WEBVTT NOTE This file was generated by Descript 00:00:00.510 --> 00:00:03.360 CJ: In this episode, Colin and I talk about cogen and 00:00:03.360 --> 00:00:04.980 how we use Cogen at Stripe. 00:00:05.250 --> 00:00:07.950 We'll also mention a conference talk at Ruby Conf. 00:00:08.039 --> 00:00:12.090 That already happens, so if you're curious and you wanna see that, head over to the 00:00:12.090 --> 00:00:13.530 description and we'll have a link for you. 00:00:14.010 --> 00:00:14.490 Enjoy the show. 00:00:22.003 --> 00:00:23.083 Colin: Welcome to Build and Learn. 00:00:23.083 --> 00:00:24.093 My name is Colin. 00:00:24.193 --> 00:00:28.543 CJ: And I'm cj, and we weren't gonna record today, but we're suddenly both free 00:00:29.513 --> 00:00:29.933 Colin: Yep. 00:00:31.303 --> 00:00:33.853 CJ: Yeah, Twitter was gonna have a developer conference 00:00:33.858 --> 00:00:36.193 called chirp, but it's canceled. 00:00:36.343 --> 00:00:37.813 The entire conference is canceled. 00:00:38.398 --> 00:00:39.238 Colin: It is canceled. 00:00:39.238 --> 00:00:42.688 Yeah, we, I think we talked about it a few episodes ago where we were pretty 00:00:42.688 --> 00:00:48.988 excited to see, uh, Twitter getting back to its developer roots, and they kind 00:00:48.988 --> 00:00:53.038 of burned to the developer community a few times over the last decade or more. 00:00:53.038 --> 00:00:57.598 And this was looking like a good sign that that would be different. 00:00:57.658 --> 00:01:01.228 So if you're listening to this in 2023, you already know about 00:01:01.228 --> 00:01:05.098 all this, but obviously Elon bought Twitter what, in October? 00:01:05.788 --> 00:01:09.773 , I wouldn't have wanted to necessarily be at that conference working at Twitter 00:01:09.773 --> 00:01:12.233 and having to answer questions, right? 00:01:12.233 --> 00:01:15.053 Like, imagine being a dev advocate, right? 00:01:16.138 --> 00:01:18.803 and being asked all these questions that you don't have answers to. 00:01:19.958 --> 00:01:20.288 CJ: Right. 00:01:20.318 --> 00:01:24.038 I don't think there are any more develop, like all of the people. 00:01:24.968 --> 00:01:27.563 at Twitter who were on the developer advocacy team that 00:01:27.563 --> 00:01:28.314 I was interacting with. 00:01:28.369 --> 00:01:30.799 I don't know any of them that are still there. 00:01:30.949 --> 00:01:35.839 Like, uh, half of them or more were laid off in the first round. 00:01:36.229 --> 00:01:40.039 Some of them quit shortly after, and I think the rest took the. 00:01:41.804 --> 00:01:44.714 Opportunity to bail when Elon was like, Hey, if you don't want to be 00:01:44.714 --> 00:01:48.674 here and grind it out, then like, you can, here's the door , and they 00:01:48.719 --> 00:01:50.969 Colin: Yeah, he sent everyone that email of, do You Like Me? 00:01:50.969 --> 00:01:51.659 Yes or no? 00:01:51.779 --> 00:01:52.139 Right. 00:01:54.134 --> 00:01:54.434 CJ: yeah. 00:01:54.824 --> 00:01:55.994 Circle yes or no. 00:01:56.279 --> 00:01:59.249 Colin: uh, so, so chaotic, I think. 00:01:59.279 --> 00:01:59.489 Yeah. 00:01:59.489 --> 00:02:05.069 Chris Mesina tweeted that it was like 97% of the, I don't know if it's, 00:02:05.459 --> 00:02:08.309 he said platform, so I don't know if that means developer platform 00:02:08.309 --> 00:02:10.649 and in like even the developers. 00:02:11.514 --> 00:02:16.104 of the platform, including the developer, obviously team are just not there anymore. 00:02:16.104 --> 00:02:19.534 So that's gonna be interesting to see where it goes. 00:02:20.324 --> 00:02:20.684 CJ: Yeah. 00:02:20.684 --> 00:02:27.534 There's also just a ton of people going out in like flames of glory , you know, 00:02:27.534 --> 00:02:34.114 like just calling out Elon on Twitter for showing bad charts or using bad 00:02:34.114 --> 00:02:39.269 metrics or, saying like, oh, I think he said something like, oh, the A the 00:02:39.269 --> 00:02:42.599 app is slow in certain countries because it's doing thousands of RPC calls. 00:02:42.659 --> 00:02:44.219 And then someone was like, not true. 00:02:44.219 --> 00:02:48.869 I've been working on this for like six years and , you know, blah, blah, blah. 00:02:48.929 --> 00:02:51.449 And then, you know, five seconds later that person's fired. 00:02:51.939 --> 00:02:55.149 Just today I was watching some tos where people were like counting 00:02:55.149 --> 00:02:56.679 down until they got locked out. 00:02:56.739 --> 00:03:01.689 And then, you know, they had champagne ready to go when they got locked 00:03:01.689 --> 00:03:03.009 outta their laptops and stuff. 00:03:03.059 --> 00:03:06.539 Colin: I mean, even, uh, Justin Jackson was sharing some tweets today 00:03:06.569 --> 00:03:10.409 about how, like there's been on some news posts that came out this morning 00:03:10.409 --> 00:03:13.319 about how the building is locked down again and no one can get in. 00:03:13.829 --> 00:03:16.949 And then he shared a tweet from someone who used to run their access 00:03:16.949 --> 00:03:20.639 control and that Elon's calling him back to help, like they're all locked 00:03:20.639 --> 00:03:23.729 out of the building and that he needs help getting back into the building. 00:03:23.789 --> 00:03:23.879 It's 00:03:23.879 --> 00:03:26.009 like, oh man. 00:03:27.259 --> 00:03:27.569 , I don't know. 00:03:27.599 --> 00:03:31.019 He should have bought it and not touched it, I think is what it's like. 00:03:31.319 --> 00:03:34.469 It's like a no make, don't make any big decisions right after 00:03:34.469 --> 00:03:35.849 you make a big purchase decision. 00:03:35.849 --> 00:03:36.209 It's like 00:03:37.154 --> 00:03:37.634 CJ: Yeah. 00:03:37.904 --> 00:03:43.874 The way that I've been thinking about it is that, He has started or made big 00:03:43.874 --> 00:03:48.314 influential moves on the cultures of Tesla and SpaceX and all these other companies 00:03:48.674 --> 00:03:50.354 when they were really in their infancy. 00:03:50.384 --> 00:03:54.644 And so the people I know who work at SpaceX or worked at 00:03:54.649 --> 00:03:58.874 SpaceX made the conscious decision when they joined SpaceX. 00:03:59.104 --> 00:04:01.684 To sleep in their van in the parking lot because they knew they were gonna 00:04:01.684 --> 00:04:03.364 have to work like 90 hours a week. 00:04:03.664 --> 00:04:06.094 But they did that because they wanted to be part of like this 00:04:06.094 --> 00:04:06.934 thing that he was building. 00:04:07.384 --> 00:04:10.354 But on the other hand, there was like all of these people at Twitter who I 00:04:10.359 --> 00:04:15.034 think were like really happy with the quality of life and they chose Twitter 00:04:15.039 --> 00:04:20.524 for, the community and like all of these, great work-life balance benefits and, 00:04:20.524 --> 00:04:21.964 working remotely and all this stuff. 00:04:21.994 --> 00:04:26.104 And then he comes in and shifts the culture from the top down and you. 00:04:26.859 --> 00:04:31.274 , 80% of the people who worked there didn't want to work in that 00:04:31.274 --> 00:04:34.514 environment, and so it just kind of like blew up in his face a bit. 00:04:35.414 --> 00:04:35.744 But 00:04:35.939 --> 00:04:37.799 Colin: Well, and I think the rhetoric right now is that 00:04:37.859 --> 00:04:38.999 they don't wanna work hard. 00:04:38.999 --> 00:04:40.609 And it's just a different culture. 00:04:40.609 --> 00:04:44.809 It's like there are definitely people who got let go, who know how Very, very 00:04:44.809 --> 00:04:46.379 specific things inside of Twitter were. 00:04:47.089 --> 00:04:49.399 And now no one will know how those things work. 00:04:49.909 --> 00:04:53.029 And sure, that might be bad documentation or whatever, but like 00:04:53.029 --> 00:04:54.589 I'm looking at my Twitter right now. 00:04:54.889 --> 00:04:57.409 I've been on Twitter since March of 2007. 00:04:57.859 --> 00:05:01.549 There is more than 15 years of culture that has already 00:05:01.554 --> 00:05:03.269 been developed, at Twitter. 00:05:03.789 --> 00:05:05.799 If people have been there fr since that time. 00:05:06.159 --> 00:05:09.119 Some people have been there only a few years, but You could probably look 00:05:09.119 --> 00:05:12.189 around at all the tech layoffs and realize that yeah, everyone probably 00:05:12.194 --> 00:05:14.279 overhired since the pandemic started. 00:05:15.914 --> 00:05:21.044 Just wholesale cutting half of your team, your company because you bought 00:05:21.044 --> 00:05:25.334 something with a bunch of debt is gonna have some waves for sure. 00:05:25.634 --> 00:05:28.064 And I think everyone's trying to figure out like, how do you make money? 00:05:28.064 --> 00:05:30.584 And like we, we pay for Twitter's api like. 00:05:30.604 --> 00:05:35.194 I don't think a lot of people know that there is a paid offering for the a p I. 00:05:35.244 --> 00:05:38.784 And I think some people were like making calls on Twitter for like, make 00:05:38.784 --> 00:05:42.324 the a p I fully open again and let us make as many calls as we want. 00:05:42.324 --> 00:05:46.434 And it's does anyone remember what it was like when that was the case? 00:05:46.439 --> 00:05:48.354 Like fail whale was prevalent. 00:05:48.504 --> 00:05:50.514 It wasn't a reliable api. 00:05:50.514 --> 00:05:54.084 I, the guy who works, or probably the team now that works on. 00:05:54.989 --> 00:06:00.239 Tweet bot, you know, I've followed him for years and just watched like their issues. 00:06:00.239 --> 00:06:02.849 Cause they even were like limited to the number of API keys. 00:06:02.849 --> 00:06:03.449 They were allowed. 00:06:04.049 --> 00:06:08.609 And so it's like when you buy Tweetbot, you get one of our hundred thousand 00:06:08.759 --> 00:06:11.699 p i keys and only a hundred thousand people can use our Twitter client. 00:06:11.704 --> 00:06:13.869 It's like in what world is that normal 00:06:14.449 --> 00:06:15.059 CJ: Mm-hmm. 00:06:15.119 --> 00:06:20.399 . Part of his thing early on was like, there's too many bots on Twitter, 00:06:20.399 --> 00:06:24.269 and so I could see him arguing for just shutting the whole API down. 00:06:24.269 --> 00:06:26.909 Like, you have to be a human That's on the app that's tweeting 00:06:26.914 --> 00:06:28.284 and we're only gonna let that. 00:06:28.488 --> 00:06:31.338 Fly and not let people have API access. 00:06:31.338 --> 00:06:35.148 But gosh, I don't want that to be the world. 00:06:35.178 --> 00:06:35.508 Colin: true. 00:06:35.748 --> 00:06:38.568 And that's something that I'm paying attention to cuz we integrate 00:06:38.573 --> 00:06:40.788 heavily with Twitter at orbit, right? 00:06:40.788 --> 00:06:42.558 So I'm like, okay, what does this mean for us? 00:06:42.563 --> 00:06:46.348 And some people are asking us for Mastodon, integration now, 00:06:46.348 --> 00:06:47.098 and we're looking at that. 00:06:47.098 --> 00:06:51.358 But like some people, I still, I looked at the API and it's pretty similar. 00:06:52.123 --> 00:06:55.138 but, and this will get us into our topic today, but the, the tricky 00:06:55.138 --> 00:06:58.708 thing is it's just not a really mature set of tools around it. 00:06:59.278 --> 00:07:02.558 And Mastodon servers are just melting right now because people 00:07:02.558 --> 00:07:04.208 don't know how to host Mastodon. 00:07:04.208 --> 00:07:08.288 People are realizing that a centralized service with a whole DevOps team, 00:07:08.288 --> 00:07:09.878 it turns out, is pretty important. 00:07:09.928 --> 00:07:12.488 Instead of a bunch of, ma on fail whales, 00:07:12.908 --> 00:07:15.848 CJ: So you are, are you on Macedon, like as a user? 00:07:16.298 --> 00:07:20.078 Colin: Yeah, I joined the Ruby Social one, so ruby.social. 00:07:20.168 --> 00:07:24.638 Um, I like it, you know, it's a little, it's definitely rough. 00:07:24.668 --> 00:07:26.438 You can tell it's an open source project. 00:07:26.528 --> 00:07:26.828 CJ: Yeah. 00:07:27.758 --> 00:07:30.788 .is it is it like implemented with Ruby? 00:07:30.788 --> 00:07:31.658 Is it the Rails app? 00:07:31.658 --> 00:07:32.018 Do you know? 00:07:32.018 --> 00:07:32.498 Like, 00:07:32.598 --> 00:07:33.528 Colin: I actually don't know. 00:07:33.858 --> 00:07:34.208 Um, 00:07:34.298 --> 00:07:37.598 CJ: Because there was someone who posted in the Ruby social 00:07:37.598 --> 00:07:38.948 Mastodon main feed or whatever. 00:07:39.858 --> 00:07:42.143 also, I don't know what the terms are, . I'm still trying 00:07:42.143 --> 00:07:43.073 to figure out what, like this 00:07:43.343 --> 00:07:44.543 Colin: Those toot toots, 00:07:44.783 --> 00:07:48.413 CJ: yeah, you gotta make some toots, someone tooted and, uh, , their toot 00:07:48.418 --> 00:07:52.763 was about how like there's all these issues on one of the master on repos 00:07:52.763 --> 00:07:56.093 and they were basically making a call to action for all the rubus that were 00:07:56.093 --> 00:07:57.773 in ruby.social to help maintain it. 00:07:58.253 --> 00:08:00.713 I was like, is this thing like implemented in Ruby? 00:08:00.713 --> 00:08:01.103 I don't know. 00:08:01.538 --> 00:08:04.868 Colin: Well, I think, uh, Mike from Sidekick runs the Ruby Social one. 00:08:05.198 --> 00:08:08.768 Um, yeah, so if anyone knows how to scale things, I think 00:08:08.768 --> 00:08:10.298 he's gonna be pretty set there. 00:08:10.298 --> 00:08:12.248 But yeah, I'll have to check and see. 00:08:12.248 --> 00:08:12.998 It'd be interesting. 00:08:12.998 --> 00:08:16.598 Maybe we, we should do a, a chat and maybe even we find someone 00:08:16.598 --> 00:08:19.838 who's more familiar with Mastodon to talk about it in the next episode. 00:08:20.528 --> 00:08:22.608 Actually getting Mike on the show would be really good too. 00:08:22.638 --> 00:08:25.338 Just talk about sidekick stuff eventually. 00:08:25.488 --> 00:08:28.878 But with all that said, like there is an API for Mastodon. 00:08:29.118 --> 00:08:35.448 When we get into working with APIs, we can work with those like raw and make API 00:08:35.448 --> 00:08:41.268 calls, but most of us are using SDKs and you're actually gonna be doing a talk. 00:08:41.953 --> 00:08:45.613 I guess the talk will have happened when this episode comes out, but, uh, 00:08:45.913 --> 00:08:49.993 it'll probably be released on YouTube a little bit after this podcast comes out. 00:08:49.993 --> 00:08:55.153 So, um, let's dive into s TK Generation and kind of what you're 00:08:55.153 --> 00:08:56.953 gonna be chatting about at Ruby Comps. 00:08:57.453 --> 00:08:57.693 CJ: Yeah. 00:08:57.693 --> 00:09:02.478 So a lot of times if you're implementing an api, , you can 00:09:02.478 --> 00:09:09.018 have the API spit out an artifact that has like the shape of the api. 00:09:09.558 --> 00:09:11.388 A lot of times it's like the open API spec. 00:09:11.393 --> 00:09:13.968 Are you are, do you have the open API spec for Orbit? 00:09:14.118 --> 00:09:14.538 Uh, 00:09:14.958 --> 00:09:20.628 Colin: Yeah, we use, uh, open api, which I think is the same as swagger these days. 00:09:20.628 --> 00:09:20.898 Right. 00:09:20.898 --> 00:09:21.678 It's, yeah. 00:09:22.133 --> 00:09:22.608 CJ: Swagger. 00:09:22.608 --> 00:09:25.218 I think Swagger got like renamed to open api. 00:09:25.588 --> 00:09:25.863 Colin: Exactly. 00:09:25.863 --> 00:09:26.193 Yeah. 00:09:27.223 --> 00:09:32.683 CJ: So it could be open API speck, or it could be, I think jsun API is another one. 00:09:32.683 --> 00:09:37.993 And then, uh, or you could have like G R P C proto buff, like your 00:09:37.993 --> 00:09:39.763 proto buff schema definitions. 00:09:39.763 --> 00:09:42.493 Could be like another artifact, but basically you just want like, 00:09:43.303 --> 00:09:44.983 or even in, in GraphQL too, right? 00:09:44.988 --> 00:09:46.223 The GraphQL schema. 00:09:46.693 --> 00:09:49.813 All of these are versions of some. 00:09:51.463 --> 00:09:55.603 Sort of declarative thing that gives you the shape of your api. 00:09:55.603 --> 00:09:58.543 And it's usually like, what are the endpoints, what are all the 00:09:58.543 --> 00:10:01.633 different parameters it expects what's required and not required. 00:10:02.083 --> 00:10:06.613 And what you can do is take that spec and then run it through a tool 00:10:06.613 --> 00:10:08.833 to generate and spit out an sdk. 00:10:09.553 --> 00:10:15.083 And if we really wind back the clocks to, talking about like XML APIs. 00:10:15.443 --> 00:10:19.563 Then, I don't know, did you ever have to work with those SDS 00:10:20.543 --> 00:10:21.623 Colin: I had to do. 00:10:22.703 --> 00:10:26.093 I had to do soap APIs with wisdom files. 00:10:26.123 --> 00:10:26.453 Yeah. 00:10:26.543 --> 00:10:27.563 CJ: Yes, yes. 00:10:27.568 --> 00:10:27.873 Yeah. 00:10:27.873 --> 00:10:29.753 So, uh, way back in the day. 00:10:30.633 --> 00:10:36.008 , this is like Prej, S o uh, a lot of APIs were implemented with xml, and we had to 00:10:36.008 --> 00:10:41.428 do this at my VR with, some of the big booking engines or like listing platforms. 00:10:41.908 --> 00:10:43.408 They, they didn't have JS O APIs. 00:10:43.408 --> 00:10:48.358 They had XML APIs, and you could get an X S D, which is a XML schema 00:10:48.358 --> 00:10:51.658 definition, which is similar to an open API spec or any of these things. 00:10:52.433 --> 00:10:56.788 but then there were tools that could take this schema definition and generate 00:10:56.788 --> 00:11:01.288 bindings or an sdk, client library, whatever, in a given programming language. 00:11:01.678 --> 00:11:05.128 So in that case, we were, you know, taking x sd, converting it to Python. 00:11:05.758 --> 00:11:10.678 Um, and for the Stripe api we spit out the open API spec. 00:11:11.128 --> 00:11:15.058 And then we have this really, really badass tooling that takes 00:11:15.058 --> 00:11:19.828 the Open API spec and generates all seven of our official support. 00:11:20.518 --> 00:11:23.188 Server languages like the SDKs for all of those languages. 00:11:23.188 --> 00:11:27.988 So Ruby, Python, PHP, node, go Java and.net. 00:11:28.018 --> 00:11:32.458 We have Stripe SDKs for all of those languages and they're 00:11:32.458 --> 00:11:36.268 all now automatically generated using the open API spec. 00:11:36.868 --> 00:11:41.848 And I like, I know that I've used Orbit Ruby or some sort 00:11:41.853 --> 00:11:44.038 of Ruby clients for Orbit. 00:11:44.038 --> 00:11:47.428 Are those also generated or are those main maintained by hand? 00:11:48.373 --> 00:11:52.243 Colin: You know, I think they were written by hand and I would 00:11:52.243 --> 00:11:54.103 say they're not maintained, so 00:11:54.368 --> 00:11:55.378 CJ: Oh, they're not maintained at all. 00:11:55.603 --> 00:11:59.683 Colin: like, I haven't touched them since I've been here. 00:12:00.403 --> 00:12:04.393 So, and yeah, the folks who built them don't work here anymore. 00:12:04.453 --> 00:12:08.383 So I think this is one of those things where, you know, I wrote down here like, 00:12:08.383 --> 00:12:10.813 what does SDK generation help to solve? 00:12:10.813 --> 00:12:13.783 Because the a API I has changed, right? 00:12:13.783 --> 00:12:16.573 And so we get to spit out at this new artifact of a new 00:12:16.663 --> 00:12:18.433 spec, a new open API doc. 00:12:19.093 --> 00:12:23.028 And it would be really nice to have that always be in sync right now. 00:12:23.028 --> 00:12:27.348 Once that changes, the docs need to be updated and the SDKs need to be updated. 00:12:27.858 --> 00:12:30.258 And then you need to communicate those SDK updates. 00:12:30.318 --> 00:12:35.268 You know, if it's a gem, if it's a MPM module, those need to make it out as 00:12:35.273 --> 00:12:39.048 as version changes, you know, whether they're breaking or non breaking changes. 00:12:39.558 --> 00:12:44.388 Do you think that most s STKs today are done automatically, or do you think that 00:12:44.388 --> 00:12:45.918 a lot of people are still handwriting. 00:12:47.463 --> 00:12:47.753 CJ: I. 00:12:48.398 --> 00:12:51.618 . So there's a, there is an interesting trade off when you're looking at Generat 00:12:51.618 --> 00:12:56.568 like automatic generation, and that is like if you want your SDK to be readable, 00:12:57.048 --> 00:13:01.878 then using the automatic generation tools don't really generate very readable code. 00:13:01.908 --> 00:13:05.748 And so if you just take like the open API spec and you run it through the Swagger 00:13:05.753 --> 00:13:12.108 Code Gen or the Open API code cogen stuff, it spits out like a giant, huge library. 00:13:13.098 --> 00:13:15.573 it's automatically generated, but it's really tough to use. 00:13:16.083 --> 00:13:20.643 Um, and so your trade off is like, do we want to use the automatic generation 00:13:20.643 --> 00:13:25.323 and have like a really tough to use tool or do we wanna write our own or do we 00:13:25.323 --> 00:13:28.923 want to like build our own generation tooling that does like a hybrid of 00:13:28.923 --> 00:13:32.293 both?, so that's the tooling that I'm excited to share and talk about is like 00:13:32.293 --> 00:13:37.363 this hybrid of, of both worlds where we've, we've built this pipeline, but. 00:13:38.683 --> 00:13:42.524 The whole point of SDK generation is to make sure that your s like 00:13:42.529 --> 00:13:45.373 all of your client libraries are staying up to date with the api. 00:13:45.973 --> 00:13:48.973 And so, yeah, going back to the top of the show, right, you're talking about 00:13:48.973 --> 00:13:54.943 how like, yeah, we might just make API calls directly to the API using Curl 00:13:54.943 --> 00:13:56.293 or Postman or something like that. 00:13:56.703 --> 00:13:59.953 But when we're integrating that into our applications, we're gonna be 00:13:59.953 --> 00:14:04.073 using Ruby code or JavaScript code or whatever to make those API calls. 00:14:06.023 --> 00:14:11.853 One option is to write your own like class that represents a client of 00:14:11.853 --> 00:14:17.403 the API and then use, HTP party, rest client, whatever sort of built-in 00:14:17.403 --> 00:14:22.443 library there is for making HTP calls or to use a library that is provided 00:14:22.443 --> 00:14:24.873 by the third party to integrate with. 00:14:24.873 --> 00:14:28.263 For instance, transistor, the host, the podcast host that we have for this 00:14:28.263 --> 00:14:33.893 platform, I found like an unofficial community, Ruby Library for it. 00:14:33.893 --> 00:14:35.243 And I was like, ah, I don't think so. 00:14:35.903 --> 00:14:40.433 . And so, um, I want to add a list of all of our podcasts to my website. 00:14:40.643 --> 00:14:42.443 I was like, oh, I'll just use the transistor API and 00:14:42.443 --> 00:14:43.873 you can like get an API key. 00:14:44.153 --> 00:14:50.213 And so I just built, I wrote a class, very simple transistor RV like Ruby class. 00:14:51.783 --> 00:14:54.843 , has a couple of helper methods on it that will just make the 00:14:54.843 --> 00:14:58.923 request to fetch all the episodes and page and iterate over those. 00:14:59.313 --> 00:15:01.653 So that's the trade off between the two is right, like you're gonna either 00:15:01.653 --> 00:15:08.403 have, um, no, no SDK and you're expect your all developers to go and implement 00:15:08.403 --> 00:15:10.593 all of those helper methods themselves. 00:15:11.493 --> 00:15:15.543 Or you have an SDK that's auto-generated by a really generic. 00:15:16.583 --> 00:15:21.333 Platform like swagger, cogen, you end up with this massive, huge library 00:15:21.333 --> 00:15:22.653 that is like really hard to read. 00:15:22.953 --> 00:15:26.013 Or the third option is that you build your own generation tooling. 00:15:26.493 --> 00:15:30.123 Colin: And when you say it's hard to read, you're talking like, I mean now it's just 00:15:30.123 --> 00:15:33.813 as hard, like you're, you're having to reference the docs all the time because 00:15:33.813 --> 00:15:35.703 it's not nece, it's machine readable. 00:15:35.703 --> 00:15:38.073 It's not necessarily like a, an amazing developer experience. 00:15:38.718 --> 00:15:39.378 CJ: exactly. 00:15:39.378 --> 00:15:40.258 To make this more concrete. 00:15:40.993 --> 00:15:45.373 If you take the Open API artifacts that are spit out for the Stripe API and you 00:15:45.373 --> 00:15:48.013 run it through swagger, cogen, one of the. 00:15:48.988 --> 00:15:52.768 Like wrapper classes it gives you is 30,000 lines of Ruby code. 00:15:53.398 --> 00:15:58.138 And then for every single resource in the API and all of its like sub 00:15:58.138 --> 00:16:03.328 resources, it will generate CL classes and files and there's like 2,200 files. 00:16:03.718 --> 00:16:08.338 So it's like you're either reading through this ridiculously massive like file 00:16:08.338 --> 00:16:12.208 or like thousands of files to try to figure out, you know, how do I construct 00:16:12.213 --> 00:16:15.628 an API call with the right classes and the right methods and things like that. 00:16:15.628 --> 00:16:16.138 And so, 00:16:16.328 --> 00:16:16.573 Colin: right? 00:16:16.933 --> 00:16:20.923 Because you do also have to have like docs for this new client library too. 00:16:20.983 --> 00:16:21.223 When. 00:16:21.548 --> 00:16:22.298 Could build that. 00:16:22.298 --> 00:16:22.538 Right? 00:16:22.538 --> 00:16:29.248 And Stripe is infamous for having very good docs and, very human readable, SDKs. 00:16:29.253 --> 00:16:32.848 And I think a lot of people think that to get there, you have to have this huge docs 00:16:32.848 --> 00:16:37.048 team and this huge s STK team and that you're gonna write everything bespoke. 00:16:37.048 --> 00:16:42.948 But you guys have kind of fallen in that happy middle of generation plus some 00:16:42.948 --> 00:16:47.298 hybrid, is that custom handwritten stuff that's being done in the middle there? 00:16:47.948 --> 00:16:54.218 CJ: So, Yes, and for a long time, for like seven plus years, we manually 00:16:54.718 --> 00:16:56.918 maintained and hand wrote all the SDKs. 00:16:57.308 --> 00:17:01.958 So when a new feature would land in the api, there was a Slack notification and 00:17:01.963 --> 00:17:05.628 then, these volunteers would jump in and start making prs to all of the different 00:17:05.633 --> 00:17:07.818 SDKs in publishing, publishing those. 00:17:07.818 --> 00:17:10.818 And that would take several hours, sometimes a full day of. 00:17:11.598 --> 00:17:14.778 To add the features, especially to Stripe, Java, like Java's just so 00:17:14.778 --> 00:17:18.798 verbose that you know, if you change an eum, you've gotta go change and 00:17:18.798 --> 00:17:20.568 update like six files in Stripe, Java. 00:17:21.228 --> 00:17:23.628 And so we built this layer. 00:17:23.868 --> 00:17:26.868 It took two years to migrate all the SDKs to, and the other thing 00:17:26.868 --> 00:17:30.398 was that we were trying to build the generation so that it mirrored what 00:17:30.528 --> 00:17:32.418 the handwritten versions were before. 00:17:32.568 --> 00:17:35.748 So like you've gotta back into your existing patterns. 00:17:36.258 --> 00:17:38.638 And so we built this f. 00:17:39.993 --> 00:17:45.543 Uh, this guy Alex Ratray, he runs a company called stainless.dev that 00:17:45.543 --> 00:17:50.373 does this like SDK generation and developer experience as a service now. 00:17:50.373 --> 00:17:52.083 So you can like hire stainless.dev. 00:17:52.653 --> 00:17:57.363 And, uh, he built like the first versions of this tool that we use 00:17:57.363 --> 00:17:58.743 internally called Prettier Poet. 00:17:59.373 --> 00:18:03.243 And Prettier poet is a react like framework where you. 00:18:04.063 --> 00:18:09.763 Components that are language specific, and they give you things like method 00:18:09.763 --> 00:18:16.333 calls or, you know, literals or strings or, uh, as like, you know, in instance, 00:18:16.703 --> 00:18:20.113 variable assignment or class definitions. 00:18:20.113 --> 00:18:23.233 So you can use these, language specific ones. 00:18:23.233 --> 00:18:25.273 They also have like all the tooling for comments. 00:18:25.753 --> 00:18:28.943 And then when you build Your high level react component. 00:18:29.513 --> 00:18:32.753 You can use the open API spec as the data. 00:18:32.843 --> 00:18:34.613 That's kind of like your original data tree. 00:18:34.613 --> 00:18:35.933 That's the input into that. 00:18:35.938 --> 00:18:37.553 And so you can generate all the classes. 00:18:37.553 --> 00:18:40.913 So in the Stripe api, there's like customers and subscriptions 00:18:40.913 --> 00:18:42.273 and payment intents. 00:18:43.293 --> 00:18:45.063 Those are all defined in the open API spec. 00:18:45.063 --> 00:18:48.843 You can like iterate overall the resources in the open API spec and have those feed 00:18:48.843 --> 00:18:54.393 into this React component that is gonna be like, oh, this is the resource component 00:18:54.723 --> 00:18:56.553 and it knows how to generate the class. 00:18:56.558 --> 00:18:59.193 And then it can use the spec to know like what's the class name and what 00:18:59.198 --> 00:19:03.873 are all the methods and what are my, what are the documentation strings 00:19:03.873 --> 00:19:06.723 that should be added to the top of the class and to each of the methods and. 00:19:08.493 --> 00:19:11.958 What's really cool about this is that like at first we did it just for SDK 00:19:11.958 --> 00:19:17.358 generation, but now we have like this tooling that generates the Postman client. 00:19:17.358 --> 00:19:20.208 It generates the API reference. 00:19:20.448 --> 00:19:24.588 All of the code snippets that you see in the docs we're generating like tons 00:19:24.588 --> 00:19:30.288 and tons of those with this like higher order markdoc.dev component called 00:19:30.288 --> 00:19:34.908 cogen snippet, where you basically say like, okay, this is a, an api. 00:19:35.223 --> 00:19:40.713 To v1/customers it's a post request and here's the arguments, and that will spit 00:19:40.713 --> 00:19:42.543 out the code in all seven languages. 00:19:42.543 --> 00:19:45.393 It shows the code in curl, it shows how to do it with Stripe cli, and 00:19:45.393 --> 00:19:47.133 that's like all autogenerated now. 00:19:47.493 --> 00:19:52.233 And before that was all like just string and interpolation in erb files, so 00:19:52.893 --> 00:19:53.283 Colin: Wow. 00:19:53.403 --> 00:19:53.613 Yeah. 00:19:53.613 --> 00:19:57.033 I was gonna ask, because it sounds very similar to Mark Docs, so you 00:19:57.033 --> 00:20:02.533 guys have Mark Docs and your SDKs are being fed by this generation, 00:20:03.793 --> 00:20:06.153 if I'm, if I'm thinking of the trees properly. 00:20:06.213 --> 00:20:06.513 Right. 00:20:06.513 --> 00:20:11.133 So you have your open API spec, which then, feeds into prettier poet and then. 00:20:11.733 --> 00:20:17.283 Generates those, those artifacts, fors, stk, STKs, and your docs. 00:20:18.008 --> 00:20:18.723 CJ: Yeah, exactly. 00:20:18.723 --> 00:20:23.133 And even things like, we have a visual studio code extension. 00:20:23.133 --> 00:20:26.313 So if you install like the stripe for VS code extension, then you get. 00:20:27.048 --> 00:20:28.188 Snippet completion. 00:20:28.188 --> 00:20:32.088 So you can type like stripe payment intent, create and hit tab, and that 00:20:32.088 --> 00:20:34.218 will auto complete a snippet in vs code. 00:20:34.248 --> 00:20:36.228 Like that's also powered by this thing. 00:20:36.708 --> 00:20:40.548 And so, uh, I think even like the stripe shell, so if you go to the documentation 00:20:40.548 --> 00:20:45.348 and you hit like tilde or whatever, it pops open a little drawer at the bottom 00:20:45.348 --> 00:20:49.373 that lets you have this interactive shell for making API calls directly in the docs. 00:20:49.883 --> 00:20:51.823 And all of that is backed by, 00:20:52.033 --> 00:20:52.543 Colin: Wait, this. 00:20:52.543 --> 00:20:54.193 is, this is in the docs that you 00:20:54.283 --> 00:20:57.823 CJ: Yeah, if you go to like stripe.com/docs and then hit . Yeah. 00:20:57.823 --> 00:21:01.153 Hit like, you know, back tick or whatever, it should pop open. 00:21:01.813 --> 00:21:05.363 Stripe Shell which is like a browser-based version of the Stripe c l 00:21:05.558 --> 00:21:06.338 Colin: Oh, wow. 00:21:06.968 --> 00:21:07.178 Yeah. 00:21:09.788 --> 00:21:11.438 we're finding out all the secrets today. 00:21:12.008 --> 00:21:12.368 Yeah. 00:21:12.788 --> 00:21:17.288 The other one we just found, if you go to the, uh, if you just go to customer.new 00:21:17.948 --> 00:21:18.878 or do you know about this one? 00:21:20.033 --> 00:21:20.363 CJ: No. 00:21:20.363 --> 00:21:21.233 What is customer not new? 00:21:21.458 --> 00:21:25.268 Colin: If you go to customer.new as a, as a u r url, it launches into 00:21:25.268 --> 00:21:28.988 Stripe and tries to create a new customer if you're logged into Stripe. 00:21:29.478 --> 00:21:32.418 Yeah, we've just found that cuz we also, we have member new for 00:21:32.418 --> 00:21:36.468 Orbit, so if you just like even make a bookmark for member.new. 00:21:37.123 --> 00:21:38.263 And click on it. 00:21:38.263 --> 00:21:41.863 It just launches a little screen to add a new member to your workspace. 00:21:42.583 --> 00:21:46.213 But um, I think a lot of people are start, there's like a whole list of all the 00:21:46.213 --> 00:21:47.893 ones that you can do around the web now. 00:21:47.943 --> 00:21:51.813 I'm sure there's like a notion, page.new or something out there, but 00:21:52.658 --> 00:21:53.768 CJ: so I definitely knew. 00:21:53.798 --> 00:21:57.488 Okay, so invoice.new works, subscription.new. 00:21:58.613 --> 00:21:59.333 Works. 00:21:59.393 --> 00:22:02.243 So there were a couple that I knew about, I did not know about customer New 00:22:03.458 --> 00:22:07.538 . Colin: Yeah, it's whatever company gets to the TLD before anyone else because 00:22:08.398 --> 00:22:12.208 Now no one else can do subscription.new, only, only Stripe powers that. 00:22:12.208 --> 00:22:14.418 But yeah, fun little Easter eggs for the internet. 00:22:14.418 --> 00:22:17.568 I'm trying to think if we have any other ones for, but I'll have to share 00:22:17.568 --> 00:22:21.408 that console with, uh, with uh, Alex who works on a lot of our Stripe stuff. 00:22:21.708 --> 00:22:25.818 And even the, I didn't know that there was a VS code extension, you know, cause in 00:22:25.823 --> 00:22:29.838 my mind I was trying to think of like how all this is really exciting when you think 00:22:29.838 --> 00:22:33.408 about even the open API spec teaching. 00:22:33.813 --> 00:22:35.973 GitHub co-pilot, right? 00:22:36.003 --> 00:22:39.783 Like, just being able to be like, I don't want to go, always look at 00:22:39.783 --> 00:22:43.683 the docs for every little thing, but like GitHub co-pilot or Stripe 00:22:43.953 --> 00:22:45.453 cli, or Stripe, you know, vs. 00:22:45.458 --> 00:22:49.413 Code extension helped me with what I'm working on here. 00:22:49.513 --> 00:22:52.673 And when I've used s STKs, like some SDKs are just as. 00:22:53.358 --> 00:22:59.148 Like, I would rather go just do curl calls or like http party or something 00:22:59.148 --> 00:23:02.728 like that, because I know what I'm doing., there are some amazing ones. 00:23:02.728 --> 00:23:07.793 I'm betting that the Discord one is probably auto-generated, but what's 00:23:07.798 --> 00:23:10.823 really cool is that like once you get the table stakes of just good 00:23:11.203 --> 00:23:16.143 s stk generation done, the discord, client libraries also help you. 00:23:17.183 --> 00:23:19.163 Rate limiting like automatically. 00:23:19.853 --> 00:23:24.833 Um, and it's, so it's got a bunch of like Texs and like backgrounding that 00:23:24.833 --> 00:23:28.673 happens to make sure that you're, like, you're not writing the code yourself to 00:23:28.678 --> 00:23:31.343 make sure you're staying within those rate limits, which is really nice. 00:23:31.863 --> 00:23:34.933 But you probably don't get to work on stuff like that if you're having 00:23:34.933 --> 00:23:38.863 to hand write these things every single time that the API changes. 00:23:38.953 --> 00:23:40.183 And then you get the drift. 00:23:40.238 --> 00:23:43.863 I mean honestly this conversation's inspiring me to, go look at our Ruby. 00:23:43.863 --> 00:23:46.383 I think we have a Ruby one and a JavaScript one, and 00:23:46.383 --> 00:23:47.343 I think that's probably it. 00:23:47.943 --> 00:23:51.663 Um, but I do know that the JavaScript one, it's like very opinionated 00:23:51.663 --> 00:23:53.193 and I think that's the challenge. 00:23:53.198 --> 00:23:55.383 It's like whoever wrote that one, that's how you. 00:23:56.388 --> 00:23:59.358 It's not necessarily how you might write it or how I might write it. 00:23:59.358 --> 00:24:03.558 And so there's like that kind of like an opinion about the developer 00:24:03.558 --> 00:24:07.713 experience instead of just saying taking a customer object, if there's a class 00:24:07.718 --> 00:24:11.913 for it, and being able to like add an invoice really easily and let it fill 00:24:11.913 --> 00:24:15.333 in all the parameters for customer and do all that stuff behind the scenes. 00:24:15.338 --> 00:24:19.053 Or do you have to go build each object by hand and then send it to the 00:24:19.053 --> 00:24:20.853 client in order to save those things? 00:24:21.258 --> 00:24:24.738 CJ: Right, and I think sometimes what's interesting about building this. 00:24:24.818 --> 00:24:29.538 generation layer is say that you make a choice early on when you're building 00:24:29.538 --> 00:24:34.818 the SDK that you want to have people interact with the API through like 00:24:34.998 --> 00:24:37.278 class methods on some resource, right? 00:24:37.278 --> 00:24:44.118 So you have like Stripe::Customer.Create and that takes in some a hash that will 00:24:44.118 --> 00:24:48.198 like pass those arguments to the api, but later you decide that, oh, it's actually 00:24:48.198 --> 00:24:54.033 way better if instead of having these class methods globally available, That we 00:24:54.033 --> 00:24:59.643 have a client where you can like new up and say like, Stripe know, stripe.new, and 00:24:59.643 --> 00:25:05.193 then you call like stripe.customers.Create And so what's crazy is like you can 00:25:05.193 --> 00:25:10.083 use the, these like generation layer to implement both patterns and then you can 00:25:10.083 --> 00:25:14.073 slowly start migrating people over to this new pattern or just enable both. 00:25:14.193 --> 00:25:15.423 Both like methods and. 00:25:17.103 --> 00:25:20.313 If you look right now, there's definitely parts of the stripe docs 00:25:20.313 --> 00:25:24.403 that are still one foot in the old world for like some of the SDKs, 00:25:24.963 --> 00:25:28.533 like php, where we had all of the. 00:25:29.768 --> 00:25:33.258 Documentation written as class methods, but the API ref and all 00:25:33.258 --> 00:25:35.748 of the auto-generated snippets are now migrated over to this 00:25:35.748 --> 00:25:37.488 like client and services pattern. 00:25:37.878 --> 00:25:40.098 And so that is definitely one of the benefits. 00:25:40.098 --> 00:25:45.198 But yeah, if, if your original authors took like one approach that 00:25:45.198 --> 00:25:49.788 you disagree with and you don't wanna break everyone, then it's ki it can 00:25:49.788 --> 00:25:56.178 be challenging to like, Release a new pattern and have everyone upgrade or, 00:25:56.282 --> 00:25:56.672 Colin: yeah. 00:25:56.822 --> 00:25:57.662 just alias. 00:25:57.692 --> 00:25:58.622 Alias everything. 00:25:58.652 --> 00:25:58.832 Yeah. 00:25:59.102 --> 00:26:00.072 CJ: Yeah, exactly. 00:26:00.122 --> 00:26:03.302 Colin: I think one of the considerations that I've heard with SDK generation 00:26:03.302 --> 00:26:06.277 is if you're not a rubus, for example, and a Java person writes 00:26:06.282 --> 00:26:07.747 a Ruby client library, right? 00:26:07.752 --> 00:26:09.787 It's probably not gonna follow. 00:26:10.247 --> 00:26:11.477 Ruby conventions. 00:26:11.567 --> 00:26:14.987 And so like, this is something to consider in your sstk generation, is that these 00:26:14.987 --> 00:26:19.407 tools probably should know a little bit about the standards in Java look 00:26:19.407 --> 00:26:21.417 like this and Python looks like this. 00:26:21.847 --> 00:26:24.157 Even just variable naming and stuff like that. 00:26:24.157 --> 00:26:28.087 Class names Camel Case, you know, snake Case, all those kinds of 00:26:28.087 --> 00:26:31.047 things where you can tell they're like, oh, a Java person wrote this. 00:26:31.047 --> 00:26:33.657 Or a JavaScript person type script, whatever it looks like. 00:26:33.687 --> 00:26:35.167 I'm just trying to think through like pros and cons here. 00:26:35.647 --> 00:26:39.477 It sounds like, you probably have more contained like testing 00:26:39.957 --> 00:26:41.187 and continuous deployment 00:26:41.697 --> 00:26:45.747 and being able to have failures and tests and all that kind of stuff along the way. 00:26:46.292 --> 00:26:46.682 CJ: totally. 00:26:46.682 --> 00:26:46.982 Yeah. 00:26:47.462 --> 00:26:52.282 So what's cool is that if you have patterns built into the api, then 00:26:52.342 --> 00:26:55.342 you can test your generation layer and be pretty confident that it's 00:26:55.342 --> 00:26:56.422 gonna spit out the right code. 00:26:56.422 --> 00:26:59.122 And if it's spitting out the wrong code, then you can fix it at the 00:26:59.122 --> 00:27:03.112 generation layer and that will fix it all, like in all the downstream places. 00:27:03.112 --> 00:27:06.442 So it fixes the docks and it fixes the SDKs and it fixes 00:27:06.442 --> 00:27:07.822 kind of everything ideally. 00:27:09.957 --> 00:27:13.042 But yeah, it's, it's also important I guess to mention that like we're 00:27:13.042 --> 00:27:19.942 taking the open API spec and by itself using the spec was really challenging 00:27:19.972 --> 00:27:25.612 in order to go from like this giant J s O object into all of these SDKs 00:27:25.617 --> 00:27:26.962 and client libraries and whatever. 00:27:27.082 --> 00:27:31.402 And so there is this like intermediate step that we call, 00:27:31.402 --> 00:27:34.882 like compiling the open API spec into this thing called Stripe type. 00:27:35.362 --> 00:27:39.622 And so I th we initially started with Flow, but I think we're using Ty, uh, 00:27:39.652 --> 00:27:46.522 type script now, where we take the, uh, the original j s o blob that came 00:27:46.522 --> 00:27:51.372 out of the open API spec, and then we build all of these different like 00:27:51.372 --> 00:27:56.352 discriminated types for what the high level resources should all look like. 00:27:56.712 --> 00:27:59.852 And then we use that as our input into prettier poet and our 00:27:59.852 --> 00:28:01.472 input into all these docs tools. 00:28:01.922 --> 00:28:03.792 That way we have this much. 00:28:04.682 --> 00:28:09.332 More like strongly typed and wired up tool because the, the Open API spec 00:28:09.332 --> 00:28:13.047 will actually give you back the resource and here's the methods, but also 00:28:13.052 --> 00:28:15.357 like down here is in a separate area. 00:28:15.357 --> 00:28:19.617 Here's the responses, and you kind of need to combine both of those in 00:28:19.622 --> 00:28:24.477 order to have all the information to implement a strongly typed language. 00:28:25.317 --> 00:28:26.877 Go and see Sharp or whatever. 00:28:26.877 --> 00:28:30.277 And there is that intermediate step and if people are interested in, learning 00:28:30.277 --> 00:28:35.687 more about that, Richard from our team did a talk at Strange Loop last year, and so 00:28:35.687 --> 00:28:37.577 we'll put a link to that, uh, on YouTube. 00:28:37.582 --> 00:28:40.587 And it's super interesting to watch, how that evolved. 00:28:40.767 --> 00:28:43.887 But yeah, when you were, when you were talking about how, like, like snake 00:28:43.887 --> 00:28:47.547 case versus Camel case versus whatever, have you ever used a Ruby motion? 00:28:47.757 --> 00:28:48.687 Ruby Motion, like. 00:28:48.907 --> 00:28:49.327 Colin: I'm. 00:28:49.762 --> 00:28:55.252 CJ: Okay, so there was, I think, 20 12, 20 13 or something, you could use 00:28:55.252 --> 00:28:59.932 Ruby to implement like iOS apps and it had like some tooling to spit out. 00:28:59.992 --> 00:29:02.872 I, it must have been like objective C or something like this, but it was so 00:29:02.872 --> 00:29:06.712 funny because the class ne like the class names would be like, you know, 00:29:06.802 --> 00:29:09.922 application on a knit with whatever. 00:29:10.282 --> 00:29:13.132 You know, and completion block or whatever, , like kind of 00:29:13.132 --> 00:29:16.792 the apple, the apple like, uh, objective seed name for the method. 00:29:16.792 --> 00:29:19.672 But it was just like in Ruby, so it was like deaf and then like 00:29:19.672 --> 00:29:22.352 this 85 character method name. 00:29:22.692 --> 00:29:26.432 But I think probably, uh, everyone who's been around for a while has 00:29:26.462 --> 00:29:29.852 encountered some SDK where you're like, this is not, this was not like 00:29:29.882 --> 00:29:34.112 implemented by Rubus based on all the different weird names and stuff that are. 00:29:35.522 --> 00:29:38.252 Colin: But I, what I like is that everything is then in code. 00:29:38.432 --> 00:29:41.642 You can commit, you can add to a change log. 00:29:41.642 --> 00:29:44.192 The change log goes out to all the different s STKs. 00:29:44.402 --> 00:29:47.402 So does that mean that you guys release, when you release an 00:29:47.402 --> 00:29:50.402 update to one s stk, you always release an update to all of them? 00:29:50.402 --> 00:29:54.032 Or is there some like, yeah, so there's not really like, oh, we're only gonna 00:29:54.032 --> 00:29:56.552 bump one because it affects all of them. 00:29:57.137 --> 00:29:59.987 CJ: Yeah, so we will release. 00:30:00.527 --> 00:30:05.957 So there are some changes that only impact the strongly typed languages because we do 00:30:05.957 --> 00:30:07.727 use some meta programming for some things. 00:30:07.727 --> 00:30:12.257 And so like if there is a change to an eum, like there's already an 00:30:12.262 --> 00:30:16.997 existing eum and we're just adding like a new value for that enum or 00:30:16.997 --> 00:30:19.217 something, then it only needs to change. 00:30:20.067 --> 00:30:26.247 Java go.net and the type definitions for Stripe node, but it doesn't need 00:30:26.247 --> 00:30:30.717 to change Ruby or Python or PHP because those are all just gonna say like, ah, you 00:30:30.717 --> 00:30:32.157 give me a string, you give me a string. 00:30:32.157 --> 00:30:35.062 It must work . Because they're not like checking types or anything, 00:30:35.392 --> 00:30:38.692 As soon as a feature is generally available, it kicks off this 00:30:38.692 --> 00:30:40.612 domino chain that is pretty wild. 00:30:40.612 --> 00:30:44.602 Like it will automatically generate like all the internal versions 00:30:44.602 --> 00:30:48.112 of it and tests and snapshot tests and like all of this stuff. 00:30:48.532 --> 00:30:51.022 And then it will also kick off the Slack message. 00:30:51.022 --> 00:30:54.142 And then the S STKs team goes and runs like generate, generate, 00:30:54.142 --> 00:30:55.882 generate, generate on all the SDKs. 00:30:55.882 --> 00:30:58.522 And that all makes prs and those prs get reviewed. 00:30:58.527 --> 00:31:01.522 And once those are merged, those all get deployed out to like, 00:31:01.522 --> 00:31:03.022 you know, Ruby Gems and npm. 00:31:03.592 --> 00:31:05.032 Maven and Gradle and whatever. 00:31:05.442 --> 00:31:09.142 It's interesting to be the person that's gonna tick the box that says 00:31:09.142 --> 00:31:10.342 this is generally available now. 00:31:10.347 --> 00:31:12.322 And it's like, you can't take that back very easily. 00:31:12.417 --> 00:31:15.207 Colin: Yeah, I was thinking like as a small company or a smaller company, 00:31:15.207 --> 00:31:17.847 it's like, whew, we're gonna be running a lot of, we're gonna run a lot of 00:31:17.847 --> 00:31:20.907 servers and spend a lot of money right now when we took this box and a lot of 00:31:21.387 --> 00:31:25.407 prs and CDs are gonna fire up and you're gonna get a bunch of servers running 00:31:25.407 --> 00:31:29.967 a bunch of code and now we got a bunch of PRS to review and yeah, it's pretty 00:31:30.042 --> 00:31:35.352 CJ: Yeah, that all that said, like I think it is really valuable and 00:31:35.352 --> 00:31:41.562 important to have features immediately available in the SDKs because that 00:31:41.562 --> 00:31:46.122 is like the main way that devs are using and feeling stripe, right? 00:31:46.122 --> 00:31:50.652 They're like, Their first impression is, okay, I installed Stripe, 00:31:50.652 --> 00:31:53.082 Ruby, and then I tried to make a customer or something like that. 00:31:53.082 --> 00:31:59.015 Like usually they're not, interacting with the Stripe API directly. 00:31:59.475 --> 00:32:03.835 The actual numbers of people who . Are using the Stripe API directly 00:32:03.840 --> 00:32:05.985 through official SDKs is massive. 00:32:05.985 --> 00:32:08.925 Like, it's surprising how many people use the SDKs instead 00:32:08.925 --> 00:32:11.085 of just making raw requests. 00:32:11.475 --> 00:32:14.445 Colin: Well, I mean, with Stripe there's so many requests that you 00:32:14.445 --> 00:32:15.705 need to chain together, right? 00:32:15.705 --> 00:32:17.085 To do some things. 00:32:17.505 --> 00:32:19.455 So I feel like that one makes a lot of sense. 00:32:19.460 --> 00:32:22.905 Cause a lot of people aren't used to, you know, carrying around all the 00:32:22.905 --> 00:32:25.395 responses and like, let's make this call. 00:32:25.395 --> 00:32:28.875 Like I've done the, like, you know, do the full process of checking 00:32:28.875 --> 00:32:30.575 somebody out in Postman, right? 00:32:30.575 --> 00:32:34.365 And it's like, okay, it's good to know how it works, but man, a library really. 00:32:35.025 --> 00:32:36.585 A lot more sense for this one, right? 00:32:36.915 --> 00:32:40.395 With Orbit, it's like, okay, sending an activity is just one call. 00:32:41.265 --> 00:32:45.855 Now talking and pulling data outta orbit requires a lot of chaining of APIs, 00:32:45.855 --> 00:32:47.805 calls to get IDs and stuff like that. 00:32:47.810 --> 00:32:50.715 And some of the things we're thinking about is like, how do we, I really like 00:32:50.715 --> 00:32:53.025 when APIs have like the side loading. 00:32:53.295 --> 00:32:57.015 I think you guys have that with yours where you can say like, This is an id, 00:32:57.015 --> 00:32:58.665 but I want it to be the full object. 00:32:58.665 --> 00:32:59.955 I don't want the ID of it. 00:32:59.955 --> 00:33:03.735 I want the customer with the invoice attached and those kinds of 00:33:03.735 --> 00:33:04.935 things I'd really like us to have. 00:33:04.935 --> 00:33:08.475 Cause right now it's like, here's an organization id, you go get the 00:33:08.475 --> 00:33:11.805 organization that it's attached to and make another call when we already 00:33:11.805 --> 00:33:13.305 know who that is and all that. 00:33:13.310 --> 00:33:15.950 I think the thing that we're not talking about here that we should probably talk 00:33:15.950 --> 00:33:20.410 about in a future episode is that like, This assumes you have an open API spec and 00:33:20.410 --> 00:33:22.690 that you have like a well designed api. 00:33:22.690 --> 00:33:24.550 Like this isn't designing your API for you. 00:33:24.550 --> 00:33:27.490 This is just the tooling around your api. 00:33:27.495 --> 00:33:33.160 And you know, I've been thinking of what RV two API I looks like so that we can 00:33:33.160 --> 00:33:39.850 get a, a different API spec cuz our a p is a J S O N api and it turns out J 00:33:39.850 --> 00:33:44.110 S O API is really good for like machine to machine computers understand it. 00:33:44.110 --> 00:33:45.460 I mean it's most of these right? 00:33:45.460 --> 00:33:46.630 Like a soap api. 00:33:47.300 --> 00:33:51.590 Has a wisdom, like you were talking about these X S D bindings, 00:33:51.680 --> 00:33:53.090 um, all those kinds of things. 00:33:53.090 --> 00:33:58.100 It's like teaches the computer what it can call, but does not necessarily 00:33:58.100 --> 00:34:02.785 always give the developer like a fun, light bulb experience of sending a, an 00:34:02.785 --> 00:34:06.595 SMS with Twilio or sending an email with SendGrid or whatever those things are. 00:34:06.595 --> 00:34:08.245 It's like it should be fun. 00:34:08.245 --> 00:34:10.405 And I, you know, I, I've always thought of APIs. 00:34:11.095 --> 00:34:14.485 Turning on a light bulb across the world, right, is it's just 00:34:14.485 --> 00:34:16.615 fun to have that as a superpower. 00:34:16.615 --> 00:34:21.175 And now we can charge a customer and stripe or, you know, send an 00:34:21.175 --> 00:34:24.385 activity to segment and have it populate everywhere we need it. 00:34:25.105 --> 00:34:28.735 CJ: So I was curious and I, I jumped in, I did write a couple of orbit 00:34:29.275 --> 00:34:35.935 integrations, like one for piping in some dev.to blog post activity, and 00:34:35.935 --> 00:34:37.495 another one for piping in YouTube. 00:34:38.360 --> 00:34:41.056 and I was looking at like, oh, did I just hit the a p I 00:34:41.526 --> 00:34:45.316 directly or did I use the S D K? 00:34:45.346 --> 00:34:48.736 And in one of them I used the, the Orbit s d K, the other one I didn't 00:34:49.696 --> 00:34:55.216 So yeah, I'm using Orbit activities request and then passing in just 00:34:55.216 --> 00:34:57.886 like a giant blob of arguments and the body and everything. 00:34:58.521 --> 00:34:58.761 Colin: Mm-hmm. 00:34:58.936 --> 00:35:02.956 CJ: And then in the other one, I'm using REST client to just hit, uh, 00:35:03.886 --> 00:35:06.616 like orbit.love/api v1 or whatever. 00:35:06.631 --> 00:35:07.051 Colin: Nice. 00:35:07.801 --> 00:35:08.101 Yeah. 00:35:08.101 --> 00:35:11.371 I mean, it's something that we all have to make decisions when we, we actually have 00:35:11.371 --> 00:35:15.581 our own Orbi HTTP client now, in Ruby and. 00:35:16.451 --> 00:35:17.771 I think it'll be interesting. 00:35:17.771 --> 00:35:18.761 I'll have to see if. 00:35:19.556 --> 00:35:22.826 Coworker Steve has any interest in eventually opening that up. 00:35:22.826 --> 00:35:26.036 But it has a lot of that rate limiting stuff built into 00:35:26.036 --> 00:35:27.416 it like we had with Discord. 00:35:27.956 --> 00:35:32.876 We sometimes will use official SDKs for our integrations, but sometimes, 00:35:32.876 --> 00:35:38.126 like if our API is only going to ingest data and never create data, 00:35:38.586 --> 00:35:42.756 some APIs don't give you API keys that are read only, things like that. 00:35:42.756 --> 00:35:44.016 So like we're like, we only. 00:35:44.841 --> 00:35:46.461 The endpoints that get the data. 00:35:46.461 --> 00:35:51.711 We do not want to even have a method that lets us delete or 00:35:51.711 --> 00:35:53.361 create or any of those things. 00:35:53.811 --> 00:35:59.091 Um, but also just sometimes like, just being able to standardize, re responses, 00:35:59.421 --> 00:36:03.231 standardize what happens after, cuz every sdk like we're talking about here, 00:36:03.591 --> 00:36:07.791 the experience, the opinions of the developer team that put together the sdk. 00:36:07.791 --> 00:36:12.396 So like, Because of how we built out our newer integrations. 00:36:12.726 --> 00:36:16.536 They almost all use that Orbit HTTP client, which I think is 00:36:16.536 --> 00:36:17.826 a wrapper around rest client. 00:36:17.826 --> 00:36:21.396 But there's just like a bunch of really nice helpers and it's more declarative. 00:36:21.396 --> 00:36:25.356 So you can just say these are the end points and we're not rewriting. 00:36:25.361 --> 00:36:27.546 Cause that's like the challenge when you do a lot of integrations 00:36:27.546 --> 00:36:29.076 is that you're doing the same thing. 00:36:29.666 --> 00:36:33.846 And translating it to every API , and being able to have some standards 00:36:33.846 --> 00:36:36.846 once the response comes back, we know what it looks like every 00:36:36.846 --> 00:36:38.586 single time, which is really nice, 00:36:38.636 --> 00:36:42.486 CJ: So this is the client that you're using to integrate against other 00:36:42.486 --> 00:36:44.226 third parties, not against the orbit? 00:36:44.226 --> 00:36:44.466 Right. 00:36:44.466 --> 00:36:44.736 Okay. 00:36:44.736 --> 00:36:51.981 And then are, gosh, I, I almost wonder if it's possible to like, Take the third 00:36:51.981 --> 00:36:55.221 parties and see like, oh, do all of them have open API specs or something 00:36:55.221 --> 00:36:57.321 like that, that you could then use to 00:36:57.321 --> 00:36:58.341 build that? 00:36:58.346 --> 00:36:58.851 Like, yeah, 00:36:58.851 --> 00:37:02.601 the, so like you have your H C D B client, and then you could build like that 00:37:02.601 --> 00:37:05.901 middle layer that we've talked so much about, which is like the adapter layer. 00:37:06.441 --> 00:37:12.081 And I, it sounded like you had a pretty cool, uh, like pipeline 00:37:12.316 --> 00:37:12.926 Colin: Mm-hmm. 00:37:13.191 --> 00:37:13.611 CJ: Right. 00:37:13.731 --> 00:37:14.091 But. 00:37:15.111 --> 00:37:15.261 Colin: yeah. 00:37:15.261 --> 00:37:20.361 Each integration essentially ends up with its own like Stripe client for 00:37:20.361 --> 00:37:23.061 example, that uses the Orbit HTTP client. 00:37:23.061 --> 00:37:27.491 And it's just the methods we need and because everything is built on 00:37:27.491 --> 00:37:30.521 that client, we don't have to go implement everything all over again. 00:37:30.521 --> 00:37:35.291 It's just really like, is to get, what's the endpoint slash customers or whatever. 00:37:35.418 --> 00:37:39.588 Obviously once you release something as like a public gem or whatever, now 00:37:39.588 --> 00:37:44.148 you comes with a burden of supporting it and documenting it and taking issues 00:37:44.148 --> 00:37:45.708 and prs and all that kind of stuff. 00:37:45.708 --> 00:37:49.248 So do you, you guys, I'm assuming, don't take PRS necessarily 00:37:49.248 --> 00:37:50.988 against these SDKs then, right? 00:37:52.053 --> 00:37:55.653 CJ: So there are parts of them that are not generated, 00:37:55.773 --> 00:37:57.693 and we will take prs on that. 00:37:57.698 --> 00:37:59.853 It's kind of like, I, I refer to that as like the core. 00:37:59.853 --> 00:38:05.043 So if it's, if it is that sort of, um, item potency or retries layer or 00:38:05.043 --> 00:38:10.083 interacting with, you know, SSL or, um, the patterns, like, oh, is this a client? 00:38:10.413 --> 00:38:12.483 Versus client services or class method or whatever. 00:38:12.573 --> 00:38:14.913 We will take PRS to modify that core. 00:38:15.433 --> 00:38:19.073 But all of the pieces that are boiler plate, that mirror the api, those are 00:38:19.073 --> 00:38:20.303 all generated and it wouldn't make 00:38:20.303 --> 00:38:21.563 sense for anyone to make a PR 00:38:22.403 --> 00:38:25.673 Colin: But they can record an issue so you guys can go check it out and see like, 00:38:25.703 --> 00:38:27.233 oh yeah, that's actually not correct, 00:38:27.773 --> 00:38:28.403 CJ: Exactly. 00:38:28.433 --> 00:38:32.153 And I would love to see us actually move the generation code 00:38:32.153 --> 00:38:33.923 into the, into the open source. 00:38:34.688 --> 00:38:35.408 Libraries. 00:38:35.408 --> 00:38:37.568 The generation code right now is all private sourced. 00:38:37.988 --> 00:38:42.954 And I was looking at, AW s gem, I think this might be the most popular Ruby gem 00:38:43.404 --> 00:38:44.784 by downloads or something like that. 00:38:45.294 --> 00:38:49.284 Uh, and they also generate this client. 00:38:49.644 --> 00:38:52.734 But they're doing it in a generator that's open source and 00:38:52.734 --> 00:38:55.234 it's like part of the library. 00:38:55.324 --> 00:38:56.854 And it was interesting to look at. 00:38:56.854 --> 00:38:59.974 They have basically these giant mustache templates and then they 00:38:59.974 --> 00:39:04.654 iterate over some input, some input that's like similar to open API spec. 00:39:04.714 --> 00:39:07.264 They don't have this intermediate stripe type. 00:39:07.424 --> 00:39:08.714 Style compiler thing. 00:39:08.714 --> 00:39:14.414 But, uh, I do love to see the generator there because as an open source 00:39:14.414 --> 00:39:17.564 contributor, you could make a change to the generator that's like, oh, it would 00:39:17.564 --> 00:39:22.184 be better if it was like this, or it'd be more perf like performant if we use, I 00:39:22.184 --> 00:39:24.674 don't know, yield instead of block.call. 00:39:24.704 --> 00:39:27.584 You know, I don't know, like the actual, you know, Ruby, 00:39:27.644 --> 00:39:29.204 whatever, magic stuff, but. 00:39:29.644 --> 00:39:33.484 Colin: And someone might learn from this, uh, you know, AW S one and start 00:39:33.484 --> 00:39:35.584 generating their own SDKs for their own. 00:39:35.584 --> 00:39:38.284 Like, we, we would just end up with better SDKs all around. 00:39:38.284 --> 00:39:38.524 Right? 00:39:38.524 --> 00:39:43.354 It's like putting it out there so that more people do this as a, as a, as a norm. 00:39:43.354 --> 00:39:44.794 Instead of it being like, you do what 00:39:45.499 --> 00:39:45.799 CJ: Yeah. 00:39:46.759 --> 00:39:51.079 , it's super hard to explain all of the benefits of Prettier poet 00:39:51.079 --> 00:39:53.239 in like the way that it looks and feels and the shape of it. 00:39:53.239 --> 00:39:56.869 So, . I'm hopeful that, uh, if this is interesting to you or you're 00:39:56.869 --> 00:39:59.839 inspired, you go, he head over and like check out the talk on YouTube, 00:40:00.269 --> 00:40:00.659 Colin: Nice 00:40:00.719 --> 00:40:02.339 CJ: there will be like code examples and such. 00:40:03.149 --> 00:40:04.629 Colin: is Prettier poet open. 00:40:05.309 --> 00:40:09.059 CJ: Prettier poet is not open and that's another thing that I'm like hopeful 00:40:09.059 --> 00:40:10.709 that we will eventually open source. 00:40:11.159 --> 00:40:11.304 Colin: Yeah. 00:40:11.724 --> 00:40:12.174 So you're 00:40:12.174 --> 00:40:12.954 you're here in Nevada. 00:40:13.044 --> 00:40:16.314 Here first, you know, or if you're at Ruby Comp, you'll, you'll get 00:40:16.344 --> 00:40:17.544 to hear some of this as well. 00:40:17.784 --> 00:40:20.694 That's, uh, Ruby Comp is the one that's in Texas. 00:40:20.724 --> 00:40:20.934 Is that 00:40:21.009 --> 00:40:21.579 CJ: Yes. 00:40:21.699 --> 00:40:25.869 I'm gonna go to the, yeah, Ruby Comp Texas, because the Ruby Comp Mini, that 00:40:25.869 --> 00:40:28.689 was just in Rhode Island right down the street, was gonna be the same time 00:40:28.689 --> 00:40:30.549 as the develop like chirp developer 00:40:30.549 --> 00:40:31.659 conference that we just missed. 00:40:31.659 --> 00:40:32.019 So 00:40:32.079 --> 00:40:32.829 Colin: Thanks Elon. 00:40:32.859 --> 00:40:33.459 Thanks Elon. 00:40:35.319 --> 00:40:35.709 Awesome. 00:40:35.709 --> 00:40:38.529 Well, I think we're at time for today, so it's awesome to kind of 00:40:38.529 --> 00:40:41.409 dive into SDKs and all the stuff you guys are doing at Stripe. 00:40:41.919 --> 00:40:45.579 I'm definitely gonna go take a, take a peek at our SDKs after this and 00:40:45.579 --> 00:40:47.289 think about what we might need to do. 00:40:47.749 --> 00:40:50.319 you said you wrote them for seven years, were not. 00:40:50.889 --> 00:40:53.919 At that point, like we're probably gonna be handwriting ours for a while, 00:40:53.919 --> 00:40:57.699 but at least getting some eyes on our s STKs and just thinking about how we're 00:40:57.699 --> 00:40:59.199 doing that would be, would be good. 00:40:59.649 --> 00:41:00.319 CJ: totally. 00:41:00.779 --> 00:41:04.439 The show notes for this episode are available@buildandlearn.dev. 00:41:04.769 --> 00:41:06.269 That wraps it for this episode. 00:41:06.269 --> 00:41:06.839 Thanks everyone. 00:41:07.449 --> 00:41:08.029 Colin: See ya.