Datasets:
willtheorangeguy
/
2017-JS-Party-Transcripts

Dataset card Data Studio Files
xet
 Community
text
stringlengths
0
1.36k
[2447.22 --> 2448.20] It would generate code.
[2449.16 --> 2451.80] Like you would just say Rails new controller.
[2452.22 --> 2452.66] Scaffolding.
[2452.84 --> 2453.56] Yeah, scaffolding.
[2453.78 --> 2454.70] That's the word.
[2455.14 --> 2457.26] And it definitely feels kind of like scaffolding.
[2457.26 --> 2460.88] And there's a bit of scaffolding like in the initial like create React app.
[2461.48 --> 2467.00] But I think it focuses less on generating code for you and more about providing tools
[2467.00 --> 2470.48] and examples and kind of a baseline for you to build on.
[2470.64 --> 2475.92] And then allowing like the kind of one of the things of scaffolding is like once it generates
[2475.92 --> 2481.00] that code, that code is is stuck there forever in that format.
[2481.00 --> 2487.36] Whereas I think more of the strategy with create React app is that hopefully it scaffolds little
[2487.36 --> 2492.76] enough to where it can update those things that it has generated like on the fly.
[2493.36 --> 2501.72] I think the last thing that's interesting in the not Webpack, create React app 1.0 release
[2501.72 --> 2503.86] is the code splitting stuff.
[2504.10 --> 2506.36] And that's part of Webpack as well.
[2506.36 --> 2511.26] But there's a standard that no one uses for dynamic imports.
[2511.50 --> 2515.10] It mixes async await with import.
[2515.72 --> 2520.54] And I hadn't looked into it much because there wasn't really a great place to use it.
[2520.58 --> 2522.28] But it's like it's part of the standards track.
[2524.06 --> 2526.70] And like I don't know where it is in that.
[2526.82 --> 2531.24] But you can have an async function and then you can import something.
[2531.48 --> 2532.92] You can do await import.
[2532.92 --> 2538.16] And then that will automatically build into a separate like all the dependencies of the
[2538.16 --> 2544.12] thing that you're asynchronously importing can be built into a separate bundle.
[2544.82 --> 2547.26] Yeah, I think you're complicating it a little bit.
[2547.32 --> 2551.92] Like it's a piece of syntax that allows you to with a function do the same thing you do
[2551.92 --> 2553.36] with syntax for import.
[2553.56 --> 2553.72] Right.
[2553.94 --> 2558.52] So and the nice thing about that is that at some point in the future, which is like not
[2558.52 --> 2563.98] part of the initial interpretation phase of the browser, you can say import this module.
[2564.88 --> 2568.04] And then what you're saying is that like now we can actually use that for code splitting
[2568.04 --> 2571.78] because you can say, oh, well, like these these little pieces that you don't necessarily
[2571.78 --> 2575.56] need, we can now import dynamically using the same kind of module system.
[2575.68 --> 2575.74] Right.
[2575.80 --> 2579.60] I guess I was complicating it because it would be invalid syntax to just throw an import
[2579.60 --> 2580.86] there.
[2580.86 --> 2582.90] So it needs to be like supported syntax.
[2583.04 --> 2586.52] It's not just like something you could do before, but people didn't know about it.
[2586.52 --> 2592.64] I think is like awaiting an import is not like it needs to be statically analyzable or or
[2592.64 --> 2598.68] at least be known to be a part of it that isn't statically analyzable because it doesn't
[2598.68 --> 2600.32] need to be something, you know, something like that.
[2600.68 --> 2608.16] And that's why I think it's part of the standards track to do asynchronous imports like like this.
[2608.16 --> 2614.36] And so create react app supports this in order to do bundles, which is a huge part of like
[2614.36 --> 2616.88] the PWA communities problem.
[2617.06 --> 2621.24] Like if you follow Alex Russell or whatever, you'll you'll know that your JavaScript that
[2621.24 --> 2623.64] you're serving by default is far too large.
[2624.64 --> 2630.38] And so if you can do so, if you can turn on HTTP two and then do something like a handful
[2630.38 --> 2634.70] of these asynchronous imports for large portions of your application.
[2634.70 --> 2640.86] I think it could go a long way to like loading far less JavaScript on load, which is which
[2640.86 --> 2641.50] is really nifty.
[2641.64 --> 2647.24] I think this is such a good direction to like automatically for like give to people.
[2647.40 --> 2651.44] I hope they use it in the baseline example that they generate, you know, that way people
[2651.44 --> 2651.86] use it.
[2652.88 --> 2652.96] Yeah.
[2653.06 --> 2657.42] Sort of following on with your talk about scaffolding, it seems like the big difference between
[2657.42 --> 2661.76] this and what Rails does is like you said, Rails will generate a lot of boilerplate code.
[2661.76 --> 2664.86] This seems to generate a lot of configuration, right?
[2664.92 --> 2670.78] Like we have like like the joke about Webpack is that like you you only write one Webpack
[2670.78 --> 2673.46] configuration and then you copy paste it into every project.
[2673.68 --> 2675.48] I mean, that's a make file joke, but yeah.
[2675.90 --> 2676.22] Yeah.
[2676.32 --> 2676.46] Yeah.
[2676.62 --> 2677.02] Exactly.
[2677.72 --> 2677.92] Right.
[2678.72 --> 2683.86] But I think also like like you were saying, one of the things this does is really standardize,
[2683.96 --> 2689.02] you know, what is the the proper path for writing a React app with all these different
[2689.02 --> 2689.56] configurations.
[2689.56 --> 2695.06] And so this allows you to sort of add features over time to that configuration without trying
[2695.06 --> 2699.20] to get, you know, thousands and thousands of developers to update their, you know, this
[2699.20 --> 2700.88] particular line in their Webpack config.
[2701.38 --> 2701.50] Right.
[2701.98 --> 2703.34] It's a noble cause.
[2703.50 --> 2707.76] And other people are doing it like Ember CLI and stuff are doing this well as well.
[2707.84 --> 2716.36] Like when you upgrade these like new world configuration CLI tools, you get instant improvements in
[2716.36 --> 2717.72] your applications, which is cool.
[2717.72 --> 2722.74] I really like like everything still works and now it's 20 percent faster.
[2723.00 --> 2729.94] It's like whenever Ember did the Glimmer update, all you did was upgrade Ember CLI and suddenly
[2729.94 --> 2731.02] everything was using Glimmer.
[2731.18 --> 2735.78] It was all supported unless you're doing something weird, you know, and suddenly your website rendered
[2735.78 --> 2736.12] faster.
[2736.22 --> 2738.04] And I think that's a cool world.
[2738.04 --> 2745.88] Like for I think that's a good goal for these well-used frameworks to to to go after.
[2746.50 --> 2747.26] Yeah, definitely.
[2747.46 --> 2752.32] One thing I can't wait to see is not compiling down to ES5 anymore, but compiling down to,
[2752.32 --> 2756.16] you know, a set of features that are actually mostly supported because there's a lot of performance
[2756.16 --> 2760.10] benefits to like arrow functions are faster than regular functions and in V8.
[2760.10 --> 2764.00] And for the most part, you know, people that are working with compilers aren't getting those
[2764.00 --> 2765.20] performance benefits right now.
[2765.28 --> 2765.34] Yeah.
[2765.60 --> 2766.28] You can choose.
[2766.40 --> 2771.46] You can configure that not in create react app, but in a generic Babel config, you can
[2771.46 --> 2772.64] say these are the things.
[2773.10 --> 2776.76] This is the target set of features that I want to compile down to.
[2776.88 --> 2777.76] So it's certainly possible.
[2777.76 --> 2780.00] But I don't think many people go that far.
[2780.00 --> 2783.58] Well, and also there's only one minifier that supports it as well.
[2783.74 --> 2787.76] So and it's still under really active development.
[2788.00 --> 2790.18] So that's one of the things that you kind of lose.
[2790.56 --> 2792.76] Anyway, I think that we're about time for another break.