857.78 --> 860.94] It's here's this very reactive model that can do these things. |
[860.94 --> 873.90] Um, uh, one of the fundamental things I think, uh, that, that web components adds, uh, is the ability to do some of the, uh, data binding that, uh, some of these libraries do via dom diffing and re-rendering every time. |
[873.90 --> 878.56] Uh, so I think that is actually another interesting reason to use it. |
[878.72 --> 890.42] Um, not necessarily like a killer because a lot of that's very fast and can get faster and all sorts of that stuff, but it's certainly an interesting thing where you can kind of bind to two sections together. |
[890.42 --> 897.22] You can bind, uh, properties on attributes of the, uh, web component to the inside of the web component. |
[897.66 --> 908.60] But yeah, I wouldn't expect that really react or Ember ends up with like a web component version, but someone would do the react for web components and it's called we act or whatever. |
[908.90 --> 911.46] And that becomes a cool popular thing. |
[912.42 --> 912.94] Interesting. |
[913.18 --> 914.04] Very interesting. |
[914.04 --> 931.20] I'm trying to play out in my head, like how, how much of the web affects this, like in the future is like the way that, um, say if you use Stripe for instance, or, uh, I was using like the Tito embed the other day, like you get this JavaScript include. |
[931.76 --> 935.02] And, um, and then you kind of use like this custom element. |
[935.02 --> 942.08] And right now it has to do like all this crazy stuff to like find that element and, and do a bunch of stuff kind of after load. |
[942.08 --> 948.44] Is it really going to change the model of how that kind of stuff is implemented where when you're like, Hey, you know, include my custom element in your page. |
[948.74 --> 951.84] Is it, is it going to work like really differently and a lot smoother than it does today? |
[952.56 --> 958.56] Yeah, I think, um, there, there are HTML imports, which I don't know if I have made it in, uh, to browsers yet. |
[958.56 --> 964.18] And there, there are a few things that, that make a lot of those things really cool. |
[964.46 --> 972.06] Um, so I, I have implemented a long time ago, uh, the, the Stripe JS, uh, credit card form as a web component just internally to try it out. |
[972.08 --> 985.28] And like the amount of work that I have to do to style safely and do all the third party JavaScript things in the current world versus the web component world is pretty vastly different. |
[985.28 --> 995.10] Uh, and, and the speed at which our component can kind of render and then be attached versus execute the JavaScript and then be injected, um, is also pretty different. |
[995.10 --> 1003.24] And if we know one thing about the performance of checkout pages is that like everybody who's ever tested is like, this matters a lot. |
[1003.82 --> 1010.46] So, um, I think it could be a pretty good fundamental change, uh, in the direction of rendering. |
[1010.46 --> 1020.70] And, and I think that that's what a lot of, like a lot of the cool wins are the modularity and the composability and the scoping and all those things that we've had trouble with on the web whenever you're building a large application. |
[1020.88 --> 1025.08] And I think those will be the things that people think about more than, than some of this stuff. |
[1025.40 --> 1032.52] But, um, the kind of the fundamental turn is that things can render and exist prior to JavaScript executing. |
[1032.52 --> 1049.10] And, and so the, uh, server side rendering isomorphic stuff changes in the way, like you don't necessarily need to do rehydration, uh, as much as you can do, uh, just like render things as web components. |
[1049.10 --> 1051.22] And then the JavaScript can, can kind of run after. |
[1052.20 --> 1053.04] What is rehydration? |
[1053.16 --> 1054.42] You just ran right over that one. |
[1055.14 --> 1055.94] Oh, sorry. |
[1056.06 --> 1058.54] Uh, rehydration and a server side render. |
[1058.54 --> 1068.28] So, so like the competition to web components in the, in the, uh, world of frameworks these days is that you can get no JS to render your entire page. |
[1068.58 --> 1073.74] Uh, and as long as that is a deterministic output, uh, the render is a deterministic output. |
[1073.74 --> 1077.66] You can, you know, like to HTML it and then serve it as the initial load. |
[1077.66 --> 1083.10] And so, uh, no JavaScript has run and it's just CSS and HTML and you can see the entire page. |
[1083.16 --> 1086.24] And I don't know if it works yet, but, uh, you can see the entire page. |
[1086.24 --> 1096.96] And then, uh, because that same render function can then run once the JavaScript has executed, it can come up with the exact same deterministic DOM that Node.js did. |
[1096.96 --> 1114.94] And instead of killing the whole page and then re-rendering it with the, the client side JavaScript, it can just kind of attach itself to the server side rendered thing and say, we claim these, uh, elements as the ones that we would have rendered had they not already existed. |
[1114.94 --> 1119.28] Kind of like a re-render that, uh, that occurs in, in React all the time. |
[1119.36 --> 1126.26] It's, it's kind of, uh, a basic property of React is that if you try to render something and all of it's still there, it's a no op. |
[1126.32 --> 1127.26] It's kind of the DOM diff. |
[1127.34 --> 1134.38] It's what's the diff between this virtual DOM that we created based on all the data and the one in the actual, uh, window. |
[1134.38 --> 1142.38] And if there's no difference, we won't do anything, but we'll kind of know that all these things are attached to like all of our handlers and stuff like that. |
[1142.58 --> 1145.06] So, uh, that's what rehydration is. |
[1145.12 --> 1150.66] It says, um, we can just attach ourselves to a server side rendered page without re-rendering it. |
[1150.70 --> 1161.52] And that is a pretty good, like if you, if you need speed, if you're a content website, especially you need speed and SEO and all that stuff, you should absolutely be doing a server side rendering with, with, uh, rehydration. |
[1161.52 --> 1166.16] Well, you just mentioned SEO, which means it's time for a break and we get off this topic. |
[1166.54 --> 1169.62] So, so we're going to take a quick break. |
[1169.76 --> 1171.48] Uh, when we come back, we're going to talk about conferences. |
[1172.84 --> 1178.64] First sponsor of the show today is our friends at Century, helping you to find and fix your errors in your applications. |
[1179.18 --> 1180.74] You can start tracking your errors today. |
[1180.84 --> 1181.58] Totally free. |
[1181.84 --> 1187.16] They support React, Angular, Ember, Vue, Backbone, and Node 3 more like Express and Koa. |
[1187.16 --> 1198.14] You can view actual code and stack traces, including support for source maps, see the errors URL, parameters, and session information, and even prompt your users for feedback when you have front end errors. |
[1198.48 --> 1200.30] Head to changelaw.com slash century. |
[1200.50 --> 1202.00] Start tracking your errors today for free. |
[1202.34 --> 1203.28] No credit card required. |
[1203.56 --> 1204.86] Get off the ground with their free plan. |
[1205.16 --> 1207.72] And when you're ready to expand your usage, simply pay as you go. |
[1208.10 --> 1210.58] Once again, changelaw.com slash century. |
[1210.86 --> 1211.96] And now back to the show. |
[1211.96 --> 1216.06] Now we're going to get into conferences a little bit. |
[1216.40 --> 1220.34] So, um, JavaScript has an amazing conference scene. |
[1220.56 --> 1223.66] There's a million, uh, little community conferences out there. |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.