WEBVTT NOTE This file was generated by Descript 00:00:00.000 --> 00:00:02.830 Phil Sturgeon: Hello everybody and welcome to another episode of 00:00:03.420 --> 00:00:07.990 APIs You Won't Hate With me, Phil Sturgeon and no Mike, the co-host. 00:00:08.090 --> 00:00:11.350 So I'm still muddling through trying to figure out what buttons I meant to 00:00:11.350 --> 00:00:13.629 press, but slowly get in the hang of it. 00:00:14.129 --> 00:00:16.190 And thankfully, I'm joined today by Lorna Mitchell. 00:00:17.240 --> 00:00:17.700 Say hello, 00:00:17.700 --> 00:00:17.919 Lorna Mitchell: Hi. 00:00:18.410 --> 00:00:18.760 Hello. 00:00:18.810 --> 00:00:19.680 Happy to be here. 00:00:20.450 --> 00:00:20.740 Track 1: Nice. 00:00:20.740 --> 00:00:21.220 How's it going? 00:00:21.300 --> 00:00:22.740 I think you've been on the podcast before. 00:00:22.820 --> 00:00:26.580 I meant to actually look that up back in the day talking about something or other, 00:00:26.880 --> 00:00:27.460 but if not. 00:00:27.470 --> 00:00:30.070 Lorna Mitchell: I actually don't think I have, I, I have written 00:00:30.290 --> 00:00:31.870 for APIs you won't hate, and 00:00:31.910 --> 00:00:32.630 I like to write. 00:00:32.870 --> 00:00:33.590 I should do that again. 00:00:34.170 --> 00:00:36.190 I'm not sure we've done the podcast thing before. 00:00:36.410 --> 00:00:38.850 Track 1: alright, well either way welcome. 00:00:39.590 --> 00:00:43.650 And for folks who dunno who you are, could you describe a little bit about yourself? 00:00:44.440 --> 00:00:44.730 Lorna Mitchell: Sure. 00:00:44.890 --> 00:00:45.570 I am Lorna. 00:00:45.850 --> 00:00:49.410 I am currently VP of Developer experience at Redley. 00:00:49.740 --> 00:00:53.780 But my background is I am been a software engineer for a long time. 00:00:54.360 --> 00:00:55.260 I'm a conference speaker. 00:00:55.640 --> 00:01:00.260 I'm a published author and I am mad passionate about APIs and open source. 00:01:01.550 --> 00:01:04.650 Track 1: Ha, sorry, I was just doing some research live on a podcast like 00:01:04.690 --> 00:01:07.460 a real professional . Fantastic. 00:01:07.850 --> 00:01:10.660 Yeah, no, your, your name pops up everywhere in open API. 00:01:11.000 --> 00:01:16.300 And yeah, you were pretty helpful in the open API 3.1 stuff when we were trying 00:01:16.300 --> 00:01:17.420 to get all of that across the line. 00:01:17.760 --> 00:01:18.340 And what was it? 00:01:19.145 --> 00:01:20.395 What was the main thing you were working on? 00:01:20.435 --> 00:01:23.475 I completely forgot because I, my whole thing was like, Jason Schema needs to 00:01:23.475 --> 00:01:25.435 match open API or what are we even doing? 00:01:25.815 --> 00:01:27.155 But you were working on other bits? 00:01:27.865 --> 00:01:30.675 Lorna Mitchell: Yeah, so for 3.3 0.1, which is when I really came 00:01:30.705 --> 00:01:33.875 into working on open API itself, I was working on the webhooks thing. 00:01:34.415 --> 00:01:39.475 That's when I was at Vonage and with their two-way messaging APIs, you need 00:01:39.475 --> 00:01:41.075 webhooks callbacks are not the same. 00:01:41.495 --> 00:01:45.155 So that was I in my not humble at all opinion. 00:01:45.375 --> 00:01:48.515 One of the best things about the 3.1 release of open API. 00:01:49.230 --> 00:01:49.550 Track 1: Absolutely. 00:01:49.750 --> 00:01:50.550 I, I forgot that. 00:01:50.550 --> 00:01:52.710 Was that you that helped push that through and yeah. 00:01:52.710 --> 00:01:56.840 We're using it, using it right now for some webhook stuff 00:01:56.840 --> 00:01:57.800 that protect Earth's doing. 00:01:57.980 --> 00:01:59.840 So yeah, it, it's really nice that that's there. 00:02:00.000 --> 00:02:02.520 'cause it was always a bit weird trying to hack that into callbacks. 00:02:02.520 --> 00:02:03.000 You'd have to like. 00:02:03.760 --> 00:02:07.290 Pretend you'd have to make some sort of fake URL called like web 00:02:07.290 --> 00:02:10.410 hooks or something, and then just kind of pretend what it was for. 00:02:10.430 --> 00:02:14.050 But no, this is, it's useful to be able to say, sometimes there are 00:02:14.050 --> 00:02:16.970 messages that are gonna come from this server that aren't simply in 00:02:17.170 --> 00:02:18.690 response to a request you just made. 00:02:18.690 --> 00:02:22.010 You can register for a web hook through the website or whatever. 00:02:22.170 --> 00:02:23.410 I mean, Shopify's super weird. 00:02:23.410 --> 00:02:24.850 We're we're registering for webhooks. 00:02:25.210 --> 00:02:27.890 With them through like a CLA command, which is awkward. 00:02:28.190 --> 00:02:32.050 But then it will just forever spit web hooks at us on, on topics that 00:02:32.050 --> 00:02:33.250 aren't related to a single thing. 00:02:33.270 --> 00:02:35.050 So it's really handy to have that in there. 00:02:35.750 --> 00:02:37.120 Lorna Mitchell: Yeah, I think there's loads of use cases. 00:02:37.340 --> 00:02:39.560 You know, callbacks gave us half of it. 00:02:39.900 --> 00:02:43.920 You know, it's in response to an incoming request, but webhooks lets 00:02:43.920 --> 00:02:47.570 you have that payload going across in response to something else. 00:02:48.650 --> 00:02:48.940 Track 1: Yeah. 00:02:49.340 --> 00:02:49.620 Brilliant. 00:02:50.320 --> 00:02:53.100 And so you've been involved in loads of stuff with Open api. 00:02:53.100 --> 00:02:55.040 I, and obviously you're working for Redoc re Dock Lee now. 00:02:55.420 --> 00:02:59.560 But yeah, we, we, we've both been kind of in the same spaces working on trying 00:02:59.560 --> 00:03:01.840 to make like a sidekick for open API. 00:03:01.840 --> 00:03:02.040 Right. 00:03:02.280 --> 00:03:02.480 I did 00:03:02.480 --> 00:03:04.960 that recent blog post about how Redoc Lee has kind of got there 00:03:04.980 --> 00:03:08.520 now, but you were around in, in Specky trying to make that useful. 00:03:09.030 --> 00:03:10.290 And, and spectral. 00:03:10.510 --> 00:03:13.770 And can you just talk a little bit about those tools and where you feel 00:03:13.770 --> 00:03:17.130 like they fell short and you know, what they're good at, what they're bad at? 00:03:17.130 --> 00:03:19.970 And then, and then what's going on with Redoc Lee? 00:03:20.190 --> 00:03:22.250 And its kind of open API psychic now. 00:03:22.950 --> 00:03:27.520 Lorna Mitchell: Yeah, no, I think, I think there are so many great API tools now, but 00:03:27.630 --> 00:03:32.160 it's not as if they just came out of thin air now, like they're ready to use now. 00:03:32.490 --> 00:03:35.380 Like you say specky I think was one of the first ones that at least 00:03:35.580 --> 00:03:40.180 I knew of, which really helped me to validate and check things. 00:03:40.360 --> 00:03:42.220 I'm not sure I was customizing it a lot. 00:03:42.320 --> 00:03:45.580 And so that can be quite frustrating if you're using out of the box rules. 00:03:45.960 --> 00:03:49.000 I think too many people do that and then it can be a hard experience. 00:03:49.755 --> 00:03:50.045 Track 1: Yeah, 00:03:50.125 --> 00:03:53.285 I think specifically I think you were the person who, well, a lot of 00:03:53.285 --> 00:03:56.565 people really hated all of the default rules about like tags and stuff. 00:03:56.775 --> 00:03:59.215 But I seem to remember you giving a bunch of feedback about that. 00:03:59.215 --> 00:04:02.575 Like why is it constantly complaining that I need to use tags, I don't actually 00:04:02.575 --> 00:04:04.815 need to use tags, and that was just in 00:04:04.815 --> 00:04:05.535 there for some reason. 00:04:05.645 --> 00:04:07.405 Lorna Mitchell: I, I, think it was an API with like two end points. 00:04:07.485 --> 00:04:09.125 I, I really don't need to tag them, but thanks. 00:04:09.775 --> 00:04:12.315 And then we used Spectral a lot. 00:04:12.795 --> 00:04:14.195 I realized the other you'll love this. 00:04:14.675 --> 00:04:19.275 I realized the other day that my personal blog, so that's Lorna 00:04:19.275 --> 00:04:20.395 jane.net will put it in the show notes. 00:04:20.575 --> 00:04:24.315 My personal blog has instructions of how to link APIs with spectral. 00:04:25.275 --> 00:04:29.255 But it does not have how to link them with Reduc, CLI, which is, that's what I do 00:04:29.255 --> 00:04:29.455 now. 00:04:29.630 --> 00:04:30.880 Track 1: time for an updated blog post. 00:04:31.000 --> 00:04:31.240 I think 00:04:31.705 --> 00:04:32.375 Lorna Mitchell: definitely. 00:04:32.645 --> 00:04:34.615 Well, and we'll just write all the blog posts for 00:04:34.615 --> 00:04:36.664 all the tools because I think they're all great. 00:04:36.724 --> 00:04:40.985 But yeah, I used Spectral and that's when I started customizing rules. 00:04:41.145 --> 00:04:45.185 I was working for Vonage and what's interesting about this is I think 00:04:45.224 --> 00:04:50.680 I did Specky when I was working for Doing like integration consultancy. 00:04:51.240 --> 00:04:53.360 I, I worked on spectral when I was at Vonage. 00:04:53.360 --> 00:04:57.680 They're an API provider, so that, but they have a lot of APIs and they're 00:04:57.680 --> 00:04:58.840 quite different from each other. 00:04:59.380 --> 00:05:02.240 So working on spectral there and making sure that we're 00:05:02.750 --> 00:05:04.280 enforcing that consistency. 00:05:04.340 --> 00:05:07.600 And, you know, I'm not sure developer experience was a buzzword then, 00:05:08.340 --> 00:05:10.040 but that's how I would call it now. 00:05:11.020 --> 00:05:11.479 In between. 00:05:11.940 --> 00:05:17.379 I worked at a cool, another cool Cloud SaaS and they had an API and that's when 00:05:17.379 --> 00:05:23.150 I started using Redoc Lee tools because I really enjoyed their linting and of 00:05:23.150 --> 00:05:25.469 course Redoc API reference documentation. 00:05:26.369 --> 00:05:33.830 So now I'm out redly working on Redoc Lee, CLI, which as you say is the API sidekick. 00:05:34.049 --> 00:05:38.119 And I think I'm still seeing some of those same problems like. 00:05:38.730 --> 00:05:41.230 People use the recommended rule set and then tell us it's wrong. 00:05:41.725 --> 00:05:42.855 it's, you need to cus 00:05:42.934 --> 00:05:43.224 Track 1: Yeah. 00:05:43.224 --> 00:05:43.424 Yeah. 00:05:43.635 --> 00:05:44.775 Lorna Mitchell: 1 0 1 run it. 00:05:44.775 --> 00:05:47.495 We'll give you some good feedback, you know, and I'm seeing some more 00:05:47.495 --> 00:05:49.855 of those upload your API and we'll tell you what you've done wrong. 00:05:50.284 --> 00:05:53.615 Like Yeah, but every API is different, so you need to customize 00:05:53.615 --> 00:05:55.935 your rule set and use it yourself. 00:05:56.635 --> 00:05:59.455 So Redly CLI is open source. 00:05:59.835 --> 00:06:03.615 It does do the linting, but it does a couple of things that I would describe 00:06:03.615 --> 00:06:06.125 as Either side of that in the process. 00:06:07.025 --> 00:06:14.534 One is it has functionality for splitting your open API up into multiple 00:06:14.534 --> 00:06:18.385 files so you don't have to deal with a hundred thousand lines of diff. 00:06:19.034 --> 00:06:22.174 You can, and also you can look at like, which segments changed and 00:06:22.174 --> 00:06:23.655 that can be easier for review. 00:06:24.354 --> 00:06:25.804 It's also easier on your editor. 00:06:26.504 --> 00:06:27.484 And then you 00:06:27.484 --> 00:06:27.684 can 00:06:27.835 --> 00:06:30.955 Track 1: yeah, whenever you are trying to like send a pull request on a file that's 00:06:31.455 --> 00:06:35.434 10,000 lines long, firstly it's just not gonna show you the preview half the time. 00:06:35.694 --> 00:06:37.815 And secondly, it's more, more likely to conflict. 00:06:38.015 --> 00:06:40.174 'cause two people worked on something completely different, but it 00:06:40.174 --> 00:06:42.335 didn't realize it was completely different and therefore, but, 00:06:42.875 --> 00:06:44.005 Lorna Mitchell: Yeah, and it's harder 00:06:44.005 --> 00:06:46.045 to reason about which sections should have changed 00:06:46.635 --> 00:06:47.885 when it's just one long thing. 00:06:48.725 --> 00:06:51.245 Track 1: I think before, before we go past it too quickly. 00:06:51.405 --> 00:06:54.205 I think one of the most impressive things, so when I was doing the 00:06:54.205 --> 00:06:57.105 review of re doley I was like, okay, yeah, I know what this is about. 00:06:57.105 --> 00:06:59.425 They got like a liny thing and they got like a Bundy thing and 00:06:59.425 --> 00:07:00.745 they got like a preview docs. 00:07:00.745 --> 00:07:02.625 These are all good, these aren't new things. 00:07:02.625 --> 00:07:05.825 They've been kind of brought together in a, in, in one new tool that does a 00:07:05.825 --> 00:07:06.705 whole bunch of things, which is great. 00:07:07.420 --> 00:07:11.680 But what I was really, really excited about discovering was the split thing. 00:07:12.080 --> 00:07:14.360 'cause the number of times, like I'm looking at some massive open 00:07:14.480 --> 00:07:18.240 API file going, ugh, like it's been generated from HCTP or whatever. 00:07:18.610 --> 00:07:21.135 You're kinda like just, you're just staring at that massive far going. 00:07:21.205 --> 00:07:22.375 This is gonna take some time. 00:07:22.875 --> 00:07:26.575 And I think like stoplight studio at some point had vaped some functionality 00:07:26.635 --> 00:07:27.975 in like, move this to a model. 00:07:28.315 --> 00:07:31.895 But then the button broke and it was disabled, and I begged and begged and 00:07:31.895 --> 00:07:34.135 begged to get that button put back in and they were, oh, we're too busy. 00:07:34.255 --> 00:07:38.175 I was like, no, because taking a, taking a giant opiate open, API file. 00:07:38.410 --> 00:07:40.580 Like, I feel like no one's doing enough refactoring, 00:07:40.990 --> 00:07:44.380 automating, refactoring, like move this to move this to models. 00:07:44.469 --> 00:07:47.469 These two, these two in place things look the same. 00:07:47.759 --> 00:07:48.879 Change that to one ref. 00:07:48.939 --> 00:07:51.159 And I think, you know, split, split does a lot of that. 00:07:51.159 --> 00:07:54.159 They, they noticed that you've got refs and they were like bung 'em off 00:07:54.159 --> 00:07:57.359 into files and it did also notice that two things were the same thing 00:07:57.699 --> 00:07:58.999 and put them into the same file. 00:07:59.019 --> 00:08:00.519 And I was like, God damn, that's useful. 00:08:00.939 --> 00:08:03.199 So I was really glad to see that even though it's like a, 00:08:03.399 --> 00:08:05.279 a undersung kind of feature. 00:08:05.429 --> 00:08:05.719 Yeah. 00:08:06.379 --> 00:08:06.799 Lorna Mitchell: it is. 00:08:06.939 --> 00:08:12.469 But I think it does really help, and particularly As we mature in our API 00:08:12.749 --> 00:08:16.829 practice, it's not just about, I wrote an open API file or I generated one. 00:08:16.829 --> 00:08:17.309 There you go. 00:08:17.999 --> 00:08:22.809 I've done that tick You know, those are living, changing documents and so 00:08:22.809 --> 00:08:24.489 the maintenance aspect gets quite big. 00:08:24.989 --> 00:08:30.049 So yeah, the, I think the, the splitting bundling and we also have join, so where 00:08:30.129 --> 00:08:31.809 people are maintaining different, every 00:08:31.809 --> 00:08:36.449 team maintains an API each, but you publish it as a single surface 00:08:37.269 --> 00:08:38.449 that's actually really messy. 00:08:38.519 --> 00:08:39.889 I've seen some horrible merge. 00:08:40.574 --> 00:08:41.694 Attempts and 00:08:41.694 --> 00:08:44.694 join attempts to handle that, which I think is really nice. 00:08:44.694 --> 00:08:48.924 So from an API management point of view, the other thing that I love about Reduc, 00:08:49.124 --> 00:08:56.084 CLI, which I am now seeing in other places, but Reduc CLI has had it for a 00:08:56.084 --> 00:08:58.804 long time, is decorators the ability to 00:08:59.174 --> 00:09:01.684 repeatably transform something. 00:09:02.979 --> 00:09:07.179 an open API document so you get something it's not perfect or 00:09:07.179 --> 00:09:10.339 you want to add something for the next stage in your pipeline. 00:09:10.950 --> 00:09:12.630 There's a decorator's set. 00:09:12.630 --> 00:09:16.230 There are some built-in decorators, but you can also extend it by saying 00:09:16.590 --> 00:09:22.150 whenever you see a, I dunno, an an operation parameter, do this. 00:09:23.550 --> 00:09:23.790 Track 1: Interesting. 00:09:23.820 --> 00:09:24.750 Like, like what? 00:09:24.940 --> 00:09:27.390 Make it more tangible for my struggling brain, 00:09:28.245 --> 00:09:30.505 Lorna Mitchell: So we do things like you can add, you can 00:09:30.505 --> 00:09:31.545 replace all the descriptions. 00:09:31.975 --> 00:09:34.025 This, this operation Id should have this description. 00:09:34.025 --> 00:09:35.265 This one should have this description. 00:09:35.485 --> 00:09:35.825 So if 00:09:35.825 --> 00:09:41.255 you're getting poor quality generated content or you wanna add, you know, 00:09:41.314 --> 00:09:46.214 the descriptions in open API support markdown, so you can add really 00:09:46.214 --> 00:09:48.535 beautiful, rich links and whatever. 00:09:49.035 --> 00:09:51.415 Do you want to maintain that in your source code? 00:09:51.964 --> 00:09:54.814 When you are generating from that, you know, 00:09:54.895 --> 00:09:55.555 Track 1: Oh, I see. 00:09:56.064 --> 00:09:56.354 Yeah. 00:09:57.034 --> 00:10:01.055 Lorna Mitchell: you are maintaining the YAML for the markdown in your source code. 00:10:01.294 --> 00:10:04.214 I just, I do understand why people don't enjoy that. 00:10:05.094 --> 00:10:05.374 Track 1: Absolutely. 00:10:05.374 --> 00:10:05.614 Yeah. 00:10:05.734 --> 00:10:11.694 I think it's the, the people that advocate most for the kind of code first B it all 00:10:11.694 --> 00:10:17.134 into annotations in our code approach are the people who aren't ever gonna put very 00:10:17.134 --> 00:10:18.775 much effort into their documentation. 00:10:18.775 --> 00:10:23.555 And therefore you know, you're not, you're not gonna see these like . Hundred lines 00:10:23.555 --> 00:10:27.514 of markdown showing up in the source code purely because they wouldn't bother 00:10:27.514 --> 00:10:30.754 to write it and the tech writers can't be faf to go and like get in there. 00:10:31.199 --> 00:10:31.419 Um, 00:10:31.774 --> 00:10:32.854 Lorna Mitchell: Not even have access. 00:10:33.474 --> 00:10:34.374 That's really common. 00:10:34.434 --> 00:10:38.174 You get frustrated tech writers who could add really great, 00:10:38.474 --> 00:10:39.694 you know, descriptive content. 00:10:39.884 --> 00:10:40.814 Example values. 00:10:41.424 --> 00:10:44.324 If it's generated from source code and not, and they're not giving 00:10:44.344 --> 00:10:47.044 access to the source code, it's very difficult to include that. 00:10:47.584 --> 00:10:49.244 And we're starting to see tools as well. 00:10:49.244 --> 00:10:53.444 Obviously the documentation tools, Redoc and friends are the very good 00:10:53.524 --> 00:10:55.124 API Documentation tools are available. 00:10:57.144 --> 00:10:58.144 Track 1: Standard BBC response. 00:10:58.214 --> 00:10:58.504 Yeah. 00:10:58.714 --> 00:11:01.524 Lorna Mitchell: Well, I just feel like, I don't mind which one you use, but please 00:11:01.524 --> 00:11:03.124 give me some documentation, you know, 00:11:03.155 --> 00:11:03.444 Track 1: Yeah. 00:11:03.474 --> 00:11:03.764 Yeah. 00:11:03.834 --> 00:11:05.854 Lorna Mitchell: And the SDK generators as well. 00:11:05.914 --> 00:11:09.859 You know, we've just seen speakeasy release the overlays tool, Because 00:11:10.029 --> 00:11:14.539 their tool does better with some added metadata and you don't wanna 00:11:14.819 --> 00:11:17.979 maintain that in wherever upstream your open API. 00:11:17.979 --> 00:11:18.939 Things are coming from. 00:11:19.704 --> 00:11:20.054 Track 1: Great. 00:11:20.124 --> 00:11:21.174 Well that brings us on nicely. 00:11:21.194 --> 00:11:23.374 So I've been meaning to pick your brain about overlays a bit 00:11:23.574 --> 00:11:27.294 'cause it's, it's something that there's increasing interest in. 00:11:27.374 --> 00:11:30.254 I mean, I, I work with the, the folks at Bump as well, and they were asking 00:11:30.264 --> 00:11:33.674 about, about a bit more about overlays and there's a lot of people in the 00:11:33.674 --> 00:11:36.754 docs space who were interested in that idea for exactly what you're saying. 00:11:36.824 --> 00:11:41.154 Like, if you know the, the code first, people dump out some open API. 00:11:41.389 --> 00:11:42.729 And it's whatever it is. 00:11:42.829 --> 00:11:46.049 And if you edit it, then it's just, you know, gone the next time they dump it 00:11:46.049 --> 00:11:46.249 out. 00:11:46.629 --> 00:11:49.769 Or you've got the people that are design first where the, the YAML is the source 00:11:49.769 --> 00:11:53.329 of truth and they do all their edits and, and that's that, and that's probably a 00:11:53.329 --> 00:11:54.889 little bit easier to add things into. 00:11:55.109 --> 00:11:59.559 But then maybe you are trying to expand somebody else's published design first. 00:11:59.559 --> 00:12:01.159 Stuff that you don't have any control over. 00:12:01.259 --> 00:12:06.779 So there's this kind of constant idea of taking . Some open API and 00:12:06.779 --> 00:12:10.179 then like programmatically adding stuff to it for whatever reason. 00:12:10.249 --> 00:12:13.379 Tech writing, you could do translation, you could add a bunch of stuff in. 00:12:13.749 --> 00:12:15.839 I've, I've seen your name come up for quite a few blog 00:12:15.839 --> 00:12:17.199 posts and overlay overlays. 00:12:17.499 --> 00:12:19.839 And while I was doing some research, I noticed the folks over at Redoc 00:12:19.839 --> 00:12:21.079 aren't quite so impressed with it. 00:12:21.099 --> 00:12:23.719 So I was wondering if you could explain where it all sits. 00:12:24.149 --> 00:12:24.999 Lorna Mitchell: controversial? 00:12:25.469 --> 00:12:25.759 Yeah. 00:12:25.879 --> 00:12:26.359 I think so. 00:12:26.359 --> 00:12:32.549 There's an older blog post on the Redoc website where Adam, you know, 00:12:33.914 --> 00:12:35.694 It didn't love it when he saw it. 00:12:35.694 --> 00:12:38.854 It's, it's not a brand fresh site, but it is top hit for overlay. 00:12:38.874 --> 00:12:39.374 So it's 00:12:39.374 --> 00:12:44.744 not a secret And I think the point there is a good one, that it's 00:12:44.764 --> 00:12:46.944 yet another complex structure. 00:12:47.034 --> 00:12:50.264 We've got a specification with no tooling at all, and it 00:12:50.344 --> 00:12:51.984 requires use of JSO path. 00:12:52.684 --> 00:12:58.984 Now, if you've been using Spectral, that is very JSON path heavy. 00:12:59.749 --> 00:13:00.039 Track 1: Yeah. 00:13:00.534 --> 00:13:06.794 Lorna Mitchell: But v Doley, CLI isn't, we do that by considering 00:13:06.794 --> 00:13:10.514 the open API as like a tree structure and the different data 00:13:10.514 --> 00:13:10.794 types. 00:13:10.794 --> 00:13:12.554 So everything's very type aware. 00:13:13.134 --> 00:13:17.324 And you can filter on fields and it knows, oh, this is a parameter you know, 00:13:17.324 --> 00:13:23.354 it'll have an in then We don't, with JSON path, it's just blind data structure. 00:13:24.014 --> 00:13:27.674 So we don't have JSON path currently in the Marine Dole CLI 00:13:27.674 --> 00:13:29.354 ecosystem I don't think at all. 00:13:30.174 --> 00:13:33.314 And so that extra learning curve, you sort of think, well, 00:13:33.434 --> 00:13:34.714 I don't know who could use this. 00:13:34.714 --> 00:13:40.114 Like it's a good solution, but there's no validator for the format. 00:13:40.534 --> 00:13:44.314 And the learning curve puts it beyond a lot of the users that 00:13:44.614 --> 00:13:46.234 at re doley we're trying to enable. 00:13:47.139 --> 00:13:47.489 Track 1: Right. 00:13:47.669 --> 00:13:50.529 And so I think when we are talking about the use cases for decorators, like 00:13:50.609 --> 00:13:55.519 overlays kind of intended to be a more generic implementation of that, that 00:13:55.539 --> 00:13:57.919 anyone could theoretically use, whereas decorators is 00:13:57.919 --> 00:13:59.279 more of a like redly feature. 00:13:59.729 --> 00:14:02.429 And could theoretically be, you know, ji to, to be the same thing. 00:14:02.489 --> 00:14:04.269 But what, what you're saying is yeah, the. 00:14:05.564 --> 00:14:06.074 right now. 00:14:06.354 --> 00:14:10.234 Overlays are designed to use JSON Path and for anyone who's not 00:14:10.514 --> 00:14:14.074 familiar if you've been using Spectral yet, then yeah, that's, that's the 00:14:14.434 --> 00:14:16.394 absolutely batshit bunch of things. 00:14:16.394 --> 00:14:18.954 You're typing into those rules to try and make anything, make work. 00:14:19.374 --> 00:14:19.864 Make work. 00:14:19.924 --> 00:14:20.944 You can tell I'm tired. 00:14:22.279 --> 00:14:25.199 I was saying before the show, I've just come back from API days and, and normally 00:14:25.439 --> 00:14:28.719 API days gets you a hangover, but this time I've just been like sick for a couple 00:14:28.719 --> 00:14:30.799 of days and now I've just not slept. 00:14:31.319 --> 00:14:34.979 I decided to go and get myself into a hotel room early and then like got into 00:14:34.979 --> 00:14:37.499 bed at eight o'clock and there were just like homeless people screaming 00:14:37.499 --> 00:14:41.099 outside my window until 4:00 AM and then the fire alarm went off at 8:00 00:14:41.099 --> 00:14:43.539 AM and I'm like, what is reality? 00:14:44.189 --> 00:14:47.959 So, yeah, JSON Path is even more confusing than my brain right now. 00:14:48.019 --> 00:14:50.959 And basically you have to kind of tell it what you would like to filter through, 00:14:50.959 --> 00:14:54.279 kind like a CSS selector but for js. 00:14:54.739 --> 00:14:56.919 And then, so JSON Path doesn't have a standard. 00:14:57.399 --> 00:14:59.764 They're kind of working on one, but it's not there yet. 00:15:00.304 --> 00:15:03.964 And so much like what markdown had, like markdown plus and a 00:15:03.964 --> 00:15:05.244 million different markdown flavors. 00:15:05.624 --> 00:15:10.074 JSON path now has JSON path plus, which stoplight was using for quite a long 00:15:10.074 --> 00:15:11.794 time and, and used as the main thing for. 00:15:12.714 --> 00:15:15.334 For spectral rules, but then that wasn't quite enough. 00:15:15.634 --> 00:15:20.664 And then like Nier appeared, which is Jason Path Plus with some other bits 00:15:21.004 --> 00:15:26.544 and like it's even Jason Path Plus not Nier is a mixture of a lot of just 00:15:26.624 --> 00:15:27.944 squiggles and doodles and back ticks. 00:15:28.619 --> 00:15:30.759 But also you can bung red rejects in there. 00:15:31.099 --> 00:15:34.559 So you can kind of say, I would like to dot gets you into the property of 00:15:34.559 --> 00:15:38.839 something and then square brackets can get you into like the array of, of something. 00:15:39.059 --> 00:15:40.919 And then there's like stars to say all of them. 00:15:40.979 --> 00:15:45.639 And there's tilts to say, parent, and you can kind of fetch all of the 00:15:45.819 --> 00:15:52.029 headers where the parent is on this and the, the name begins with an X 00:15:52.409 --> 00:15:53.829 and you can do really powerful things. 00:15:54.089 --> 00:16:01.359 But it's . Almost impossible in any sort of . I've made lots of spectral 00:16:01.359 --> 00:16:06.399 rule sets because I built an NPM rule set and, and I built an NPM test suite 00:16:06.729 --> 00:16:10.719 using jest so that I can type in these crazy things and go, am I close yet? 00:16:10.719 --> 00:16:11.319 Am I close yet? 00:16:11.319 --> 00:16:11.959 Am I close yet? 00:16:12.099 --> 00:16:13.759 And the result will show me what I've done. 00:16:14.339 --> 00:16:16.319 But most people are just writing a YAML file. 00:16:17.174 --> 00:16:18.674 And, and that doesn't work. 00:16:18.734 --> 00:16:21.554 So you have to like, literally switch over to your terminal, run it, and it 00:16:21.554 --> 00:16:22.874 just blows up at you in strange ways. 00:16:22.934 --> 00:16:24.594 And it, it can be very hard to do. 00:16:24.694 --> 00:16:28.494 So it's one of the things I really appreciated about about Redoc Redoc 00:16:28.594 --> 00:16:34.214 Lee, CLI was the, was the switch to using just named selectors. 00:16:34.354 --> 00:16:35.214 So you can say like. 00:16:35.874 --> 00:16:40.294 Header instead of going like, you know and I think spectral iss 00:16:40.514 --> 00:16:44.774 trying it, it made early attempt to fix that by adding aliases. 00:16:45.074 --> 00:16:47.014 But I'm just going, going off on a bit of a tirade now. 00:16:47.014 --> 00:16:47.374 I've gotta get 00:16:47.374 --> 00:16:48.094 some stuff off my chest. 00:16:49.004 --> 00:16:49.594 Lorna Mitchell: Carry on. 00:16:50.424 --> 00:16:53.044 Track 1: The roadmap for aliases was powerful and impressive, and 00:16:53.044 --> 00:16:56.084 I've smoked to Smart Bear who have now inherited this to try and get 00:16:56.084 --> 00:16:56.884 this back on the table. 00:16:57.464 --> 00:17:01.094 But basically aliases, the idea was let's come up with a list of aliases. 00:17:01.094 --> 00:17:05.334 Common open, API, aliases common A and KPI, aliases, and then 00:17:05.434 --> 00:17:06.854 people can use those instead. 00:17:07.074 --> 00:17:08.014 So instead of dollar. 00:17:09.064 --> 00:17:14.414 Which is root and then paths, brackets slash post you, you would just say 00:17:14.414 --> 00:17:18.294 like you would just say paths and then you could do the square bracket slash 00:17:18.294 --> 00:17:19.734 post to pull out all the post ones. 00:17:19.734 --> 00:17:23.574 So you would cut out a lot of the nonsense by using aliases to replace 00:17:23.694 --> 00:17:28.354 a certain chunk of of the path. 00:17:28.354 --> 00:17:32.314 Whereas I think Redley would've said, you know, paths and then you would have. 00:17:33.425 --> 00:17:35.314 Header if, if, if post or something, right? 00:17:35.314 --> 00:17:38.794 So yours is more of a DSL approach, but aliases seemed like a good step forward. 00:17:39.415 --> 00:17:47.095 And yeah, the, the aliases ended up getting defined in the core rule set 00:17:47.514 --> 00:17:49.534 of like the core open API rule set. 00:17:49.995 --> 00:17:52.324 And they weren't extendable. 00:17:52.384 --> 00:17:55.004 So if you extend the core open API rule set, you can. 00:17:55.554 --> 00:17:58.495 Get those aliases out, they're only available inside the place. 00:17:58.495 --> 00:17:59.054 They're defined. 00:17:59.314 --> 00:18:01.654 So that meant you couldn't actually use those to help. 00:18:01.654 --> 00:18:04.975 Like no one using, no one writing their own rule sets, extending open 00:18:05.054 --> 00:18:08.094 API or no one writing their own rule sets at all, could use those aliases 00:18:08.094 --> 00:18:09.215 unless you copy and paste them. 00:18:09.394 --> 00:18:13.414 And so I'm, I've been suggesting that like not only should they be extendable, but 00:18:13.564 --> 00:18:15.374 they should be defined in like the format. 00:18:15.914 --> 00:18:18.254 So you should, you know, anyone using the open API format. 00:18:19.100 --> 00:18:22.269 Whether it's the core rule set or not, you should get those aliases. 00:18:22.370 --> 00:18:25.310 But that's, that's a bit more of a, like a, a roadmap. 00:18:25.370 --> 00:18:27.790 One day they might fix it kind of thing, but 00:18:28.149 --> 00:18:28.270 I, 00:18:28.270 --> 00:18:31.110 completely recognize the need to not Jason path and therefore 00:18:31.280 --> 00:18:32.909 avoid that entirely in overlays. 00:18:32.909 --> 00:18:32.990 Right. 00:18:33.575 --> 00:18:33.865 Lorna Mitchell: yeah. 00:18:34.095 --> 00:18:36.544 Well, I don't think it is avoidable in overlays. 00:18:36.965 --> 00:18:40.225 And I feel a little bit like I've done it, an injustice there. 00:18:40.225 --> 00:18:44.185 Like leading with that, leading with that negative, the controversial blog post. 00:18:44.245 --> 00:18:45.424 You know, not everyone's a fan. 00:18:46.344 --> 00:18:48.344 I really think that overlays. 00:18:49.199 --> 00:18:50.779 Solves a real problem. 00:18:51.169 --> 00:18:54.979 Like I see this all the time we just talked about when it's generated from 00:18:55.009 --> 00:18:57.659 code and it hasn't got enough anything. 00:18:58.199 --> 00:19:00.860 I'm also seeing much more complex pipelines. 00:19:00.959 --> 00:19:01.179 Now. 00:19:01.799 --> 00:19:05.779 People have multiple teams maintaining different bits of the a p surface. 00:19:06.409 --> 00:19:08.340 They all ship an open API. 00:19:08.879 --> 00:19:14.100 We join it into one, and then the, the downstream outputs, whether 00:19:14.100 --> 00:19:17.819 that's documentation, SDK, generation API, gateways, like whatever. 00:19:19.119 --> 00:19:24.699 All of them need their own enrichment and overlays solves that problem where you 00:19:24.699 --> 00:19:30.259 don't have to, you can just, yeah, add a translation, add an example, add the 00:19:30.259 --> 00:19:35.379 hint for the correct data type for the target language in the SDK and the team 00:19:35.379 --> 00:19:37.259 that deals with the SDKs can do that. 00:19:37.259 --> 00:19:42.139 We don't have to go all the way back in a, in a, in a big organization that's hard 00:19:42.679 --> 00:19:43.259 to find. 00:19:43.279 --> 00:19:44.939 The team that maintains the thing. 00:19:45.644 --> 00:19:46.374 That lives in the house. 00:19:46.429 --> 00:19:47.329 The Jack bill, right? 00:19:47.869 --> 00:19:51.369 So ,we, we can do that here. 00:19:51.569 --> 00:19:53.289 Overlays solves that problem. 00:19:53.969 --> 00:19:59.329 I don't think it's a coincidence that one of the implementations, there are 00:19:59.329 --> 00:20:00.969 not many overlays, implementations. 00:20:01.269 --> 00:20:02.289 One of them is mine. 00:20:02.429 --> 00:20:03.409 We can link to it. 00:20:03.519 --> 00:20:04.689 It's very hobby. 00:20:05.449 --> 00:20:07.689 I just, I, this is how I get things out of my brain. 00:20:07.889 --> 00:20:10.729 I had to write code that did it so that I could look at 00:20:10.729 --> 00:20:10.889 it. 00:20:11.629 --> 00:20:12.070 I need to 00:20:12.070 --> 00:20:12.230 do 00:20:12.370 --> 00:20:12.850 Track 1: that's really helpful. 00:20:13.170 --> 00:20:16.480 I, I remember, I remember reading the, the official page 00:20:16.480 --> 00:20:17.800 of the overlays working group. 00:20:18.300 --> 00:20:19.950 And oh, just a quick thing on working groups. 00:20:20.000 --> 00:20:24.520 So open API used to do everything by everyone getting on a weekly call 00:20:24.540 --> 00:20:27.600 and then getting homework assigned, and then like, maybe it was done 00:20:27.600 --> 00:20:28.880 by the next weekly call, but that 00:20:28.880 --> 00:20:30.000 was how everything ever happened. 00:20:30.260 --> 00:20:32.680 And if there wasn't enough interest in certain things, 00:20:32.680 --> 00:20:33.680 then they just wouldn't happen. 00:20:33.820 --> 00:20:37.680 But now things have been split up into working groups and I think one of them 00:20:37.700 --> 00:20:41.140 was like workflows that I was . Briefly involved with, it's quite exciting. 00:20:41.680 --> 00:20:45.670 And there's a bunch of other ones and SLAs and, and yeah, so 00:20:45.670 --> 00:20:46.950 overlays was, was one of them. 00:20:46.970 --> 00:20:51.400 And it's great to see an early kind of finished I don't know what phases and 00:20:51.400 --> 00:20:55.360 stages there are, but I read the repo and it said like, we're basically stable. 00:20:56.230 --> 00:20:59.450 Now but no one's really implemented this, so can, you know, we need some 00:20:59.450 --> 00:21:00.930 big implementations to go and do it. 00:21:00.930 --> 00:21:03.570 And, and it's cool to hear there's some hobbyist ones or 00:21:03.570 --> 00:21:04.650 some implementations out there, 00:21:04.670 --> 00:21:05.930 but that, that's the hard part. 00:21:05.930 --> 00:21:07.010 We had this problem at stoplight. 00:21:07.290 --> 00:21:11.730 Everyone has this problem where you want to support a feature, but you 00:21:11.730 --> 00:21:14.650 get a few user requests come in saying like, could you use this feature? 00:21:14.700 --> 00:21:15.170 Could we. 00:21:15.530 --> 00:21:16.400 Could we have this feature? 00:21:16.515 --> 00:21:18.775 And then you're like, oh, I'm not sure we should implement 00:21:18.775 --> 00:21:20.295 it because it's not stable yet. 00:21:20.315 --> 00:21:23.175 And the people who have like come up with it have said, well, we'll only 00:21:23.175 --> 00:21:26.055 call it version one when it's had some implementation and some feedback. 00:21:26.055 --> 00:21:29.255 So you get in this chicken and egg problem of everyone hoping somebody else 00:21:29.645 --> 00:21:32.215 will implement it, so that then it stabilizes. 00:21:32.475 --> 00:21:35.935 But then if they're the first to market, then you lose out, but you're not 00:21:36.375 --> 00:21:37.775 prepared to commit the resources to it. 00:21:37.835 --> 00:21:38.055 So 00:21:38.395 --> 00:21:39.215 So where are we at 00:21:39.570 --> 00:21:40.100 Lorna Mitchell: Good news. 00:21:40.460 --> 00:21:43.900 I have a hobbyist thing and I think Mike Sson has something as well. 00:21:43.960 --> 00:21:45.100 He donated me a 00:21:45.140 --> 00:21:45.380 Tests. 00:21:45.600 --> 00:21:46.020 Mine's 00:21:46.020 --> 00:21:46.140 open 00:21:46.320 --> 00:21:48.080 Track 1: course Mike does brilliant 00:21:48.100 --> 00:21:48.260 Lorna Mitchell: does. 00:21:48.490 --> 00:21:50.460 It's like he's got his finger on the pulse. 00:21:50.920 --> 00:21:54.380 But there's another implementation, which is from speakeasy. 00:21:54.680 --> 00:21:58.260 Now they do call modern SDKs. 00:21:58.480 --> 00:22:02.380 It is not a coincidence that it is them that have gone to 00:22:02.380 --> 00:22:03.740 implement with overlays because. 00:22:04.370 --> 00:22:09.460 Open API does not have all of the metadata that you need to generate 00:22:09.620 --> 00:22:11.220 a great SDK in every tech stack. 00:22:11.490 --> 00:22:15.660 Some of them need more data type instructions or serialization 00:22:15.660 --> 00:22:19.780 information, or, you know, there's a lot that you need to add that wouldn't 00:22:19.780 --> 00:22:21.980 be in a standard open API description. 00:22:23.285 --> 00:22:27.675 So they have these extra extensions, the x dash fields that they add. 00:22:28.575 --> 00:22:32.275 And rather than everyone, like you say, you can't add them because then when 00:22:32.275 --> 00:22:34.635 you update your open API, they are lost. 00:22:34.975 --> 00:22:37.555 So not a coincidence that they implemented overlays. 00:22:37.725 --> 00:22:43.315 Their tool is available, it's open source, it includes a validator. 00:22:44.430 --> 00:22:48.080 Kind of embrace and extended the, the implementation a little bit, but I think 00:22:48.080 --> 00:22:49.680 that's an interesting use case too. 00:22:50.580 --> 00:22:51.280 So yeah. 00:22:51.710 --> 00:22:53.000 More than one tool to look at. 00:22:53.100 --> 00:22:53.920 And I'm 00:22:53.920 --> 00:22:56.360 excited because I think it, it really solves a real problem. 00:22:57.130 --> 00:22:57.900 Track 1: That is really cool. 00:22:58.000 --> 00:23:01.460 And once again, I'm googling mid podcast 'cause I am extremely professional. 00:23:01.480 --> 00:23:03.640 But I'm looking at the speakeasy dev. 00:23:03.720 --> 00:23:06.500 I'm just seeing AI powered end-to-end API maintenance. 00:23:06.680 --> 00:23:08.580 And again, there's like, I've just come from API days. 00:23:08.580 --> 00:23:12.420 Every other talk was about AI and I just, you know, my, my eyes are in 00:23:12.420 --> 00:23:16.000 pain from rolling, but this looks like a, a really useful implementation. 00:23:16.000 --> 00:23:19.010 They're like . They've noticed that you've type that they've noticed that 00:23:19.010 --> 00:23:20.830 you've typed in a duplicate schema and 00:23:20.830 --> 00:23:23.870 then just like changed it to a ref , which is amazing. 00:23:24.330 --> 00:23:28.230 And there's a lot of other pretty cool looking stuff on this marketing page. 00:23:28.320 --> 00:23:31.280 Let's, let's be clear, but I like their intentions. 00:23:32.010 --> 00:23:32.300 Lorna Mitchell: Yeah. 00:23:32.600 --> 00:23:33.140 And I think 00:23:33.160 --> 00:23:33.680 Track 1: dig in more. 00:23:34.040 --> 00:23:36.580 Lorna Mitchell: I'm with you in in team Rolling your eyes. 00:23:36.990 --> 00:23:38.080 When it comes to Yeah. 00:23:38.373 --> 00:23:40.033 AI is gonna do everything but 00:23:40.353 --> 00:23:46.233 actually . The open API, that standard machine readable specification is gonna 00:23:46.233 --> 00:23:48.313 enable a bunch of AI applications. 00:23:49.133 --> 00:23:54.513 So, and also chat GT four knows quite a lot of open API and can 00:23:54.513 --> 00:23:57.393 like quickly give you, quickly give you back some hilarious 00:23:57.573 --> 00:23:59.353 and sensible example fields. 00:23:59.493 --> 00:24:01.233 For example, I wouldn't let it write 00:24:01.233 --> 00:24:02.473 my API, but when I'm 00:24:02.473 --> 00:24:05.033 working on examples, it's like, ah, this could be better. 00:24:05.343 --> 00:24:05.633 Like, 00:24:05.633 --> 00:24:07.993 make me, make me something fun with circus animals. 00:24:08.133 --> 00:24:08.713 It can do that. 00:24:09.918 --> 00:24:10.198 Track 1: Brilliant. 00:24:10.228 --> 00:24:10.518 Okay. 00:24:10.518 --> 00:24:10.718 Yeah. 00:24:10.838 --> 00:24:13.918 'cause writing, you know, arbitrary examples was always the hardest part. 00:24:14.038 --> 00:24:17.478 I would always really struggle to do tutorials or demos 'cause 00:24:17.478 --> 00:24:21.638 I was like, just coming up with contrived examples is so frustrating. 00:24:21.658 --> 00:24:24.078 And like, everyone's got a bloody to-do app and no one cares. 00:24:24.218 --> 00:24:28.348 And, and just, I, I fundamentally refuse to use the pet store for anything ever 00:24:28.798 --> 00:24:29.018 So, 00:24:29.018 --> 00:24:29.898 yeah, I, I have 00:24:29.963 --> 00:24:32.663 Lorna Mitchell: we can share a link to Redley just published a museum, 00:24:32.783 --> 00:24:34.103 API for more or less these reasons. 00:24:34.453 --> 00:24:34.983 Also, the 00:24:34.983 --> 00:24:36.263 pet store's really outdated. 00:24:36.813 --> 00:24:40.943 SmartBear only just gone to 3.1, so we've been using a variation of 00:24:40.943 --> 00:24:43.783 this museum thing internally and we were like, we should ship this. 00:24:44.538 --> 00:24:44.888 Track 1: right? 00:24:45.078 --> 00:24:51.008 Well, So I mean, , the, what really annoys me about the pet store is that it's a. 00:24:52.193 --> 00:24:52.413 Bad. 00:24:52.453 --> 00:24:54.213 API like it's describing a bad API. 00:24:54.213 --> 00:24:56.573 It's got like a bunch of really bad conventions in the actual 00:24:56.613 --> 00:24:57.693 API that it's describing. 00:24:58.113 --> 00:25:02.813 it it uses a weird amount of, it uses like a, an odd amount of open API to do it. 00:25:02.893 --> 00:25:05.253 I think it's being copied from like swagger two and, and 00:25:05.353 --> 00:25:06.653 not being fleshed out. 00:25:07.073 --> 00:25:10.293 And then, so now the pet store lives in open API three when 00:25:10.363 --> 00:25:14.933 like 3.1 came out, 20 20, 20 21. 00:25:15.318 --> 00:25:15.818 Lorna Mitchell: 2021. 00:25:15.838 --> 00:25:17.538 By the time we press the button, yeah. 00:25:17.668 --> 00:25:21.278 Track 1: Yeah, so that's, you know, two, it's nearly, yeah, it's a while. 00:25:22.018 --> 00:25:23.558 So don't use the pet store. 00:25:23.888 --> 00:25:26.848 There's other better examples and I like that AI can help with that for sure. 00:25:27.088 --> 00:25:32.528 I mean, I have been using copilot built into VS code and it's been 00:25:32.528 --> 00:25:33.888 helping me out for the protector. 00:25:34.008 --> 00:25:34.728 API quite a lot. 00:25:34.728 --> 00:25:39.948 Like I was just adding an array of species to sites so that we can . Ahead of time, 00:25:40.008 --> 00:25:43.108 we can say, look, these are the only species we're gonna plant at this site. 00:25:43.448 --> 00:25:45.868 So when I'm going through the field and it's raining and I've 00:25:45.868 --> 00:25:48.628 gotta take a photograph of 4,000 bloody trees, we just planted for 00:25:48.628 --> 00:25:49.788 proof for our funding partners. 00:25:50.108 --> 00:25:53.588 I don't wanna be scrolling past like sea buckthorn, which we've never planted, 00:25:53.958 --> 00:25:56.638 Just to get to like the, the, the subset of the eight species that 00:25:56.638 --> 00:25:57.998 are actually in that bloody field. 00:25:57.998 --> 00:25:58.238 Right. 00:25:58.578 --> 00:26:00.658 'cause we've got like 50 different species we may plant there. 00:26:00.998 --> 00:26:04.538 So yeah, like I was just putting that in and I just typed on the, on the sites. 00:26:04.988 --> 00:26:09.038 List I typed in species and it was like, oh, would you like to add 00:26:09.098 --> 00:26:12.438 all of these obvious properties and this whole schema and then reference 00:26:12.498 --> 00:26:14.738 the docs slash schema no sorry. 00:26:14.938 --> 00:26:17.068 Schema slash species that I see over there. 00:26:17.108 --> 00:26:18.508 I was like, yes, yes, I bloody wood. 00:26:18.508 --> 00:26:20.388 And it just built like loads of stuff for me. 00:26:20.638 --> 00:26:24.418 And even building, building that that one schema was great, but there have been 00:26:24.418 --> 00:26:27.338 lots of times where it's just completely hallucinated and done something absolutely 00:26:27.338 --> 00:26:29.418 bonkers, and I've just like pushed it. 00:26:29.558 --> 00:26:29.778 So 00:26:30.218 --> 00:26:33.578 I, I still don't trust it to do anything much apart from like, kind 00:26:33.578 --> 00:26:37.578 of assist me, but I, I, I'll never do dictated but not read with AI 00:26:38.113 --> 00:26:38.333 Lorna Mitchell: No, 00:26:38.568 --> 00:26:39.338 Track 1: it's not there yet. 00:26:39.373 --> 00:26:42.453 Lorna Mitchell: think as well, like it's a, it's, it's a time enhancer for me. 00:26:42.873 --> 00:26:47.613 One of the things that I find time consuming is I'm no longer particularly 00:26:47.643 --> 00:26:50.013 technically specialist in any given. 00:26:50.743 --> 00:26:51.548 Programming language. 00:26:51.828 --> 00:26:52.308 I write about 00:26:52.308 --> 00:26:53.868 four with equal levels of danger. 00:26:54.228 --> 00:26:56.828 PP is my original community, but it's moved on. 00:26:57.188 --> 00:26:59.348 I wouldn't recognize it if I walked past it in the street. 00:26:59.848 --> 00:27:04.548 And so and so actually I can write all of those languages and I can 00:27:04.558 --> 00:27:07.148 debug all of those languages, but I just like throw the thing I want. 00:27:08.213 --> 00:27:10.873 In and I get something back and then that's enough. 00:27:11.213 --> 00:27:13.633 So otherwise you're like, how do I iterate? 00:27:13.773 --> 00:27:14.993 How do I concatenate? 00:27:15.383 --> 00:27:16.433 What have I done wrong with 00:27:16.458 --> 00:27:16.698 Track 1: Yeah. 00:27:16.848 --> 00:27:17.138 Yeah. 00:27:17.253 --> 00:27:17.673 Lorna Mitchell: on this? 00:27:18.253 --> 00:27:21.153 You know, like I know that it JavaScript array is a reference, 00:27:21.153 --> 00:27:22.433 but I need it to not be. 00:27:22.853 --> 00:27:27.473 And so that kind of speed up is quicker for the things that I don't know. 00:27:28.053 --> 00:27:30.033 It just lies and I can't catch it. 00:27:30.093 --> 00:27:31.233 So yeah, I'm with you. 00:27:31.233 --> 00:27:31.433 That 00:27:31.608 --> 00:27:33.058 Track 1: That that is absolutely the problem. 00:27:33.118 --> 00:27:36.258 And it's totally fine if this decay like this podcast decay 00:27:36.288 --> 00:27:37.378 into complaining about ai. 00:27:37.538 --> 00:27:39.538 'cause we've been meaning to talk more about AI on it for a while. 00:27:39.548 --> 00:27:41.978 We've got someone coming on soon to talk about things that it is 00:27:41.978 --> 00:27:43.978 useful for and so we can get some of the bullshit outta the way. 00:27:44.788 --> 00:27:47.248 But yeah, even with PHP, like people have said, oh, it can help 00:27:47.248 --> 00:27:48.688 you write your tests or whatever. 00:27:48.908 --> 00:27:52.168 And I was writing a test for a model and it just kept recommending 00:27:52.168 --> 00:27:56.208 that I like write these really complicated tests for relationships, 00:27:56.338 --> 00:27:57.568 which you don't really need to do. 00:27:57.808 --> 00:28:00.848 'cause if you have an OMM that has relationship logic and you just say, I. 00:28:01.068 --> 00:28:02.968 You know orders should have organization. 00:28:02.968 --> 00:28:04.128 You don't really need to write a test. 00:28:04.128 --> 00:28:06.168 That's, that says orders do have organization, 00:28:06.508 --> 00:28:09.608 but the test it wrote was invalid and didn't fundamentally work at all. 00:28:09.988 --> 00:28:11.248 And kept doing loads of other stuff. 00:28:11.248 --> 00:28:13.608 It was trying to test methods that didn't exist 00:28:13.918 --> 00:28:16.128 like it was trying to find out if that I. 00:28:16.128 --> 00:28:19.838 Order was allocated to something, but orders aren't allocated, like 00:28:19.848 --> 00:28:21.318 units are allocated to orders. 00:28:21.418 --> 00:28:26.158 So it, it sore enough to have a real college try, but it didn't 00:28:26.398 --> 00:28:27.358 actually do the right thing. 00:28:27.458 --> 00:28:29.278 And it just confused me more than if it didn't. 00:28:29.438 --> 00:28:32.998 'cause if, if I hadn't spent like 10 minutes trying to faff around with the, 00:28:33.458 --> 00:28:37.038 the bad test that it made, I would've just Googled how do I test this thing 00:28:37.458 --> 00:28:38.358 and then done it 00:28:38.658 --> 00:28:38.878 So. 00:28:39.268 --> 00:28:44.328 Lorna Mitchell: And I do worry about how much it helps us to learn new things, 00:28:44.348 --> 00:28:48.448 and especially if you are earlier in your career, like actually, is this helping 00:28:48.588 --> 00:28:51.058 you to be the best that you can be? 00:28:51.118 --> 00:28:54.098 It, I'm, I'm just not sure, and time will tell, I think. 00:28:54.908 --> 00:28:58.868 Track 1: Yeah, one, one devil's advocate there is that it's, it's 00:28:58.868 --> 00:29:02.148 just sped up stack overflow, copy pasting, which is, you know, and it 00:29:02.148 --> 00:29:03.388 has a bit of knowledge of your system 00:29:03.388 --> 00:29:04.708 where a stack overflow has none. 00:29:05.048 --> 00:29:08.588 So there's plenty of times that I've like Googled for a problem and like copied 00:29:08.648 --> 00:29:12.788 and pasted the close enough thing into my code base and, and hope for the best. 00:29:12.808 --> 00:29:16.828 And so if it's, if it's doing that, but slightly better than, okay. 00:29:17.968 --> 00:29:18.188 Lorna Mitchell: It 00:29:18.188 --> 00:29:18.508 isn't 00:29:18.783 --> 00:29:21.423 Track 1: still not really adding that knowledge and there's, you're missing 00:29:21.483 --> 00:29:26.993 the, like, the ability to see down votes and, you know I mean, I think you were 00:29:26.993 --> 00:29:28.153 around in the code United Days, right? 00:29:28.173 --> 00:29:28.393 And, 00:29:28.393 --> 00:29:32.393 and if not early PHP days where a lot of people were doing really dumb stuff 00:29:32.613 --> 00:29:33.793 and really dumb stuff was popular 00:29:34.333 --> 00:29:37.753 and, and the popularity of something, the frequency of which you see 00:29:37.753 --> 00:29:38.873 something doesn't make it better. 00:29:39.433 --> 00:29:43.413 And so you kind of have to hope that at least with Stack Overflow and 00:29:43.413 --> 00:29:46.333 some of those forums, you'd have someone posting a really terrible 00:29:46.483 --> 00:29:50.013 idea that may even be accepted and then a bunch of people down voting and 00:29:50.013 --> 00:29:52.133 going, whatcha doing underneath it? 00:29:52.423 --> 00:29:53.213 Which is good. 00:29:53.513 --> 00:29:54.893 And with this it just goes, now take that. 00:29:54.893 --> 00:29:55.413 May it'd be fine. 00:29:55.913 --> 00:29:56.353 Lorna Mitchell: Accepted answer. 00:29:56.453 --> 00:29:56.913 Go for it. 00:29:57.103 --> 00:29:57.393 Yeah. 00:29:57.393 --> 00:29:59.713 And what you want is you want the most voted answer. 00:29:59.813 --> 00:30:02.313 And you also want, did this change recently? 00:30:02.413 --> 00:30:03.713 You know, stack Overflow is not a 00:30:03.713 --> 00:30:04.003 Track 1: yeah. 00:30:04.003 --> 00:30:04.283 Yeah. 00:30:04.623 --> 00:30:06.683 It was the correct answer for years Yeah. 00:30:06.833 --> 00:30:09.633 Lorna Mitchell: was correct but after 2015, you should do it this way. 00:30:10.183 --> 00:30:13.393 Like that's a, that's a real answer, and I feel that. 00:30:14.108 --> 00:30:17.598 Machine necessarily on our side with that 00:30:17.598 --> 00:30:20.238 because they can't take in that context. 00:30:21.448 --> 00:30:24.228 Track 1: can I, can I do a, a complaint about oh, it's 00:30:24.228 --> 00:30:25.228 my podcast to do what I want? 00:30:25.518 --> 00:30:27.718 Lorna Mitchell: I love, please, please continue. 00:30:28.948 --> 00:30:32.468 Track 1: one of the talks I saw by a lovely bloke, well-intentioned 00:30:32.688 --> 00:30:35.178 talk, but the whole thing fundamentally seemed to be. 00:30:36.323 --> 00:30:40.493 Integrating with other people's APIs is really hard, so here 00:30:40.713 --> 00:30:42.773 we can get AI to do it for us. 00:30:43.353 --> 00:30:48.133 So here is an, an AI doing all of this really complicated, amazing 00:30:48.463 --> 00:30:52.853 stuff and it was truly impressive what the toy demos were doing. 00:30:53.113 --> 00:30:57.653 But towards the end of the question and answer, it was very much, 00:30:57.983 --> 00:31:02.843 how much does the AI need the API to be built in a specific way? 00:31:03.543 --> 00:31:08.283 And and he was like, oh yeah, basically there are certain ways that 00:31:08.383 --> 00:31:10.323 AI will expect the API to be built. 00:31:10.783 --> 00:31:13.843 And if it's not built following those very specific assumptions, then 00:31:13.843 --> 00:31:15.363 this basically won't work at all. 00:31:15.903 --> 00:31:20.283 So what some people do is build a new API specifically for the ai. 00:31:20.293 --> 00:31:20.713 Lorna Mitchell: Mm-Hmm. 00:31:21.228 --> 00:31:24.018 Track 1: Where it does follow all of those conventions and then 00:31:24.018 --> 00:31:25.258 it can save you loads of time. 00:31:25.258 --> 00:31:25.978 It's like, oh, okay. 00:31:25.978 --> 00:31:29.898 We're all just completely rebuilding all of our APIs and, and doing BFFs 00:31:29.998 --> 00:31:34.658 for every client just so that AI can then save some time instead of looking 00:31:34.678 --> 00:31:36.578 at the docs we spent ages making. 00:31:36.998 --> 00:31:40.918 I, that's, that's sort of thing is where my, my, my bullshit flag goes up. 00:31:41.158 --> 00:31:45.118 'cause it's like, is this just early days and the AI will eventually be able 00:31:45.118 --> 00:31:49.728 to talk to any API or are we just kind of saying . Like the wizard will fix 00:31:49.728 --> 00:31:54.408 it and pass in more work onto API teams who probably would've had a better time 00:31:54.408 --> 00:31:57.488 sitting down with their stakeholders and asking what they really need instead of 00:31:57.648 --> 00:32:00.288 building some shit that was so confusing in the first place, that everyone's 00:32:00.288 --> 00:32:03.528 scared of looking at the documentation and then ask for an AI to be built so 00:32:03.528 --> 00:32:04.808 they can figure out what the hell you did. 00:32:05.038 --> 00:32:05.328 Like 00:32:05.818 --> 00:32:07.568 maybe we could just make better APIs. 00:32:07.918 --> 00:32:08.208 Yeah. 00:32:08.703 --> 00:32:12.353 Lorna Mitchell: Well, and I, I'm an optimist, so I, I wanna flip that argument 00:32:12.373 --> 00:32:15.873 and say you're right, but it, what you're actually saying is that the people who 00:32:15.983 --> 00:32:21.193 have, well-designed, well-documented APIs described with open API. 00:32:22.198 --> 00:32:26.018 Can integrate easily and well both with other platforms and with ai if 00:32:26.018 --> 00:32:27.538 that's what people want to use it with. 00:32:27.968 --> 00:32:29.458 It's good for all integrations. 00:32:29.718 --> 00:32:34.298 People who have badly designed APIs incomplete API documentation, 00:32:34.968 --> 00:32:36.338 they are gonna have a hard time. 00:32:36.618 --> 00:32:38.418 I don't think the wrapper APIs add anything. 00:32:39.318 --> 00:32:39.538 You, 00:32:39.748 --> 00:32:43.618 you'd be better just to start again, talk to the actual users 00:32:44.328 --> 00:32:45.618 talk to the AI if you like. 00:32:45.618 --> 00:32:46.618 They're very chatty. 00:32:47.358 --> 00:32:48.458 And make, and make things better. 00:32:49.578 --> 00:32:50.198 Track 1: That's cool. 00:32:50.968 --> 00:32:53.818 Yeah, I think it's a good point is that some of those, some of those people 00:32:53.818 --> 00:32:57.178 were coming from a place of, you don't have any open API or anything similar. 00:32:57.238 --> 00:32:57.818 So if it's just 00:32:57.818 --> 00:33:02.618 like you have an implementation and then like some Jason happens, I guess then how 00:33:02.618 --> 00:33:04.018 would anyone know anything about this? 00:33:04.078 --> 00:33:07.698 And I think open API is an interesting way that it can add in more. 00:33:08.543 --> 00:33:10.883 You can, you can train an API on that a bit. 00:33:11.183 --> 00:33:14.483 And even if they don't understand what all of the objects are moving through. 00:33:15.698 --> 00:33:18.468 Part of something that a, the API community needs to do anyway. 00:33:18.598 --> 00:33:22.648 Part of my like 10 year plan, 20 year plan for like what I'm gonna be 00:33:22.728 --> 00:33:26.968 blathering about in API world has always been step one, get people defining 00:33:27.438 --> 00:33:28.928 what the hell this data is, right? 00:33:28.928 --> 00:33:32.048 Like open API isn't the end goal that we all, we do this 00:33:32.048 --> 00:33:33.128 and then everything's brilliant. 00:33:33.128 --> 00:33:36.848 Even though I talk about it plenty, it was was you have random Jason 00:33:36.988 --> 00:33:39.968 flying about the internet and no one knows what the fuck is going on. 00:33:40.158 --> 00:33:41.048 This is not great. 00:33:41.268 --> 00:33:45.103 So getting people to like . Write that down, make a contract that, that, that 00:33:45.103 --> 00:33:47.343 was step one and then, okay, great. 00:33:47.343 --> 00:33:50.063 We know what your implement, you know, your version of a tree and 00:33:50.063 --> 00:33:53.263 your version of a person and your version of a whatever account. 00:33:53.523 --> 00:33:56.303 But then like, can we try and have shared collections? 00:33:56.403 --> 00:33:56.823 And at 00:33:56.823 --> 00:33:59.583 stoplight we were doing design libraries where you can have like 00:33:59.743 --> 00:34:04.463 a repository of open API models you can reuse within your organization. 00:34:04.743 --> 00:34:07.478 'cause there was one company that had . 500 different 00:34:07.478 --> 00:34:08.958 versions of a flight, which is 00:34:08.958 --> 00:34:10.318 literally plane goes from there to there. 00:34:10.318 --> 00:34:11.598 Like that's not very complicated. 00:34:12.098 --> 00:34:14.308 With some seats, but there are a lot of different versions. 00:34:14.678 --> 00:34:18.158 And so they've, they've kind of squished that down to one version of a flight now. 00:34:18.178 --> 00:34:22.478 So it's literally describe the mess and then like reduce the, 00:34:22.618 --> 00:34:27.238 the, the mess and then like reuse those outside of the organization 00:34:27.388 --> 00:34:31.078 with, you know, shared data models and, and concepts like schema.org. 00:34:31.078 --> 00:34:32.278 But maybe not that 'cause it's like. 00:34:33.323 --> 00:34:36.803 A lot of that is biological medical data, which doesn't seem very 00:34:36.803 --> 00:34:38.203 helpful, , but that kind of concept. 00:34:38.943 --> 00:34:41.723 And then after that, like everyone is more trained, Haos becomes 00:34:41.723 --> 00:34:43.203 more useful in that world because 00:34:43.303 --> 00:34:45.923 you don't need to train it to know exactly what that URL is. 00:34:45.923 --> 00:34:48.443 You train it to know what these generic models are and how you work with them. 00:34:48.693 --> 00:34:51.893 But it also makes AI more useful 'cause it then has that same shared 00:34:51.893 --> 00:34:53.013 knowledge of what's coming back. 00:34:53.153 --> 00:34:56.573 So I, I like that that approach seems to be. 00:34:57.388 --> 00:34:58.768 The same as my existing approach. 00:34:59.468 --> 00:35:00.808 , I'd have to change my thoughts too much. 00:35:00.808 --> 00:35:01.128 Great. 00:35:01.628 --> 00:35:05.408 But yeah, the other, there is a lot of like magical Wi Wizard stuff in 00:35:05.408 --> 00:35:08.728 the world of AI and I do wanna kind of get people to be a bit cautious 00:35:08.778 --> 00:35:10.168 about what they go all in on. 00:35:10.368 --> 00:35:10.648 'cause it, 00:35:10.838 --> 00:35:11.888 some of it doesn't seem very helpful. 00:35:12.683 --> 00:35:13.103 Lorna Mitchell: agreed. 00:35:13.378 --> 00:35:17.188 Track 1: Last bit we were gonna talk about was open API four. 00:35:17.548 --> 00:35:19.068 I have been very out of the loop on this. 00:35:19.168 --> 00:35:22.348 It has a fun name, moonwalk, what's going on, help. 00:35:22.593 --> 00:35:22.883 Lorna Mitchell: Yeah. 00:35:22.903 --> 00:35:25.003 I'm not sure why it needed a project code name. 00:35:25.463 --> 00:35:29.176 But I think we've been kicking this around for a while. 00:35:29.336 --> 00:35:31.856 I mean, I went off and spent two years working in the 00:35:31.976 --> 00:35:33.056 database space and came back. 00:35:33.056 --> 00:35:35.256 We were still talking about it, so it's not going anywhere. 00:35:36.161 --> 00:35:38.481 Track 1: I was gonna say you vanished for a while and then came back to open 00:35:38.581 --> 00:35:39.141 Lorna Mitchell: I did, I had 00:35:39.241 --> 00:35:39.601 Track 1: said yay. 00:35:39.661 --> 00:35:41.901 Lorna Mitchell: absolutely lovely time doing loads of derell with loads of 00:35:41.901 --> 00:35:43.621 open source databases at Ivan, which was 00:35:43.621 --> 00:35:43.781 great. 00:35:43.841 --> 00:35:45.221 Now I'm back in the API space. 00:35:45.221 --> 00:35:46.101 That's great as well. 00:35:46.321 --> 00:35:47.621 Did I mention I'm an optimist? 00:35:48.321 --> 00:35:51.341 So news, news from the open API community. 00:35:51.341 --> 00:35:53.041 Then I mean, I think I. 00:35:53.351 --> 00:35:57.396 There's news in both the V three and the V four worlds. 00:35:57.576 --> 00:36:00.636 The V three news is just, we are seeing a lot of adoption. 00:36:00.696 --> 00:36:02.476 We are seeing people using 3.1. 00:36:03.236 --> 00:36:06.596 I think some of those educational resources are starting to get out there. 00:36:07.076 --> 00:36:10.156 Companies like Redley, and, but it takes the whole ecosystem 00:36:10.696 --> 00:36:15.636 and, and we, I think even though we are commercial competitors, we know we share a 00:36:15.636 --> 00:36:17.316 community and we collaborate really well. 00:36:17.806 --> 00:36:20.226 I'm sure that's true in other industries as well. 00:36:20.766 --> 00:36:24.226 But, so yeah, three, going from strength to strength, new tools all the time. 00:36:25.006 --> 00:36:26.936 General goodness, four. 00:36:28.116 --> 00:36:32.516 We've just put out a blog post from the Open API initiative committing to 00:36:32.516 --> 00:36:36.076 launching a version of four in 2024. 00:36:37.191 --> 00:36:37.481 Track 1: Okay. 00:36:37.481 --> 00:36:39.041 Getting people ready for it instead 00:36:39.111 --> 00:36:39.601 Lorna Mitchell: ticking. 00:36:40.101 --> 00:36:40.321 Yep. 00:36:40.861 --> 00:36:43.801 And there is, if anyone wants to see, we'll put it in the show notes. 00:36:43.831 --> 00:36:49.561 There's a repository full of discussions on GitHub about what we want to solve. 00:36:50.691 --> 00:36:52.771 At a high level um, it's, we are 00:36:52.771 --> 00:36:58.811 looking at simplifying the structure, making it a bit more resource oriented 00:36:59.431 --> 00:37:01.771 and a little bit more approachable as well. 00:37:02.461 --> 00:37:02.811 Jason 00:37:02.996 --> 00:37:06.016 Track 1: Just to explain on that a second, the resource orientated, meaning 00:37:06.016 --> 00:37:07.536 it's a bit less about like paths. 00:37:07.536 --> 00:37:08.296 What are your paths? 00:37:08.856 --> 00:37:12.776 'cause that's been a, a concern about . Not really being very resty. 00:37:13.376 --> 00:37:17.016 'cause like paths, some people will say that in a rest API like paths 00:37:17.016 --> 00:37:18.416 are the least important thing. 00:37:18.756 --> 00:37:22.576 And they could literally be like a random MD five Check sum is the most extreme 00:37:22.576 --> 00:37:26.016 example I've heard someone say because it really is just, it's about interacting 00:37:26.016 --> 00:37:28.456 with resources as like a state machine. 00:37:29.076 --> 00:37:33.336 And the homepage is a resource that then like lets you see other 00:37:33.616 --> 00:37:36.376 resources you could interact with in various ways and describes how. 00:37:36.826 --> 00:37:40.406 And so the fact that open API has always been like, tell me your paths, 00:37:40.966 --> 00:37:44.086 resources are optional, has always felt a little bit RPC ish to many people. 00:37:44.226 --> 00:37:44.446 So 00:37:44.666 --> 00:37:45.566 is that something that's 00:37:45.941 --> 00:37:46.161 Lorna Mitchell: so 00:37:46.871 --> 00:37:50.441 well, and the goal I think, with moonwalk is to get away from 00:37:50.611 --> 00:37:54.881 paths so that we can include more types of API, including the ones 00:37:54.881 --> 00:37:56.721 that are less restful. 00:37:57.181 --> 00:37:57.401 So 00:37:57.401 --> 00:38:02.201 with RPC, where you have one end point and maybe different query parameters or. 00:38:03.046 --> 00:38:05.936 Body data makes different actions happen. 00:38:06.596 --> 00:38:09.416 You can't represent that right now in 00:38:09.416 --> 00:38:10.256 open API. 00:38:10.756 --> 00:38:16.236 If you can just have, it's like the path and verb combination. 00:38:16.236 --> 00:38:17.156 Is it get, or is it post? 00:38:17.156 --> 00:38:17.836 What's the exact 00:38:17.866 --> 00:38:18.156 path? 00:38:18.536 --> 00:38:22.236 Not including query parameters, that's your unique identifier. 00:38:22.686 --> 00:38:26.866 Track 1: Yeah, someone had to make open api sorry, open RPC, which is like a 00:38:26.866 --> 00:38:30.266 copy paste of open API, but tweaked it so that you could do RPC stuff. 00:38:30.406 --> 00:38:30.626 And 00:38:30.626 --> 00:38:31.826 so yeah, that would be nice if. 00:38:32.231 --> 00:38:32.521 Lorna Mitchell: Yeah. 00:38:32.901 --> 00:38:36.201 And looking more at, there are more, the more things than that, that go into 00:38:36.271 --> 00:38:38.001 what we are now calling a signature. 00:38:38.391 --> 00:38:39.681 Like what endpoint is this? 00:38:39.981 --> 00:38:42.081 How would you root it in your code? 00:38:42.501 --> 00:38:48.201 You know, does your API do different things on header or on some other 00:38:48.531 --> 00:38:50.521 thing that we could take into account. 00:38:50.671 --> 00:38:55.881 Otherwise, you end up with this very lots of polymorphism because you can't. 00:38:56.651 --> 00:39:00.336 Correctly represent that these two things look completely different. 00:39:00.396 --> 00:39:02.936 If it's, you know, completed orders or incomplete orders, 00:39:03.716 --> 00:39:05.936 it might be a filter on an end point, but you're gonna get 00:39:05.936 --> 00:39:07.216 different types of object back. 00:39:08.096 --> 00:39:11.096 Moonwalk is gonna cater for that as well. 00:39:11.546 --> 00:39:11.946 But there's 00:39:11.946 --> 00:39:13.826 a thriving community, lots of chats. 00:39:14.116 --> 00:39:15.656 Th three isn't going anywhere. 00:39:15.656 --> 00:39:17.176 That's a lively community too. 00:39:17.786 --> 00:39:21.366 And as with the open API initiative, we're looking more at how we can 00:39:22.181 --> 00:39:24.786 Better respond to questions in that community, stuff like that. 00:39:24.986 --> 00:39:28.026 I have triaged on those repos and just trying to get everyone 00:39:28.026 --> 00:39:28.986 to be a bit more present. 00:39:30.371 --> 00:39:30.661 Track 1: Nice. 00:39:31.181 --> 00:39:33.381 I will have to dive in a little bit more. 00:39:33.501 --> 00:39:36.911 I mean, bloody Yeah, we had Darryl Miller on on the episode on the podcast, and I 00:39:36.911 --> 00:39:39.431 didn't ask him a single question about it, but we're talking about other stuff. 00:39:40.151 --> 00:39:43.711 I think the reason they came up with moonwalk was that they wanted a bit of 00:39:43.751 --> 00:39:48.351 a experimental playground where they could just try and hash out some ideas 00:39:48.351 --> 00:39:49.271 without everyone going. 00:39:49.271 --> 00:39:50.351 That's definitely what's happening. 00:39:50.351 --> 00:39:51.631 Panic, panic, don't do four. 00:39:52.001 --> 00:39:53.791 Um, Which would definitely happen. 00:39:53.961 --> 00:39:54.711 We've had that before. 00:39:55.161 --> 00:39:56.141 So I can understand that. 00:39:56.141 --> 00:40:00.021 And it sounds like it's kind of got enough agreement amongst the people 00:40:00.021 --> 00:40:02.181 that were paying attention that it. 00:40:02.861 --> 00:40:05.201 Is moving towards possibly actually being four. 00:40:05.221 --> 00:40:06.361 So that's, that's what you want. 00:40:06.651 --> 00:40:07.851 So we'll just have a normal name. 00:40:07.851 --> 00:40:09.851 It will just be called open API four. 00:40:10.061 --> 00:40:12.051 We're not doing like cool branding now 00:40:12.351 --> 00:40:12.641 Lorna Mitchell: Well, 00:40:12.681 --> 00:40:12.801 a 00:40:13.016 --> 00:40:13.496 Track 1: that's cool. 00:40:13.721 --> 00:40:14.721 ' Lorna Mitchell: cause you're actually walking backwards. 00:40:14.841 --> 00:40:15.281 I don't know. 00:40:15.701 --> 00:40:15.991 Track 1: Yeah. 00:40:16.231 --> 00:40:19.151 , this is, this is several very casual steps backwards. 00:40:19.591 --> 00:40:19.871 Brilliant. 00:40:20.221 --> 00:40:20.671 Well, yeah. 00:40:20.671 --> 00:40:20.991 Great. 00:40:21.051 --> 00:40:22.151 I'm glad to hear those changes. 00:40:22.391 --> 00:40:26.664 I mean, one, one, . Small thing in my head is that. 00:40:26.831 --> 00:40:29.701 There are certain things that open API won't let you do that. 00:40:29.731 --> 00:40:31.901 I've always been quite happy about, in a 00:40:31.901 --> 00:40:35.501 way, cause like open API will let you perfectly describe a bad API. 00:40:35.681 --> 00:40:38.461 But there are some of the things you shouldn't be doing that 00:40:38.461 --> 00:40:39.741 you can't currently describe. 00:40:40.161 --> 00:40:43.261 And so by being able to describe more of those things you shouldn't be doing, 00:40:43.441 --> 00:40:46.941 I'm a little bit like flexibility is important so that you can get 00:40:46.941 --> 00:40:50.021 all the old APIs on and then you can edge them in the right direction 00:40:50.051 --> 00:40:52.031 with, you know automated linters. 00:40:52.281 --> 00:40:54.621 And, and get them doing the right conventions that way. 00:40:54.801 --> 00:40:57.021 But part of it was also you. 00:40:57.281 --> 00:40:58.261 Man, why did you do that? 00:40:58.261 --> 00:40:59.021 Please don't do that. 00:40:59.021 --> 00:40:59.821 Change that quickly. 00:41:00.871 --> 00:41:02.161 Lorna Mitchell: Yeah, I think that's a, I think 00:41:02.161 --> 00:41:03.081 that's a real concern. 00:41:03.541 --> 00:41:07.161 And, you know, I want, I wanna represent and respect everyone else's 00:41:07.161 --> 00:41:08.721 work, but looking at the adoption for 00:41:08.721 --> 00:41:14.561 3.1, which was a very obvious improvement and upgrade for 3.0 users. 00:41:15.214 --> 00:41:16.784 some tools were available quite quickly. 00:41:17.104 --> 00:41:18.304 I know some took some time. 00:41:19.259 --> 00:41:22.264 There's a whole raft of new tools have come out for 3.1. 00:41:22.274 --> 00:41:24.104 We're really seeing the adoption now. 00:41:24.754 --> 00:41:27.244 4.0 doesn't offer anything to existing users. 00:41:27.723 --> 00:41:27.943 Track 1: Mm. 00:41:28.158 --> 00:41:30.278 Lorna Mitchell: brings more people into the fold, and if 00:41:30.278 --> 00:41:31.638 you're an existing tools vendor. 00:41:32.213 --> 00:41:33.843 Which obviously that's my job now. 00:41:34.759 --> 00:41:37.019 I'm not sure where we're going with that. 00:41:37.129 --> 00:41:39.699 It's going to be significantly more difficult to build for 00:41:40.279 --> 00:41:42.259 as things look at this very early stage. 00:41:43.119 --> 00:41:46.979 So I am supporting everything that's happening in the Open a BI initiative. 00:41:47.799 --> 00:41:50.239 Very excited to see just better. 00:41:50.608 --> 00:41:55.218 Education, better appreciation in the industry for what we do with 00:41:55.218 --> 00:41:57.458 Open API three and 3.1 and how it can 00:41:57.458 --> 00:41:57.658 help. 00:41:58.558 --> 00:42:00.138 And watching what's happening in four. 00:42:01.238 --> 00:42:05.498 It takes a long time to write these specifications and for tools to follow. 00:42:06.118 --> 00:42:07.778 So nobody needs to panic. 00:42:08.283 --> 00:42:08.573 Track 1: Yeah. 00:42:08.573 --> 00:42:08.813 Yeah. 00:42:08.813 --> 00:42:09.933 It's not appearing anytime now. 00:42:09.973 --> 00:42:15.023 I mean, even if, even if, didn't you say that mo version four is committed 00:42:15.023 --> 00:42:16.943 to appearing sometime in 2024? 00:42:17.213 --> 00:42:19.183 Well, that's, that's 12 months long, isn't it? 00:42:19.523 --> 00:42:19.743 So 00:42:20.663 --> 00:42:24.203 Lorna Mitchell: And I don't know, you know, open API initiative does 00:42:24.203 --> 00:42:27.683 not build tools, and I'm not, it's not clear to me right now who will. 00:42:27.713 --> 00:42:28.063 Track 1: Right. 00:42:28.533 --> 00:42:29.303 Well, that's the thing. 00:42:29.403 --> 00:42:32.783 So yeah, like in a year's time we could have a spec out and 00:42:32.783 --> 00:42:33.863 then still no one will use it. 00:42:34.003 --> 00:42:36.643 'cause it takes a long time for, for tool and vendors to catch up. 00:42:36.973 --> 00:42:40.653 Which is why it's important to like, you know, plow on whilst, but some 00:42:40.653 --> 00:42:42.253 tools have only just upgraded to 3.1. 00:42:42.253 --> 00:42:43.293 Yeah, this all takes a long time. 00:42:43.313 --> 00:42:45.653 So like, someone needs to be thinking about the future instead of just 00:42:45.653 --> 00:42:46.733 going, don't change anything. 00:42:46.763 --> 00:42:47.253 It's scary. 00:42:47.873 --> 00:42:48.313 Lorna Mitchell: Me it is 00:42:48.313 --> 00:42:50.793 about doing good rep, good support 00:42:51.693 --> 00:42:52.473 for 3.1. 00:42:52.743 --> 00:42:56.353 Lots of people are coming into APIs or raising their API 00:42:56.383 --> 00:42:59.353 game now with what's available today. 00:42:59.853 --> 00:43:00.273 And I 00:43:00.273 --> 00:43:02.593 think, you know, I work in developer experience. 00:43:02.633 --> 00:43:06.193 I really care about stuff that you can do today, and so that's my focus. 00:43:06.728 --> 00:43:07.008 Track 1: absolutely. 00:43:07.278 --> 00:43:07.568 Yeah. 00:43:07.568 --> 00:43:08.208 Gotta do both. 00:43:08.428 --> 00:43:12.168 But as long as, as long as the new one thinks about migration path, like 00:43:12.168 --> 00:43:15.688 as long as, as long as it's made in such a way where there is a migration 00:43:15.688 --> 00:43:18.448 path and, and that can be automated, I think that's pretty important. 00:43:18.908 --> 00:43:19.528 But also. 00:43:19.608 --> 00:43:20.728 Lorna Mitchell: is one of the requirements. 00:43:21.188 --> 00:43:21.408 Yes, 00:43:21.808 --> 00:43:24.138 Track 1: Okay, good , because various tooling vendors. 00:43:24.358 --> 00:43:24.928 Lorna Mitchell: They got it. 00:43:25.048 --> 00:43:25.338 Track 1: Yeah. 00:43:25.368 --> 00:43:28.778 Well, Mike Ralph's there saying like, he, you know, he, he made swagger to 00:43:28.778 --> 00:43:32.258 open API and he is gonna make open API three to moonwalk or whatever. 00:43:32.258 --> 00:43:33.178 Like he, he's on it. 00:43:33.588 --> 00:43:35.398 But yeah, that, that is important. 00:43:35.738 --> 00:43:39.728 And then the other thing, like Eric Wild made this really good 00:43:39.728 --> 00:43:42.728 point at a API days Paris about how 00:43:43.933 --> 00:43:48.543 Async API released a async passer, and it is like the defacto passer. 00:43:48.543 --> 00:43:50.343 You can pass a document any way you like. 00:43:50.943 --> 00:43:54.243 You can also use theirs, and I'm sure it's just written in, in TypeScript and nothing 00:43:54.243 --> 00:43:59.403 else, but this thing, this thing exists and, and the majority of tools can use it. 00:43:59.403 --> 00:44:03.403 And an open API majority of the tools are TypeScript or no something. 00:44:03.703 --> 00:44:06.483 So I'm, I'm really thinking that like that is the next most important step 00:44:06.483 --> 00:44:10.683 is for somebody probably Mike, let's be honest, he loves to write some code. 00:44:11.523 --> 00:44:13.923 Somebody to uh, just volunteered him either way. 00:44:14.183 --> 00:44:16.503 Somebody to write like an open API passer, that is the default. 00:44:17.083 --> 00:44:18.963 'cause there's been lots of these other little ones here and there. 00:44:18.963 --> 00:44:21.923 Like I've just deprecated one of them that I awkwardly inherited from some 00:44:21.923 --> 00:44:24.363 guy I tried to help and then, and then I got stuck doing all the code and I 00:44:24.363 --> 00:44:25.163 was like, I don't have time for this. 00:44:25.493 --> 00:44:28.533 There's a lot of these like awkward ones floating about, but if there can be one 00:44:28.533 --> 00:44:30.553 sample represent you know implementation. 00:44:31.388 --> 00:44:34.513 Initially written by a person, but as more people use it, more 00:44:34.513 --> 00:44:35.513 people will help maintain it. 00:44:35.973 --> 00:44:39.873 And then that can be okay, you've got, you've got your passer that works for 00:44:39.873 --> 00:44:43.553 3.1, but there's this other passer over here that works for four and that 00:44:43.553 --> 00:44:45.073 makes it a lot easier to switch to. 00:44:45.193 --> 00:44:46.913 'cause you don't have to rebuild your passer to support 00:44:47.023 --> 00:44:48.713 four, you just use the thing. 00:44:49.253 --> 00:44:52.873 And then you don't have to worry about all the intricacy, awkward bits of yaml. 00:44:52.873 --> 00:44:54.833 It's just you have some models to, to work with. 00:44:55.193 --> 00:44:56.233 I really hope they can do that. 00:44:56.233 --> 00:44:57.993 And I'll be nagging people to give that a try. 00:44:58.213 --> 00:45:00.618 Without volunteering myself 'cause I'm crap at code these days. 00:45:01.773 --> 00:45:04.773 Lorna Mitchell: I think it's a tricky one, you know, 'cause I work for a 00:45:04.773 --> 00:45:07.253 tools vendor now, so I was like, oh, you are gonna build some of the tools. 00:45:07.273 --> 00:45:10.093 Do you wanna tell me what else you're gonna build so that we don't 00:45:10.093 --> 00:45:12.333 invest our commercial expertise in 00:45:12.358 --> 00:45:12.648 Track 1: Yeah. 00:45:12.953 --> 00:45:16.863 Lorna Mitchell: you know, and I'm .I'm not really sure if the proliferation of many 00:45:16.863 --> 00:45:19.583 tools in open API is a feature or a bug. 00:45:19.603 --> 00:45:23.023 But I think feature, I think the competition has helped because 00:45:23.533 --> 00:45:25.183 it's very open source, isn't it? 00:45:25.183 --> 00:45:25.703 It's, it's a, 00:45:25.703 --> 00:45:26.623 it's an open standard. 00:45:26.933 --> 00:45:28.183 There's loads of different tools. 00:45:28.563 --> 00:45:32.143 You know, open API tools tells that story loud and clear. 00:45:32.363 --> 00:45:35.343 You can choose, it's in your tech stack. 00:45:35.493 --> 00:45:36.903 There's always something. 00:45:37.213 --> 00:45:37.703 Whereas 00:45:37.863 --> 00:45:41.743 I think if we go to one official version, we're tied down to the speed that the 00:45:42.248 --> 00:45:45.893 Central initiative works at, which isn't always the same speed the industry does 00:45:46.883 --> 00:45:49.813 Track 1: Yeah, I mean, hey, maybe that could be a working group right in, 00:45:49.813 --> 00:45:53.213 in, instead of it just being , very overworked, Mike and trying to do 00:45:53.213 --> 00:45:55.453 the thing to make all the vendors happy, who are all getting paid to 00:45:55.453 --> 00:45:56.613 work on their tools and he's not. 00:45:56.953 --> 00:45:59.543 Maybe that could be a working group where we're like, let's hash out like 00:45:59.543 --> 00:46:01.623 what we need from a, a shared passer. 00:46:01.843 --> 00:46:05.183 And then if anyone in the group ends up not really getting something that. 00:46:05.513 --> 00:46:08.323 They're happy with and they can just go and make their own however they want. 00:46:08.423 --> 00:46:12.603 But yeah, definitely it is a difficult line in general with like, should the open 00:46:12.643 --> 00:46:14.363 API foundation be making all this stuff? 00:46:14.563 --> 00:46:16.563 'cause they're like, well we've just made a documentation tool. 00:46:17.383 --> 00:46:19.563 You know, like that would be pretty messed up. 00:46:19.583 --> 00:46:22.273 Lorna Mitchell: well, and the members are documentation tool 00:46:22.743 --> 00:46:25.753 vendors, so that's awkward Async. 00:46:25.793 --> 00:46:28.833 KPII have the opposite one, and you might have seen this as well from 00:46:29.283 --> 00:46:31.613 having Async KPI support spectral. 00:46:32.468 --> 00:46:35.418 Reduc, CLI has Async API support as well, just for linting. 00:46:35.418 --> 00:46:41.448 We don't have it in docs yet, and honestly there's the Async API community doesn't 00:46:41.448 --> 00:46:42.848 really talk about anyone else's tools. 00:46:42.908 --> 00:46:44.528 It doesn't have the same ecosystem. 00:46:44.878 --> 00:46:46.248 They just build their own tools. 00:46:46.248 --> 00:46:47.728 They're all quite closely coupled. 00:46:47.728 --> 00:46:48.528 It's the same people. 00:46:48.853 --> 00:46:49.203 Track 1: Right. 00:46:49.533 --> 00:46:50.463 Lorna Mitchell: Postman pays them. 00:46:50.463 --> 00:46:51.263 That's what you get. 00:46:51.763 --> 00:46:56.383 And so it's very, very different despite the fact that those two projects have 00:46:56.383 --> 00:46:58.223 quite a lot of shared philosophy. 00:46:58.573 --> 00:47:00.743 They also don't, and the tools is 00:47:00.783 --> 00:47:01.863 a really big differentiator. 00:47:02.439 --> 00:47:03.329 Track 1: That is interesting. 00:47:03.759 --> 00:47:07.089 Well, still, I'm always excited about the potential for collaboration and 00:47:07.089 --> 00:47:09.719 at some point I'll be picking your brains about I chatted to all the 00:47:09.719 --> 00:47:12.879 other API linkers and we're all gonna have a conversation about like coming 00:47:12.879 --> 00:47:14.599 up with a shared rule set format. 00:47:14.739 --> 00:47:17.759 And even if it's not one that anyone particularly wants to make their their 00:47:17.759 --> 00:47:21.239 default, it's just like, can we make some tools that swap them in all directions? 00:47:21.559 --> 00:47:22.479 'cause that'll be cool. 00:47:23.019 --> 00:47:26.599 Although I gotta talk to the optic team about, they've just come up with um, 00:47:26.963 --> 00:47:30.923 lint, GPT, where you just write your rules as a, an array of strings and 00:47:30.923 --> 00:47:32.283 hope that the wizard is consistent. 00:47:32.383 --> 00:47:32.603 But 00:47:32.783 --> 00:47:34.123 I'm gonna give that to all a try 00:47:34.238 --> 00:47:36.278 Lorna Mitchell: I think Optic currently have the best 00:47:36.528 --> 00:47:40.238 converters, so they understand this problem space, but because. 00:47:40.293 --> 00:47:40.583 Track 1: Yeah. 00:47:40.613 --> 00:47:40.903 Yeah. 00:47:41.428 --> 00:47:45.298 Lorna Mitchell: Redoc, CLI, all our linting is very like 00:47:45.538 --> 00:47:47.498 a ST data structure driven. 00:47:47.888 --> 00:47:48.378 It's not 00:47:49.138 --> 00:47:51.778 JSON path driven like the others. 00:47:52.838 --> 00:47:55.938 My, I had a quick look and I was like, oh, this is gonna be really difficult. 00:47:55.938 --> 00:47:59.578 Interrupt because we are not describing, you 00:47:59.578 --> 00:48:03.298 know, we're like it, oh, you have a response example here. 00:48:03.768 --> 00:48:05.418 This is what we do with this response examples. 00:48:05.418 --> 00:48:08.338 It's not like it's this path, this path, this path. 00:48:08.978 --> 00:48:09.328 Track 1: Right. 00:48:09.563 --> 00:48:11.333 Lorna Mitchell: Turn, turn left at this array. 00:48:11.333 --> 00:48:13.876 Then take the third success response element. 00:48:14.056 --> 00:48:14.596 You know, like 00:48:14.686 --> 00:48:14.976 Track 1: Yeah. 00:48:15.016 --> 00:48:15.296 I gotcha. 00:48:15.616 --> 00:48:17.256 cause it's based on, it's based on OAS kit. 00:48:17.256 --> 00:48:17.376 Right. 00:48:17.376 --> 00:48:18.256 And I remember how that works. 00:48:18.256 --> 00:48:21.096 Like OAS kit just walks through and then goes, oh, I found one of these. 00:48:21.096 --> 00:48:21.976 Anyone care about this? 00:48:21.976 --> 00:48:22.576 Yeah, you do. 00:48:22.576 --> 00:48:22.816 Cool. 00:48:22.836 --> 00:48:23.616 Run some rules on it. 00:48:23.616 --> 00:48:27.181 Whereas we're like, go through the rules and then like . Delve into the 00:48:27.181 --> 00:48:29.101 depths following this arcane direction. 00:48:29.471 --> 00:48:31.791 I shouldn't have opened up this can of worms so late in the podcast. 00:48:32.041 --> 00:48:32.361 We like 00:48:32.416 --> 00:48:33.016 Lorna Mitchell: we're gonna still 00:48:33.041 --> 00:48:35.721 Track 1: 35 minutes normally, and I yeah, I could, I could 00:48:35.721 --> 00:48:36.841 talk to you for hours, honestly. 00:48:36.951 --> 00:48:40.141 Yeah, it it's really great to have you on the show and we'll have to get you back on 00:48:40.141 --> 00:48:41.301 to talk about other stuff in the future. 00:48:41.401 --> 00:48:41.741 For sure. 00:48:41.941 --> 00:48:42.231 Lorna Mitchell: That 00:48:42.231 --> 00:48:42.751 would be awesome. 00:48:43.061 --> 00:48:43.541 Track 1: for coming, 00:48:43.751 --> 00:48:44.311 Lorna Mitchell: for having me. 00:48:44.484 --> 00:48:44.904 Track 1: Cheers. 00:48:45.084 --> 00:48:45.304 Bye.