text stringlengths 0 2.02k |
|---|
[2265.06 --> 2270.52] Dino info. It'll show you your dependency tree. Generally, Dino knows all about your source code. |
[2270.58 --> 2276.24] It knows all about your dependencies and what it's doing. And you know, why not? We're distributing this |
[2276.24 --> 2282.00] executable. All of this stuff is written in Rust and so compiles down really tightly. And so, you know, |
[2282.06 --> 2289.38] our formatter that you run is not prettier. It's something called dprint and it's written in Rust |
[2289.38 --> 2295.90] and it's like a hundred times faster. And our linter is not ESLint. It's DinoLint and DinoLint is written |
[2295.90 --> 2306.64] in Rust and is very, very fast. We have a DinoLSP, which allows VS code to talk to Dino and get kind |
[2306.64 --> 2313.24] of tab completion, really nice editor interaction. So Dino can basically tell your editor, you know, |
[2313.30 --> 2319.64] documentation, all sorts of interesting interactions that Dino can help provide VS code. |
[2319.78 --> 2320.88] That's all really cool stuff. |
[2320.88 --> 2326.68] How much of that stuff is informed or inspired by other languages? Are you a language aficionado? |
[2326.84 --> 2330.92] Do you watch what the Go folks are doing, what the Elixir folks are doing, what the |
[2330.92 --> 2335.34] Haskell folks are? Are you watching these different ecosystems and saying, oh, there's a good idea? |
[2335.38 --> 2340.52] Because I think, you know, GoFoom kind of made this formatter a popular thing to do at the language |
[2340.52 --> 2345.38] level, not at like a third party level. And it seems like you're pulling in lots of good ideas. |
[2345.38 --> 2351.24] Absolutely. I mean, this heavily inspired by Go, not just in these various toolings, |
[2351.30 --> 2356.08] but all throughout the system. It's important to continually improve and take good ideas where |
[2356.08 --> 2356.94] you can get them. |
[2357.74 --> 2361.38] So the standard library is pretty fleshed out as well. I'm just wondering, like, where are the |
[2361.38 --> 2367.22] holes? Because Dino's still, it's a 1.0, but it's newer. It's not as mature. Is the ecosystem still |
[2367.22 --> 2373.00] waiting for like large holes to be filled? I know you're working on, or you recently shipped |
[2373.00 --> 2377.64] HTTP2 web server backend. So like things that you expect to be there are kind of there, |
[2377.76 --> 2381.24] but at a certain point, do you get to where you're looking for a package and it's only |
[2381.24 --> 2384.74] in Node, it's not in Dino. Is that still the situation? |
[2386.30 --> 2391.74] Yes, that is still the situation. You know, things take time. I expect this will be worked |
[2391.74 --> 2397.74] out over time. I remember back in like 2010, people would be like, how do I connect to my |
[2397.74 --> 2406.06] SQL with Node? And I would, you know, turn red face and apologize that I did not have a, |
[2406.06 --> 2410.74] my SQL library yet. And they're like, but how is that going to happen? And I said, I don't know |
[2410.74 --> 2416.34] how that's going to happen. I just hope that that is solved at some point. And like, it's embarrassing |
[2416.34 --> 2420.86] to even think that that was a concern because that was totally not a concern. Like all those problems |
[2420.86 --> 2426.30] got worked out, like essentially without my interaction, you know, Dino and Node are very |
[2426.30 --> 2430.96] similar systems. They're both built on V8. They're both JavaScript. They're pretty similar. |
[2431.18 --> 2438.54] The differences are relatively superficial. I mentioned this Dino land slash standard library, |
[2438.90 --> 2444.66] the Node compatibility layer. This is a work in progress. If you go check it out, there's something, |
[2445.22 --> 2452.58] I think we're maybe at 40% compatibility now. So we're still filling these things out. And I think |
[2452.58 --> 2459.72] over time, it will be less and less of a problem to take existing code and run it in Dino. But, |
[2459.80 --> 2465.10] you know, there's still a lot of work to do. So you mentioned this HTTP2 web server. So up until |
[2465.10 --> 2473.68] recently, Dino was using a web server written in TypeScript that was a loose port of Go's web server. |
[2473.68 --> 2483.00] So it's built on top of TCP sockets and TLS sockets. And, you know, it was a nice HTTP 1.1 server, |
[2483.24 --> 2491.60] but had some problems in that it didn't support HTTP2. And because we just made this port ourselves and |
[2491.60 --> 2497.60] are not particularly interested in writing web servers, we're kind of dead ended with that code |
[2497.60 --> 2504.74] base, like we're forced to then write an HTTP2 web server, HTTP2 being a much more complicated |
[2504.74 --> 2511.34] protocol than HTTP 1.1. What we'd really like to do, I mentioned this earlier about linking in native |
[2511.34 --> 2518.60] code libraries using Rust. Rust obviously has a web server already implemented. In fact, Dino already has |
[2518.60 --> 2525.18] that web server in its binary somewhere deep inside of it. What we really want to do is just allow people to |
[2525.18 --> 2533.42] call from JavaScript into this hyper web server in Rust and start up a nice fast HTTP2 web server. |
[2533.90 --> 2540.18] This work is still unstable. It's shipped in Dino 1.9. So, you know, people can use it if they use the |
[2540.18 --> 2546.82] dash dash unstable flag. But yeah, now we're working on this native web server, which I think is serving |
[2546.82 --> 2553.26] websites is quite important to server side JavaScript tasks. And it's quite fast. It has very good latency, |
[2553.26 --> 2560.06] very good throughput. So, you know, we have some preliminary analysis of its performance on the 1.9 |
[2560.06 --> 2565.30] release notes. We hope to stabilize this in the next couple of months, and people will have a very |
[2565.30 --> 2567.74] fast web server available right out of the box. |
[2583.10 --> 2589.14] This episode is brought to you by Cloud Zero. They help teams monitor, control, and predict their cloud |
[2589.14 --> 2595.06] value. And I talk with Ben Johnson, co-founder and CTO at Obsidian Security. They get tremendous value from |
[2595.06 --> 2600.82] using Cloud Zero. Ben shared with me the challenges they face driving innovation and customer value, |
[2600.82 --> 2605.06] while also trying to control and understand their Amazon Web Services spend. |
[2605.06 --> 2612.90] We want our engineers to move fast, to innovate, and to really focus on driving customer value. Yet, at the |
[2612.90 --> 2621.06] same time, reality is we have to pay for cloud compute and storage. And the challenge around AWS is often |
[2621.06 --> 2626.26] that you have multiple accounts, you have lots of different services, you have some people who only |
[2626.26 --> 2631.54] have access to development environments, not necessarily production. A lot of these different challenges |
[2631.54 --> 2638.18] across services, across accounts, that make it hard to understand the positive or negative impact |
[2638.18 --> 2645.46] to the costs that the new feature, the scale, you know, maybe the change in architecture are having. |
[2645.46 --> 2651.22] And so, giving our team more insight into the ramifications, again, positive or negative, |
[2651.22 --> 2655.78] of their changes in order to, maybe we need to really move fast, let's have less worry about cost right |
[2655.78 --> 2660.50] now, or maybe now we're in a more stable place, let's drive down the cost so we can, you know, |
[2660.50 --> 2666.02] give those cost savings on to our customers or improve our own margin. So, a product like Cloud Zero can |
[2666.02 --> 2671.54] really help your team get a handle on costs, get alerted to those spikes, feel good when you actually |
[2671.54 --> 2675.62] see the costs drop, and do all that without a whole lot of investment of your own time. |
[2675.62 --> 2679.70] All right, if your organization shares similar struggles as Band and Obsidian Security, |
[2679.70 --> 2684.90] check out Cloud Zero today. Learn more and get a demo at cloudzero.com slash changelog. |
[2684.90 --> 2687.22] Again, cloudzero.com slash changelog. |
[2698.42 --> 2704.90] So, we're on back March 29th, you announced the Dino company, and obviously you got a great |
[2705.46 --> 2710.66] new start to what you begin with Node. A lot of new beginnings here, a lot of wrongs made right, |
[2710.66 --> 2716.10] and to put behind that, a company. So, why are you on this path? I suppose more than 10 years, |
[2716.10 --> 2721.14] you know, 2009, now it's 2021. What's the path? What's the new company? What are you doing? |
[2721.14 --> 2728.82] David Plylar Programming languages are important. You know, the world is built on software and of the |
[2728.82 --> 2734.74] programming languages, dynamic programming languages are, I would say, what most people are interacting with |
[2734.74 --> 2740.90] day to day-to-day, you know, but most business needs are best expressed in, you know, Ruby on Rails or, |
[2741.54 --> 2748.66] you know, Node, JavaScript, or Python. Because, you know, to be honest, most problems are not compute-bound, |
[2748.66 --> 2754.18] they are engineering-bound, right? All of our time spent as programmers is kind of what we're trying |
[2754.18 --> 2761.54] to optimize for. And scripting languages, dynamic programming languages are the best way to get going |
[2761.54 --> 2769.86] fast. You know, I work on this problem because I think it's very empowering to be able to take a |
[2769.86 --> 2778.18] system like Node or Dino or Perl or Ruby and just start programming really fast. I got started with |
[2778.18 --> 2784.42] Perl myself, and I don't know if you guys feel this way, but there's that feeling of power when you first, |
[2784.42 --> 2789.06] like, really grok the language and you kind of get over, you know, you have to climb some mountains |
[2789.06 --> 2793.38] before you feel powerful in Perl. But, you know, once you get there, like, suddenly, like, the world |
[2793.38 --> 2799.22] is your oyster and you can just solve all sorts of problems very quickly. Yeah, I just don't think |
[2799.22 --> 2804.50] there's enough effort being, obviously, you know, I'm talking a lot about Rust, Go. These programming |
[2804.50 --> 2811.46] languages are very important, but they're targeted at kind of different use cases. You know, I love Rust. |
[2811.46 --> 2818.10] I would never start a new C++ project, as I said, but I'm not going to use that to rename a bunch of |
[2818.10 --> 2823.86] files on my folder. I'm just going to smash my keyboard for a while and write some JavaScript. |
[2823.86 --> 2832.34] Of the dynamic programming languages, JavaScript is the biggest. V8 is the fastest VM in JavaScript. |
[2832.34 --> 2839.70] It's the fastest dynamic language VM period. And JavaScript has this standardization, |
[2839.70 --> 2844.82] this industrial standardization process with it. It's tied to the web. So, you know, |
[2844.82 --> 2849.62] if you believe that the web is going to be here in a couple of years, JavaScript will certainly be |
[2849.62 --> 2855.30] here in a couple of years. It is as tied to the web as HTML is. So it's certainly not going anywhere. |
[2856.02 --> 2861.22] So, you know, if you're going to choose one of these languages to invest and build on, |
[2861.78 --> 2866.74] I think it's pretty clear that JavaScript is the language, ultimately, that we're going to be using. |
[2866.74 --> 2872.66] I think, you know, once you play around with this stuff a bit and you learn Python and you learn Ruby |
[2872.66 --> 2877.86] and you learn JavaScript and Perl, you get the feeling that they're pretty much the same thing. |
[2878.18 --> 2882.18] Like there's different syntax here. There's different ways to do different things, |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.