1 00:00:06,799 --> 00:00:11,080 Hello, Welcome to Adventures and Angular. It's been a while. This is 2 00:00:11,160 --> 00:00:15,519 the podcast where you keep you updated on all things Angular related. This show 3 00:00:15,640 --> 00:00:18,879 is produced by two companies, Top and Doves and Onoid. Top and Doves 4 00:00:18,920 --> 00:00:22,079 is where we great Top and Doves we get top and pay and recognition while 5 00:00:22,120 --> 00:00:27,280 working on interesting problems and making meaningful community contributions. And Onvoid, which offers 6 00:00:27,399 --> 00:00:31,440 more design and software development services on a task basis, so clients only pay 7 00:00:31,480 --> 00:00:37,359 when tests are delivered and approved. In today's episode, we will talk about 8 00:00:37,920 --> 00:00:43,799 reactiveness and Angular and how you can still use urxgs, so that's correct. 9 00:00:43,920 --> 00:00:48,000 We're not necessarily going to talk about signals today, although we might touch on 10 00:00:48,039 --> 00:00:56,479 the subject, but how you can simplify your reactivity while still using rxgs and 11 00:00:57,119 --> 00:01:03,040 well, for that we have a variouspect guests. So Dean Radcliffe is a 12 00:01:03,239 --> 00:01:07,879 web developer of many years and the creator of our ex facts, which is 13 00:01:08,000 --> 00:01:14,400 the library we're going to be talking about throughout this episode. So we're going 14 00:01:14,480 --> 00:01:19,239 to get back to Dean in a second. But also joining me in today's 15 00:01:19,239 --> 00:01:26,319 episode is Arman Vardanian. Hey, everyone, first episode this year. I 16 00:01:26,439 --> 00:01:32,319 missed you, yeah, and well I am Lucas Paganini, one of the 17 00:01:32,480 --> 00:01:40,680 other hosts in the show. So let's get started. So Dean, yes, 18 00:01:40,719 --> 00:01:47,920 sir, please pitch our XFCX to the audience. All right, our 19 00:01:48,079 --> 00:01:57,760 xffcts, that's our ex FX is framework independent state management and effect management with 20 00:01:57,920 --> 00:02:04,040 concurrency control like our extra has, and it is built on r XJS. 21 00:02:04,280 --> 00:02:09,199 It's interoperable with r XJS, but it brings a better developer experience and a 22 00:02:09,319 --> 00:02:15,560 shorter learning curve than using only r x JS. Compared to r xjs, 23 00:02:16,039 --> 00:02:23,639 RX effects will still concurrency control your ACYNC functions like with those concat map switch 24 00:02:23,759 --> 00:02:28,919 map operators, but in a way that will be simpler, safer, and 25 00:02:29,080 --> 00:02:35,479 with greater readability. Okay, that's really interesting, And I was I was 26 00:02:35,560 --> 00:02:39,199 actually taking a look at the MEPO so the route. I noticed that, 27 00:02:39,400 --> 00:02:44,879 like in the peckyjson file, you don't have RXS as a dependency. So 28 00:02:45,800 --> 00:02:49,120 I do have our xs as a dependency, but not in the root. 29 00:02:50,000 --> 00:02:55,520 Oh not in the route. Okay, repository is a is a the it's 30 00:02:55,560 --> 00:03:04,800 a like Lucas said, there's an installation of libraries and they the primary ones 31 00:03:04,919 --> 00:03:10,280 are going to be RX effects, slash bus, RX effects slash effect and 32 00:03:10,639 --> 00:03:20,240 rx effects slash service, and they each provide similar reactive and concurrency control benefits, 33 00:03:21,400 --> 00:03:28,039 but with a little bit of a flavor and different different mental models. 34 00:03:28,280 --> 00:03:31,800 So, for example, the bus is an event bus. It was actually 35 00:03:34,479 --> 00:03:43,639 derived from a simple rx JS event bus that the company I don't know, 36 00:03:43,719 --> 00:03:46,360 somebody posted I can't remember the name or the exact company name it was. 37 00:03:46,919 --> 00:03:51,240 I had to do with the RxJS creators. There was a simple event bus, 38 00:03:51,800 --> 00:03:57,520 and then I built on that event bus, and the event bus brings 39 00:03:57,800 --> 00:04:02,439 pub sub benefits. So so that's like you trigger an event somewhere and an 40 00:04:02,479 --> 00:04:09,080 event listener responds by running an async function, but that listener can be concurrency 41 00:04:09,199 --> 00:04:14,960 controlled. So imagine having your RxJS operators kind of baked in to the listener. 42 00:04:15,560 --> 00:04:23,000 So that's the bus model. It has the advantage of your code that's 43 00:04:23,079 --> 00:04:29,279 triggering events. Say you have a component that triggers an event when it's clicked 44 00:04:29,399 --> 00:04:33,319 to save something to a database. But now you want to send an analytics 45 00:04:33,399 --> 00:04:39,680 event that that button was clicked, right, So the published subscribe pub sub 46 00:04:39,839 --> 00:04:45,120 metaphor says you don't have to go into that component with the button and add 47 00:04:45,199 --> 00:04:49,480 code to it you can listen for that click event with a bus listener and 48 00:04:51,480 --> 00:04:57,000 then you know, send analytics events that way. And I've done this on 49 00:04:57,120 --> 00:05:00,560 an application that had many many questions gens. It was a survey application, 50 00:05:01,319 --> 00:05:06,800 and the goal was that once we developed a question type that we wanted to 51 00:05:06,959 --> 00:05:13,720 keep that component closed like closed to modification, we should be able to extend 52 00:05:14,680 --> 00:05:19,959 the behavior of the components without having to go in and modify the components. 53 00:05:20,480 --> 00:05:25,439 And so by having each component put an event on the event bus, it's 54 00:05:25,519 --> 00:05:30,279 easy to add event listeners like an analytics listener. And suppose you want to 55 00:05:30,360 --> 00:05:36,160 say your analytics events should be sent serially so that they don't take up more 56 00:05:36,240 --> 00:05:40,800 than one HTTP connection at a time, Well, then you just set the 57 00:05:40,879 --> 00:05:46,839 concurrency mode of that listener. And that also gets you out of the pickle 58 00:05:47,240 --> 00:05:51,800 of you if you were to modify the component and you'd have some code that 59 00:05:51,920 --> 00:05:56,959 says save to the database and some code that says analytics, Well, is 60 00:05:57,000 --> 00:06:00,399 there a correct order to put those calls? And if you put the analytics 61 00:06:00,480 --> 00:06:05,240 call first and that analytics call errors for some reason, is that going to 62 00:06:05,360 --> 00:06:09,759 keep it? Are you going to keep from needing to you know, is 63 00:06:09,759 --> 00:06:14,639 that going to prevent the database call from functioning. So it's actually much cleaner 64 00:06:15,120 --> 00:06:18,560 not to keep piling things into the one component, but to set up separate 65 00:06:18,639 --> 00:06:25,279 listeners. And another benefit of our x effects is error handling. So if 66 00:06:25,360 --> 00:06:31,360 that analytics event listener acts up, it doesn't kill the safe to database event 67 00:06:31,480 --> 00:06:39,519 listener. So you have separation through the pub sub metaphor like even more complete 68 00:06:39,600 --> 00:06:46,639 separation and error handling then with raw rx JS. That's really that's really interesting 69 00:06:46,759 --> 00:06:53,040 because in ourhgs, error handling can be a bit of a pain point for 70 00:06:53,160 --> 00:06:59,600 lots of developers, Like there are are all of those tools. For example, 71 00:07:00,519 --> 00:07:03,720 we have retry, retry when right with the count we have, we 72 00:07:03,800 --> 00:07:10,439 could just schedule and so on, but some some of those tools, especially 73 00:07:10,480 --> 00:07:15,480 I think from the perspective of the documentation, aren't readily like obvious that those 74 00:07:15,879 --> 00:07:20,519 can be used for airror handling. And yeah, I'll honestly see lots of 75 00:07:20,560 --> 00:07:25,480 people who have worked with the ourgis for a while, but they would probably 76 00:07:25,519 --> 00:07:30,319 get confused by like what means we try when mm hmmm or something, or 77 00:07:30,360 --> 00:07:36,000 how to what scenarios you can have with you know, error handling. So 78 00:07:36,680 --> 00:07:41,199 yeah, there's this's this really sounds interesting. I'm just taking a look at 79 00:07:41,240 --> 00:07:49,439 the effects library. Yeah yeah, Well you mentioned pain points with our x 80 00:07:49,560 --> 00:07:56,480 g S and I was at a conference which side is a four years ago 81 00:07:56,759 --> 00:08:01,439 called our x JAS Live, and I basically follow the RxJS core team and 82 00:08:01,759 --> 00:08:09,199 I follow and listen to what people say the top pain points are with r 83 00:08:09,319 --> 00:08:11,680 x JS, and of course as I hear them, I think, well, 84 00:08:11,720 --> 00:08:16,639 maybe I could address that in rx effects, and I believe I have, 85 00:08:16,920 --> 00:08:22,040 and so I'll touch on some of them. Tracy Lee and Ben Lesh 86 00:08:22,600 --> 00:08:28,279 and j phelps the core people that give workshops on r x JS. What 87 00:08:28,519 --> 00:08:33,279 they're suggesting you can, you know, get stuck on in which they'll help 88 00:08:33,320 --> 00:08:37,639 you with. But our x effects will also help you with. Is okay, 89 00:08:37,799 --> 00:08:41,240 So subscription management, when do you subscribe? What do you do with 90 00:08:41,360 --> 00:08:48,480 subscription objects? How do you choose an operator? And is can cat map 91 00:08:48,679 --> 00:08:56,759 or exhaust map correct for your case? And error handling? So I think 92 00:08:56,799 --> 00:09:03,240 subscriptions, concurrency, and error handling are some of the biggest problems challenges people 93 00:09:03,360 --> 00:09:11,080 have with r x JS, and our XFX has has solutions to all of 94 00:09:11,159 --> 00:09:18,320 those with a simpler, developer friendly API. Oh one thing that I'm still 95 00:09:18,399 --> 00:09:24,440 wondering about. Like, I'm seeing lots of different examples in the repole, 96 00:09:24,440 --> 00:09:31,759 and I see very little RXGS in actuality. So is this meant to be 97 00:09:31,919 --> 00:09:37,919 like a sort of total wrapper around rh just like you don't really need to 98 00:09:37,080 --> 00:09:46,440 touch RGS itself for yes, but but you can grow to use all of 99 00:09:46,519 --> 00:09:52,000 the r xjs features. It's truly interoperable. But so in the show notes, 100 00:09:52,120 --> 00:09:58,879 I provided a link to a stack overflow post where I said, here's 101 00:09:58,919 --> 00:10:03,840 a situation. You have a bell ringing function. It plays some audio, 102 00:10:03,080 --> 00:10:09,639 and so it's async because it takes some time. Now you want to create 103 00:10:09,840 --> 00:10:13,879 a web form with a button that rings the bell. Now, if that 104 00:10:15,159 --> 00:10:20,679 button is wrung or clicked while the bell is ringing, you would have two 105 00:10:20,759 --> 00:10:24,279 bells ringing at once, and it might overload your audio and sound bad. 106 00:10:24,440 --> 00:10:30,320 So we want to concurrency control the bell like put them in a queue, 107 00:10:31,120 --> 00:10:33,919 or cut off the previous bell, or throttle like. So we want a 108 00:10:35,000 --> 00:10:39,799 concurrency control mechanism. And we also want cancelation because what if the user just 109 00:10:39,840 --> 00:10:45,600 wants to cancel on that bell. Simple example, there's a there's one button 110 00:10:45,759 --> 00:10:50,600 to ring, one button to cancel and a concurrency control. Well, I 111 00:10:50,840 --> 00:10:56,320 sketched out what I thought the RxJS version of that was, and it was 112 00:10:56,399 --> 00:11:05,480 about about fifteen lines. There were two subjects, there were various imported operators 113 00:11:05,600 --> 00:11:11,080 take until there actually needs to be three subjects, uh take until a switch 114 00:11:11,200 --> 00:11:16,000 map. It was. It was what I call operator soup and the rx 115 00:11:16,080 --> 00:11:22,720 effects. Given a function that returns a promise or an observable for ringing the 116 00:11:22,799 --> 00:11:28,200 bell the audio, it's you know, one line to wrap it in uh 117 00:11:28,799 --> 00:11:33,919 an effect with the create effect function, and it's one line to request it, 118 00:11:33,080 --> 00:11:37,320 which is the same for r XJS, one line to like trigger the 119 00:11:37,399 --> 00:11:43,480 bell ringing. But then cancelation is just dot cancel current and canceling. The 120 00:11:43,639 --> 00:11:50,279 entire queue is dot cancel, current and queue. So it's a question of 121 00:11:50,360 --> 00:11:56,120 when should you like deep dive into r x JS. Well, if you 122 00:11:56,240 --> 00:12:00,480 want to tap something, if you want to do retry, when if you 123 00:12:00,519 --> 00:12:03,480 want to use any of the fancy operators, if you want to zip two 124 00:12:03,759 --> 00:12:09,799 services together, that's fine. But as far as concurrency control and cancelation, 125 00:12:09,879 --> 00:12:15,639 like you shouldn't need to go into that much detail or have that many imports, 126 00:12:15,759 --> 00:12:18,840 Like think about load dash, how you just take your function, pass 127 00:12:18,919 --> 00:12:24,559 it into a load dash function, and get a new function out that has 128 00:12:24,639 --> 00:12:28,159 new behavior. It should be that simple. That's the premise of OURXFX. 129 00:12:31,039 --> 00:12:35,200 So once really got me interested when I was thinking about this in the context 130 00:12:35,240 --> 00:12:41,000 of Angular and jar X. Could I, for example, and I will 131 00:12:41,399 --> 00:12:46,120 tell you in a moment why I'm asking this, Like I have an observable 132 00:12:46,240 --> 00:12:50,960 that does something something, Can I just pick it up and turn it into 133 00:12:50,000 --> 00:12:56,120 an effect? Like I already have an observable that works in a certain way 134 00:12:58,200 --> 00:13:03,759 that I want to sort of turn into in an RX affects effect to be 135 00:13:03,080 --> 00:13:07,559 able to I don't know, cancel it or add something to it, or 136 00:13:07,320 --> 00:13:16,960 pose it or whatever capabilities it offers. Do you have the observable as like, 137 00:13:18,039 --> 00:13:22,240 do you have a function that returns an observable like an endpoint that returns 138 00:13:22,919 --> 00:13:28,639 an AJAX observable, Like I mean a function that returns an ajax observable observable? 139 00:13:30,000 --> 00:13:31,720 Yeah? Let me yeah, let me just give you the exact thing 140 00:13:31,799 --> 00:13:35,799 that I was that I had in moment. So for example, in JRX, 141 00:13:35,879 --> 00:13:39,840 we have the jur affects package, and it's very decent to create like 142 00:13:39,960 --> 00:13:45,600 effects based on actions and so on, but it doesn't have very many features. 143 00:13:45,679 --> 00:13:50,960 So essentially it provides with the function to sort of register and observable that 144 00:13:50,559 --> 00:13:56,519 j RX will listen to whatever you wanted to do, and it gives you 145 00:13:56,159 --> 00:14:03,639 the actions observable like whatever actions have in your I don't know application, and 146 00:14:03,759 --> 00:14:07,320 you just mix them and then you have to go and use a bunch of 147 00:14:07,480 --> 00:14:11,480 rs if you want something complex like ninety scenarios, just like I make an 148 00:14:11,559 --> 00:14:16,919 HDP request and despatch this or that action. But sometimes you need like oh, 149 00:14:16,279 --> 00:14:22,720 I want to retry this, is this something specific appensode or maybe in 150 00:14:22,879 --> 00:14:26,399 some cases I want to throw away the previous requests start a new one. 151 00:14:26,480 --> 00:14:30,200 But in other scenario, I want to do something else, and it becomes 152 00:14:30,320 --> 00:14:33,240 as you mentioned, like an operator soup. In the effects it becomes really 153 00:14:35,399 --> 00:14:37,799 so. And the funny thing is like injr X, when you create an 154 00:14:37,840 --> 00:14:43,200 effect, like you have functional effects. Now you have a variable that represents 155 00:14:43,279 --> 00:14:46,240 this effect, but you never use it. It would be great if you 156 00:14:46,559 --> 00:14:52,000 could bring it to your component and for example, in some scenario button less 157 00:14:52,039 --> 00:14:54,399 click and you could say, okay, cancel this effect or pose this effect 158 00:14:54,559 --> 00:14:58,600 or something in that regard. So I was thinking if I could take this 159 00:14:58,720 --> 00:15:03,279 egg observable that that is the thing that I need from juris. I need 160 00:15:03,360 --> 00:15:05,919 to know what actions have. Yeah, and just I don't know, plug 161 00:15:07,000 --> 00:15:13,519 it into this effect and maybe use it this way. So okay, I'm 162 00:15:13,639 --> 00:15:20,679 still maybe, but let me help clarify it. So there are observables that 163 00:15:20,919 --> 00:15:24,879 are like the sources of effects running and is that what? There is one 164 00:15:24,960 --> 00:15:30,799 observable, Yeah, Jelicks gives you one observable. It's called actions actions where 165 00:15:30,799 --> 00:15:33,799 an action is dispatched somewhere this action, will you know, send the new 166 00:15:33,840 --> 00:15:39,960 EMISI. So actions that sounds that sounds like a hot observable. Yes, 167 00:15:39,000 --> 00:15:45,720 that right, Okay, it's a hot observable. So with you generally wouldn't 168 00:15:46,559 --> 00:15:54,799 have that wrapped in RX effects. You would you would maybe trigger some effects 169 00:15:54,399 --> 00:16:00,200 based on a filtered set of actions, right, yeah. So so you 170 00:16:00,240 --> 00:16:04,879 have observables like that that are hot and they are like the sources of information. 171 00:16:06,720 --> 00:16:11,080 And then you have observables that I usually call cold, that are like 172 00:16:11,879 --> 00:16:15,600 doing the effect, doing the AJAX, ringing the ball in the in the 173 00:16:15,799 --> 00:16:21,399 context, in the context of the specific what ang X does. It gives 174 00:16:21,440 --> 00:16:26,840 you a custom operator called off type and you can pass it actions that you 175 00:16:26,960 --> 00:16:30,399 want. I want this action for example, if the load data action was 176 00:16:30,480 --> 00:16:36,320 triggered, then it will all sort of filter out and only start working. 177 00:16:36,440 --> 00:16:41,679 And then you give this entire thing to the create effect function that AnGR x 178 00:16:41,720 --> 00:16:45,039 does, and then it reduces these facts. I think it has one or 179 00:16:45,080 --> 00:16:49,360 two methods that you can use. You can definitely it has methods using human 180 00:16:49,440 --> 00:16:53,399 testing like but the object that it returns itself, it doesn't have very many 181 00:16:53,480 --> 00:16:59,279 other useful things. So that's what got me wondering that sort of enhance. 182 00:16:59,360 --> 00:17:03,839 It just you the same pattern but instead just fed this effect to something from 183 00:17:04,039 --> 00:17:10,200 RX effects, so that I will be able to try stuff on demand or 184 00:17:11,519 --> 00:17:15,759 you know, sort of modified concurrency or whatever without going deep into RGS operators. 185 00:17:17,440 --> 00:17:21,000 Yes, I mean, I hope so. But my big disclaimer is 186 00:17:21,079 --> 00:17:25,680 that I don't work in Angular much these days, though I have, and 187 00:17:26,319 --> 00:17:33,400 I developed this library when I was consulting and switching between Angular and React, 188 00:17:33,960 --> 00:17:37,400 and so I wanted something that was portable I could take with me. So 189 00:17:37,680 --> 00:17:41,880 as far as how to make this interact with n g r X, I 190 00:17:42,119 --> 00:17:48,079 have not explored that use case, but my general sense of it is that 191 00:17:48,880 --> 00:17:56,440 you would have any effect like ringing a bell. You would have your actions 192 00:17:56,480 --> 00:18:02,119 of type, and then you would basically subscribe to those actions of type and 193 00:18:02,279 --> 00:18:06,559 say when those come through, ring the bell, and now you have a 194 00:18:06,640 --> 00:18:11,640 concurrency controlled cancelable bell. On the one hand, it's independent, and then 195 00:18:11,680 --> 00:18:17,119 you might use RxJS if your actions are not U n g r X. 196 00:18:17,519 --> 00:18:22,559 If your actions were already available through n g RX, then you could just 197 00:18:22,799 --> 00:18:26,680 you know, send them to the to the service or effect. So that's 198 00:18:26,759 --> 00:18:30,319 probably how it would work, but I haven't tried it. Yeah, that's 199 00:18:30,359 --> 00:18:33,319 interesting. Like from the top of my head, what I could do is 200 00:18:33,559 --> 00:18:37,440 create one and jr X SOCI just takes all actions and sentences to like x 201 00:18:37,519 --> 00:18:42,880 effects service. Sounds like effect service, yeah or yeah. And so what 202 00:18:42,960 --> 00:18:45,880 would be different is like, if you were doing this purely an n g 203 00:18:47,119 --> 00:18:52,200 RX to cancel the bell ringing, you might have another action that you would 204 00:18:52,359 --> 00:18:56,400 dispatch to do it. But with rx effects you wouldn't need that. You 205 00:18:56,400 --> 00:19:00,839 would just go to the ringing bell service or effect and just say dot cancel 206 00:19:00,920 --> 00:19:07,519 current. So in some in some ways it factors some things out of the 207 00:19:07,759 --> 00:19:15,640 rx JS g r X land that you don't necessarily need to live in that 208 00:19:15,839 --> 00:19:19,960 land. So for example, the the like I just said, cancelation, 209 00:19:21,160 --> 00:19:22,960 you can just go directly to the ringing bell and cancel it, right. 210 00:19:25,440 --> 00:19:29,480 But another thing is the status of whether the bell is ringing or not. 211 00:19:29,640 --> 00:19:33,319 And there's a linked code sandbox in the RX effects documentation that has one of 212 00:19:33,359 --> 00:19:37,920 these bells you can play with, but you want to be able to display 213 00:19:37,000 --> 00:19:41,319 to the user whether the bell is ringing or not. And prior to my 214 00:19:41,960 --> 00:19:51,359 prior to my working on RX effects, almost and almost every app I did, 215 00:19:51,960 --> 00:19:56,519 I had a field in state called loading right or for a bell ringing, 216 00:19:56,559 --> 00:20:02,039 I would explicitly code that into the state model. But then I said, 217 00:20:02,480 --> 00:20:06,559 well, if you've got a promise that isn't resolved, if you've got 218 00:20:07,079 --> 00:20:11,839 a bell that's still ringing, how come we don't automatically have access to a 219 00:20:11,960 --> 00:20:17,079 field, like how do I have to build that into the state model of 220 00:20:17,440 --> 00:20:19,880 my app? Because the problem with that is, you know, you have 221 00:20:21,119 --> 00:20:23,960 all these examples that say, all right, well you're gonna have an angular 222 00:20:25,039 --> 00:20:30,359 component that has a button and then you're going to use the HTTP client to 223 00:20:30,720 --> 00:20:33,119 you know, fetch some stuff. So the first thing you do is you're 224 00:20:33,160 --> 00:20:37,160 gonna, you know, set is loading to true, then you're going to 225 00:20:37,240 --> 00:20:40,920 do the effect. Then when it's done, you're going to set is loading 226 00:20:41,000 --> 00:20:45,720 to false. And that manual tracking of the loading state is bundled into RX 227 00:20:45,759 --> 00:20:49,319 effects. You can track the loading state. So I also have an example 228 00:20:49,680 --> 00:20:53,880 of a counter that's asynchronous. It takes two seconds to increment the counter, 229 00:20:55,720 --> 00:21:00,440 and the state model only contains the count, but RX effects keeps track of 230 00:21:00,480 --> 00:21:07,799 whether it's executing or not. Yeah, that's what I was And of course 231 00:21:07,920 --> 00:21:15,799 we will now inevitably stumble into the signals topic because I was just yeah, 232 00:21:15,839 --> 00:21:19,799 because you know, signals got now very deep integration despite them being not being 233 00:21:21,039 --> 00:21:25,200 like even one year old. But we have now njr X signals, so 234 00:21:25,319 --> 00:21:30,279 we can build your yeah like state management without like without the reduct spattern, 235 00:21:30,440 --> 00:21:37,240 just with signals without observables. And I was thinking, like, the weak 236 00:21:37,359 --> 00:21:42,279 point for me in NJRX signal is probably the fact because effects are essentially the 237 00:21:42,400 --> 00:21:48,359 same that we have in angoric effects just baked into you know the signal land. 238 00:21:49,240 --> 00:21:53,000 Yeah, and yeah, that got me thinking, especially with the loading 239 00:21:53,079 --> 00:21:59,759 example, because there you will have to do like even more work because you 240 00:21:59,839 --> 00:22:06,839 are inside a signal state, and like you start an HTP goal for instance, 241 00:22:06,960 --> 00:22:10,799 you have to go like patch state loading through something like that, and 242 00:22:10,880 --> 00:22:14,319 then you will have to you will always have to add this line of code 243 00:22:14,359 --> 00:22:22,720 and the other line that's something like finalized patch state loading falls. And it's 244 00:22:22,759 --> 00:22:27,759 not a pattern that I personally like. And you can't sort of make a 245 00:22:27,920 --> 00:22:34,279 generic function that doesn't because of the loading state can be different depending on whatever 246 00:22:34,480 --> 00:22:38,759 you are trying to accomplish, unless it resorts some magic strings or something. 247 00:22:38,839 --> 00:22:42,519 But that's not a good solution to that. But I mean, yeah, 248 00:22:42,799 --> 00:22:51,079 we've had promises since twenty fifteen and nine years later, like how come we 249 00:22:51,200 --> 00:22:56,079 don't have like an abstraction that just automatically knows, Like how come we're still 250 00:22:56,119 --> 00:23:00,839 writing manual loading states, Like why don't we have an abstraction that knows whether 251 00:23:00,920 --> 00:23:07,640 it's executing or not and is cancelable if the endpoint of course returns an observable 252 00:23:07,720 --> 00:23:12,359 Because promises are not so simple to cancel as observables. And that's one of 253 00:23:12,400 --> 00:23:18,880 the greatest things about observables, and I'm super geeked on observables. They're coming 254 00:23:18,960 --> 00:23:25,720 into the language, into the browsers are starting to integrate it slightly different API 255 00:23:26,000 --> 00:23:29,519 than r XJS observables. I'm still learning about, like how am I going 256 00:23:29,599 --> 00:23:36,599 to have to adjust my code when when those come into play? But but 257 00:23:36,720 --> 00:23:41,599 it's it's I mean, our XJS is a tremendous library. If I could 258 00:23:41,640 --> 00:23:48,640 give your listeners a little history lesson of ASYNC, I did a little research 259 00:23:48,799 --> 00:23:59,640 to find out like how rx JS and promises evolved and and could I I 260 00:23:59,680 --> 00:24:04,319 could do that an hour later. I mean, I mean we sort of 261 00:24:04,680 --> 00:24:07,799 can. Yeah, but it's intastic. I've never thought about it this way 262 00:24:07,880 --> 00:24:12,920 because I've never really worked with promises. I have worked with pies since I 263 00:24:14,559 --> 00:24:18,880 since I started doing while I worked with promises when I was doing Angular jes, 264 00:24:19,279 --> 00:24:23,480 we didn't have like we don't know much about urges back then then Angular 265 00:24:23,599 --> 00:24:29,680 two drops in twenty sixteen, and yeah, since then, I've not really 266 00:24:29,920 --> 00:24:34,880 touched promises that much unless I was I don't know, teaching someone about them 267 00:24:36,240 --> 00:24:38,559 or stuff. I was really using them in the cold. But it never 268 00:24:40,440 --> 00:24:45,319 like I just thought, you know, promises are sort of weak and observables 269 00:24:45,319 --> 00:24:48,279 are very powerful, and I can do HTP cols with observable that yeah, 270 00:24:48,279 --> 00:24:52,880 I don't care about promises. No, I mean that's that's actually very true, 271 00:24:52,920 --> 00:24:57,759 and I don't think it's judgmental or critical at all. Promises you could 272 00:24:57,799 --> 00:25:03,680 say, are a subset of observables, definitely right, and so you can 273 00:25:03,759 --> 00:25:07,640 build the promise with an observable, but the reverse is not true. And 274 00:25:10,319 --> 00:25:15,079 so yeah, so let me talk about mental models, because my mental model 275 00:25:15,160 --> 00:25:19,039 has been evolving a lot over time, and I think I can clarify for 276 00:25:19,160 --> 00:25:26,599 the listeners a lot of things that you hear about what observables are, because 277 00:25:26,799 --> 00:25:32,720 that's even kind of, you know, confusing. So observables have been described 278 00:25:32,839 --> 00:25:37,599 any number of ways. They've been described as lazy. They've been called just 279 00:25:37,839 --> 00:25:44,920 functions and observables just a function, which is the hardest explanation I have to 280 00:25:45,160 --> 00:25:51,039 ask. They've been described as streams of values, or they've been described as 281 00:25:51,240 --> 00:25:56,599 promises with more than one value, so a lot of those are confused. 282 00:25:56,599 --> 00:26:02,880 Oh. They've also been described as push collections. They've been described as is 283 00:26:02,920 --> 00:26:07,920 it unicast or multicast? Well, it's unicast, But how do I tell? 284 00:26:08,400 --> 00:26:12,039 It's very hard to piece together, like what the heck is an observable 285 00:26:12,519 --> 00:26:15,640 with all of these explanations. So I have what I think is going to 286 00:26:15,680 --> 00:26:23,319 be the final explanation of observable that people will need to hear. And here's 287 00:26:23,400 --> 00:26:30,599 this. Observables are a template for a running process. They are like the 288 00:26:30,759 --> 00:26:37,359 string ls dash l. Okay, so they're like the string ls dash l 289 00:26:37,519 --> 00:26:42,000 because the string ls dash l is not executing. It's just a template from 290 00:26:42,039 --> 00:26:47,039 which you can create a process that lists the files in a directory. Right, 291 00:26:48,279 --> 00:26:56,359 So observables aren't really lazy anymore than a template is lazy because when you 292 00:26:56,559 --> 00:27:00,640 click on the template to create a new document, it creates new document. 293 00:27:00,759 --> 00:27:03,599 When you have a string, a command string to run a process, it 294 00:27:03,680 --> 00:27:07,440 runs a new process. So yes, you can call them lazy because they 295 00:27:07,480 --> 00:27:11,519 don't give you values until you subscribe. But that's exactly like the way that 296 00:27:11,599 --> 00:27:18,480 a Unix command line string doesn't give you lines of output until you execute it 297 00:27:18,519 --> 00:27:23,680 and hit enter. So you will get what can a Unix process do? 298 00:27:25,279 --> 00:27:29,119 And this is what's so strong about observables. They can do everything that a 299 00:27:29,240 --> 00:27:33,480 Unix process can do. They can have an exit code of true or false. 300 00:27:33,519 --> 00:27:37,599 I mean basically, they can complete or error like an observable, and 301 00:27:37,720 --> 00:27:44,559 they can produce one or more values. They can also produce zero values, 302 00:27:44,640 --> 00:27:48,599 which is why the description of an observable as a stream of values it misses 303 00:27:48,720 --> 00:27:56,359 the mark. An observable is a template for a running process. A running 304 00:27:56,400 --> 00:28:02,200 process, which is a subscription, consumes some resources. So the main thing 305 00:28:02,240 --> 00:28:07,680 about an observable is that it explains how to consume and release some resources, 306 00:28:07,400 --> 00:28:11,640 just like a command line string tells you how to consume some resources, and 307 00:28:11,720 --> 00:28:15,640 then it has a standard way of canceling it. So it's perfectly valid to 308 00:28:15,799 --> 00:28:22,240 have a Unix command line like r M r F node modules, which doesn't 309 00:28:22,279 --> 00:28:26,319 produce a bunch of values, and it's perfectly valid to have an observable that 310 00:28:26,480 --> 00:28:33,680 isn't about values, it's about getting some work done. That's a that's an 311 00:28:33,720 --> 00:28:38,200 interesting way of looking at it. Actually, I've never thought of that sort 312 00:28:38,200 --> 00:28:42,440 of comparison. And it makes sense especially that we have like the well, 313 00:28:44,200 --> 00:28:48,799 like the pipe operator rather pipe commands, pipe results throw and yeah, yes, 314 00:28:48,920 --> 00:28:52,720 yeah, we sort of have the same thing. We can take observable. 315 00:28:53,240 --> 00:28:56,920 Yeah, yeah, yeah, that's really interesting way looking at it. 316 00:28:57,759 --> 00:29:02,119 Yeah, it's it's just so oh, it's just such a perfect match. 317 00:29:02,440 --> 00:29:07,519 It's a template, it's a command line string. That's like a perfect match 318 00:29:07,599 --> 00:29:12,680 for me. And so what I've found is that the best use of an 319 00:29:12,720 --> 00:29:18,319 observable, which is of course a description of some work that can be done, 320 00:29:18,480 --> 00:29:21,519 a template for doing work, the best use of an observable, and 321 00:29:21,680 --> 00:29:26,400 pretty much the only thing you need to do in RX effects is use an 322 00:29:26,440 --> 00:29:32,119 observable as a return value. So that's kind of talking about old observables, 323 00:29:33,519 --> 00:29:37,079 but it could also be a hot I don't really find there's much distinction between 324 00:29:37,319 --> 00:29:42,000 hot and cold. There's still a allocation of a resource that you can release 325 00:29:42,559 --> 00:29:48,440 either way, and they may produce values or they may not. But so 326 00:29:48,920 --> 00:29:52,000 observable is a great return value. And in our X effects, when you 327 00:29:52,079 --> 00:29:56,960 return an observable from a function into an effect or a bus listener or a 328 00:29:56,079 --> 00:30:02,240 service, you don't have to call dot subscribe on it, right And it 329 00:30:02,400 --> 00:30:06,519 works because if you're ringing a bell and you have an observable for ringing a 330 00:30:06,599 --> 00:30:11,599 bell, when the bell is wrong, you return the observable for ringing the 331 00:30:11,680 --> 00:30:15,240 bell. Well, RX effects knows that if you return an observable, well, 332 00:30:15,279 --> 00:30:21,279 of course I have to subscribe to it if the concurrency model says subscribe 333 00:30:21,359 --> 00:30:26,200 right now, because there are concurrency models that don't say subscribe right away, 334 00:30:26,440 --> 00:30:33,480 for example a blocking concurrency model also known as exhaust map. So observables make 335 00:30:33,599 --> 00:30:37,799 a lot of sense as a return value. And I've written very large and 336 00:30:38,039 --> 00:30:45,519 very async applications with RX effects using observables, and I never call dot subscribe, 337 00:30:47,160 --> 00:30:51,920 just return an observable. So, for example, my flagship app for 338 00:30:52,119 --> 00:30:59,319 personal use that I can share around is called Inbox Radio, and it lets 339 00:30:59,400 --> 00:31:04,920 me dream music that has been sent to me in my Gmail inbox. And 340 00:31:06,079 --> 00:31:10,079 so there's a lot of async going on there. You have to get messages 341 00:31:10,160 --> 00:31:12,960 that match a pattern, then you have to get the attachments, and you 342 00:31:14,039 --> 00:31:18,279 don't want to download them in parallel. You want to download them serially because 343 00:31:18,279 --> 00:31:22,279 you're going to play them serially, and then you have to decode the stuff 344 00:31:22,359 --> 00:31:26,519 then you have to play it. And so uh, these these chains of 345 00:31:26,599 --> 00:31:30,480 things are not explicitly chained. The bus, the bus, the event bus 346 00:31:30,000 --> 00:31:34,559 is sitting there, you know, making one of them subscribe to another. 347 00:31:36,400 --> 00:31:42,079 And basically this whole thing doesn't have a can haat map explicitly written in it. 348 00:31:44,400 --> 00:31:51,799 It has create queueing service, and it doesn't have dot subscribe because you 349 00:31:52,200 --> 00:31:56,440 return an observable and RXFFX call subscribe on it. So that kind of takes 350 00:31:56,519 --> 00:32:00,119 care of where do you put your subscription objects. Well, you don't generally 351 00:32:00,359 --> 00:32:07,000 need to see the subscription objects in this. The exception to that would be 352 00:32:07,559 --> 00:32:12,720 if you had an observable of button clicks, for example, and you needed 353 00:32:12,799 --> 00:32:19,519 to like call an effect on every button click from that observable, then you 354 00:32:19,559 --> 00:32:24,920 would subscribe to that outer observable and say, okay, each time a notification 355 00:32:25,039 --> 00:32:29,920 on the observable comes in, then call this effect. So that's where you 356 00:32:30,000 --> 00:32:34,160 would if you had like a native r XJS observable, you could totally call 357 00:32:34,279 --> 00:32:37,519 dot subscribe on it and do whatever you want. But if you were returning 358 00:32:37,599 --> 00:32:43,640 it as the return value from an effect or a bus listener, you wouldn't 359 00:32:43,680 --> 00:32:51,640 need to tag dot subscribe on the end of it. That's fascinating Actually, 360 00:32:52,440 --> 00:32:58,200 as you mentioned exhaust map and stuff, so it's nice to hear that, 361 00:32:59,400 --> 00:33:02,960 uh, those sources of things are taking care of about because for the life 362 00:33:04,079 --> 00:33:07,440 of me, I couldn't explain the difference between exhaust map and concat map. 363 00:33:07,519 --> 00:33:10,319 I've looked it out like a bunch of times. I can easily explain difference 364 00:33:10,359 --> 00:33:15,599 between Merge map and swishmap. Okay, so here's here's yeah, so here's 365 00:33:15,599 --> 00:33:20,480 a here's a gift. This is for everybody. It's a free uh way 366 00:33:20,640 --> 00:33:22,640 to keep all these things straight? Can you? Can you scroll down? 367 00:33:23,039 --> 00:33:29,720 Uh in the RX effects repo yep, until you see effects one. Yeah. 368 00:33:29,759 --> 00:33:32,440 I wasn't a service one because I think either one of them. So 369 00:33:32,559 --> 00:33:38,400 do you see some cards that you have pictures on them? Oh? Jez. 370 00:33:38,559 --> 00:33:43,400 By the way, if you're comfortable with that, we could even share 371 00:33:43,440 --> 00:33:46,079 your screens. Some people can be watching from YouTube, so they could be 372 00:33:46,160 --> 00:33:51,759 seeing the I would. I would happily do that. Sure, let's do 373 00:33:51,920 --> 00:33:58,000 that. Okay, I think you're seeing GitHub repo. Yeah, okay, 374 00:33:58,079 --> 00:34:02,559 So this is the RX effects main repo. And if you scroll down, 375 00:34:04,039 --> 00:34:07,159 there's some code examples, and then there's a list of the libraries. Yeah, 376 00:34:08,000 --> 00:34:13,039 the one. There's the fascinating set of cards. Yes, and so 377 00:34:13,239 --> 00:34:17,039 here here is the Here is the set of cards that you can use. 378 00:34:17,719 --> 00:34:23,559 I have not printed them myself yet on like true quality card stock, but 379 00:34:23,679 --> 00:34:27,440 if people are interested in that, I would I would do that. I'd 380 00:34:27,480 --> 00:34:30,679 love to have these at my desk. Well, when we collaborated with desks 381 00:34:30,800 --> 00:34:37,239 next to desks and other people. But yeah, so this gives friendly names 382 00:34:37,760 --> 00:34:40,639 to each of the each of the modes, it has pictures, and then 383 00:34:40,920 --> 00:34:47,119 in the little in this window it says also known as merge map, also 384 00:34:47,239 --> 00:34:52,079 known as cat map. So you know, my bias is to kind of 385 00:34:52,480 --> 00:34:59,280 tuck rx JS into something as a bunch of implementation details. I'm not trying 386 00:34:59,320 --> 00:35:05,599 to teach people how to, you know, necessarily use our XJS better. 387 00:35:06,119 --> 00:35:13,679 If they can get the first ninety percent of use cases under friendlier names and 388 00:35:14,039 --> 00:35:19,320 in a friendlier library, then that's kind of what I'm advocating for. I've 389 00:35:19,360 --> 00:35:24,480 gotten several teams that I'm on to adopt this early in the development life cycle, 390 00:35:25,000 --> 00:35:30,960 and then when people can use one of these, they use that, 391 00:35:30,280 --> 00:35:35,880 and then it comes up against something that doesn't fit into that paradigm. Well 392 00:35:35,920 --> 00:35:38,000 then you know, hey, this this lets you use any amount of r 393 00:35:38,119 --> 00:35:42,639 XJS on top of it. But yeah, I would say that this is 394 00:35:44,159 --> 00:35:50,280 mnemonic and friendly for people to remember the modes and just have good pictures of 395 00:35:50,360 --> 00:35:53,440 it, because these modes. This is what's amazing about what they came up 396 00:35:53,519 --> 00:35:57,480 with our x When they came up with our XJS and I'm going to go 397 00:35:57,639 --> 00:36:05,679 rewind a little bit in history up sharing though. So in two thousand and 398 00:36:05,760 --> 00:36:12,159 four, Gmail came out and ajax started to get used, like in seriousness. 399 00:36:13,079 --> 00:36:16,480 And as AJAX was getting used between two thousand and four and twenty eleven, 400 00:36:16,920 --> 00:36:23,039 call back help was like really forming. And you know when you attach 401 00:36:23,079 --> 00:36:30,360 an event listener to something, if that event listener starts some ASYNC process and 402 00:36:30,440 --> 00:36:32,760 then that event comes in again, you know, you don't have concurrency control. 403 00:36:34,159 --> 00:36:39,280 So seven years after like ajax appeared, Microsoft Labs our research and development 404 00:36:39,840 --> 00:36:43,599 went down and said we got to solve this ACYNC problem and we got to 405 00:36:43,639 --> 00:36:47,679 solve it for good and hit all of the use cases. And this is 406 00:36:47,719 --> 00:36:53,880 four years before promises, and they came up with the Reactive x you know 407 00:36:54,000 --> 00:36:59,039 project, which you know became our XJS and all the implementations in all the 408 00:36:59,079 --> 00:37:04,800 languages. So they had you know, PhD computer science people and everything really 409 00:37:04,840 --> 00:37:07,079 looking at this async problem and saying like, what are all the tools we 410 00:37:07,159 --> 00:37:13,199 need to solve every use case in async? And they built it. But 411 00:37:13,920 --> 00:37:19,039 by two fift by twenty twelve, the A plus team was working on promises 412 00:37:19,199 --> 00:37:23,559 and promises were gaining traction. I think partly they were beating winning the race 413 00:37:23,639 --> 00:37:31,400 because the documentation for our XJS was just really problematic. They you know, 414 00:37:31,679 --> 00:37:42,719 like the documentation for gosh, for merge map, it's it's obtuse. It 415 00:37:43,239 --> 00:37:49,719 projects inner observables into the output observable, and it's just it's just not descriptive 416 00:37:49,840 --> 00:37:57,119 enough. So promises from twenty eleven to twenty fifteen were gaining mind share and 417 00:37:57,280 --> 00:38:00,039 they were an improvement over just regular function, but they didn't get all the 418 00:38:00,079 --> 00:38:08,000 way to RxJS. And then promises were adopted. But promises can't control concurrency 419 00:38:08,400 --> 00:38:14,360 in all the ways that ox js can. And so those concurrency operators that 420 00:38:14,480 --> 00:38:17,400 are so hard to remember but described in those cards that I showed, those 421 00:38:17,480 --> 00:38:24,440 things are amazing because they solved for all async combinations and I can I can 422 00:38:24,480 --> 00:38:30,000 describe this in real life too. It isn't just RxJS. They didn't just 423 00:38:30,079 --> 00:38:37,800 solve like for web development. It's like extrapolatable. It can be extrapolated to 424 00:38:37,920 --> 00:38:40,960 real life. So I gave the analogy on the React show. Suppose I 425 00:38:42,039 --> 00:38:45,960 am making a sandwich, or my daughter she wants a grilled cheese sandwich. 426 00:38:46,000 --> 00:38:49,960 I'm on the stove. I'm busy. I've been requested to make a sandwich. 427 00:38:49,960 --> 00:38:53,039 I'm busy making the sandwich, and she says, Dad, I want 428 00:38:53,119 --> 00:39:00,320 chicken nuggets instead. Well, these five concurrency modes well for them, or 429 00:39:00,360 --> 00:39:04,679 in r x JAS, I added the fifth one. These five concurrency modes 430 00:39:04,800 --> 00:39:09,559 describe what I can respond with. Right, I can merge MAP, So, 431 00:39:09,679 --> 00:39:14,360 in other words, I can just start a new pan cooking the chicken 432 00:39:14,440 --> 00:39:19,840 nuggets next to the grilled cheese. I can say, okay, Sierra, 433 00:39:20,119 --> 00:39:22,639 but only after I make the grilled cheese. I'll get started on the chicken 434 00:39:22,679 --> 00:39:27,840 nuggets. That's cancat MAP. I can say, oh, you want chicken 435 00:39:27,920 --> 00:39:30,599 nuggets, Well, I better stop making the grilled cheese. I'll make you 436 00:39:30,719 --> 00:39:37,840 chicken nuggets instead. And I can also say no, Sierra, I'm already 437 00:39:37,880 --> 00:39:40,960 making your grilled cheese. I'm not going to start chicken nuggets. So that 438 00:39:42,320 --> 00:39:47,039 happens to be exhaust MAP. Finally, the final one, which I just 439 00:39:47,199 --> 00:39:52,199 came up with by looking at what are all the possibilities. The final one 440 00:39:52,320 --> 00:39:54,760 is I'm making the grilled cheese. Sierra says, can I have chicken nuggets? 441 00:39:55,239 --> 00:39:59,119 I just turn off the stove for the grilled cheese, and I say, 442 00:40:00,000 --> 00:40:01,679 I can't deal with all these requests. Just make your own lunch, 443 00:40:02,039 --> 00:40:07,000 so I can walk away from that. These are universal, right, and 444 00:40:07,519 --> 00:40:15,480 so it's really cool that RxJS has all of the universal possibilities embedded in it. 445 00:40:15,880 --> 00:40:20,880 It's not cool that everyone isn't using it. And I think the reason 446 00:40:21,000 --> 00:40:28,639 is because the documentation isn't friendly enough. And so instead of trying to improve 447 00:40:28,719 --> 00:40:36,679 the documentation and doing you know that kind of long but noble process of trying 448 00:40:36,760 --> 00:40:39,880 to contribute back to an open source repository, you use. But I just 449 00:40:39,960 --> 00:40:44,360 think, well, it's time for a higher level of abstraction. And I 450 00:40:44,360 --> 00:40:47,320 don't think I'm on the wrong track because, like I ask for comparisons, 451 00:40:47,360 --> 00:40:52,360 I say, how would you do this the RXGS way? And it's fifteen 452 00:40:52,400 --> 00:40:57,920 to twenty confusing lines with you know, seven levels of nesting versus you know, 453 00:40:59,000 --> 00:41:01,679 one or two levels nesting and like five lines. So I want to 454 00:41:01,719 --> 00:41:07,880 stick with this approach of creating a rapper library and hopefully you know, hey, 455 00:41:07,960 --> 00:41:10,880 if any of this ends up back in r x j S, if 456 00:41:10,920 --> 00:41:15,320 it ends up in react, you know, I don't care, Like I'm 457 00:41:15,360 --> 00:41:19,679 not I'm not saying everybody has to use this exactly this way, but I 458 00:41:19,760 --> 00:41:24,000 am saying I think there is an abstraction just waiting to be uh, you 459 00:41:24,079 --> 00:41:30,280 know, birthed into the community. This truly has been an evening of interesting 460 00:41:30,360 --> 00:41:38,480 analogies. Thanks for thanks for explaining me the difference between I find that I 461 00:41:38,559 --> 00:41:46,199 found it all Libray really fascinating and be surprised I haven't heard about it before. 462 00:41:46,719 --> 00:41:49,880 I will. I will research it, you know, if it's okay, 463 00:41:49,920 --> 00:41:52,719 if I PINU you with questions from this time. Yeah, you know, 464 00:41:52,840 --> 00:41:55,719 the best place to engage with me right now is uh is through the 465 00:41:55,760 --> 00:42:04,559 GitHub repo. Stars are welcome discussions and issues also, and uh you know, 466 00:42:04,760 --> 00:42:07,679 that's that's where we can talk about the repo. If we want to 467 00:42:07,719 --> 00:42:14,840 talk a little more generally about stuff, I'm on Twitter at slash x at 468 00:42:15,239 --> 00:42:20,360 Dean dev Dad and that'll go in the show notes. And yeah, I 469 00:42:20,400 --> 00:42:23,079 mean I would love to to engage on this more. Like I said, 470 00:42:23,079 --> 00:42:30,119 I have my workplace where I've gotten now three teams you know, using this 471 00:42:30,480 --> 00:42:35,880 as you know, a starting point, and so that's probably where you know, 472 00:42:36,360 --> 00:42:39,199 more of my time gets spent in the workplace than marketing it. But 473 00:42:39,519 --> 00:42:46,000 since I've reconnected with this awesome top end dev's team and and all the shows 474 00:42:46,039 --> 00:42:49,480 that they run. I was like, well, sure, I'd love to 475 00:42:49,880 --> 00:42:58,480 talk about it more widely. That's awesome. Yeah, and for anyone who's 476 00:42:58,559 --> 00:43:02,480 listening to this and once to try out our ax effects and other frameworks, 477 00:43:02,960 --> 00:43:07,159 we do have an entire episode about how this can be used and React, 478 00:43:07,639 --> 00:43:12,480 so if you're interested, you can check out the latest episode we did on 479 00:43:12,679 --> 00:43:15,840 React Round Up, where we also talk to Dean about how we can use 480 00:43:15,960 --> 00:43:22,800 our ex effects to simplify some of the issues that React developers generally run into. 481 00:43:22,679 --> 00:43:30,840 So yeah, this again, this has been great again because the React 482 00:43:30,920 --> 00:43:36,920 episode was also awesome. I really love the the analogis to your daughter, 483 00:43:37,360 --> 00:43:45,119 and it just it makes it so much easier to remember what exhaust map means 484 00:43:45,239 --> 00:43:57,559 because now I can just remember being exhausted, like I'm so thanks, thanks 485 00:43:57,599 --> 00:44:01,440 again, man. I hope you come back even even more times whenever possible. 486 00:44:04,480 --> 00:44:07,840 So let's do a quick round of promos before wrapping things up. So 487 00:44:08,480 --> 00:44:14,599 you're mentioning your Twitter, the GitHub repository for the project. Is there anything 488 00:44:14,679 --> 00:44:19,039 else you would like to promote? Just that? Just that? Okay, 489 00:44:19,360 --> 00:44:22,559 okay, awesome. So yeah, just reiterating anyone who's listening to this, 490 00:44:23,159 --> 00:44:28,280 if you can't stop what you're doing for just ten seconds, then go into 491 00:44:28,360 --> 00:44:34,320 the GitHub repository and give it a star that could create a chain reaction that 492 00:44:34,559 --> 00:44:38,840 is going to lead to eventually us having better names to our rx jazz operator. 493 00:44:39,000 --> 00:44:45,400 So please do that for the entire community and future generations. Well, 494 00:44:45,519 --> 00:44:51,800 I should mention there is a thread a discussion in the RxJS repo that opened 495 00:44:51,840 --> 00:44:59,519 up recently where they're seeking to do some renaming in RxJS eight. So you 496 00:44:59,599 --> 00:45:05,320 can find that Ben Lesh's profile on Twitter, and so if you have opinions 497 00:45:05,320 --> 00:45:07,880 about it, like I do, you can you can contribute in their discussion 498 00:45:07,960 --> 00:45:13,519 directly. I'm sure they'd appreciate the community and put awesome Rman. I'm gonna 499 00:45:13,960 --> 00:45:15,599 give this to you in a second. I know that you have a special 500 00:45:15,639 --> 00:45:19,719 announcement, so I'm going to leave you to the end. Mine is just 501 00:45:19,800 --> 00:45:28,760 going to be the regular episode interest. So if you're interested in outsourcing your 502 00:45:28,880 --> 00:45:34,119 software or design teams, or even better, extending your design and software development 503 00:45:34,199 --> 00:45:37,800 teams, you should definitely check out Onvoid dot com is U n v o 504 00:45:37,920 --> 00:45:43,719 I D dot com and yeah, that's basically gonna be my promo. So 505 00:45:44,320 --> 00:45:49,679 arman, please take away. I know you have a huge announcement. Well 506 00:45:49,760 --> 00:45:52,519 it's not that huge. The huge announcement like a month ago, as I 507 00:45:52,639 --> 00:45:58,360 mentioned on the previous episodes and also on Twitter and LinkedIn and all the other 508 00:45:58,440 --> 00:46:00,840 places I visit. I don't know. I wrote a book on Angular. 509 00:46:01,840 --> 00:46:07,559 It's called Modern Angular. It's with the many publications. It's in early access 510 00:46:07,760 --> 00:46:15,559 now. So out of the ten total chapters, we have now five out. 511 00:46:15,440 --> 00:46:21,760 I think two other chapters will drop like in a week or two, 512 00:46:22,079 --> 00:46:27,199 I guess. But the entire manuscript is finished, so no worries if you 513 00:46:27,239 --> 00:46:29,800 think I'm not going to finish it. I finished it. I'm just doing 514 00:46:29,960 --> 00:46:35,760 some polishing and listening to early access leaders comments. And by the way, 515 00:46:35,880 --> 00:46:38,480 the fifth chapter the last month, that is right now available is about our 516 00:46:38,519 --> 00:46:45,719 issues. So yeah, if this episode got you interested, maybe you can 517 00:46:45,760 --> 00:46:50,920 also take a look at that chapter. The book discusses all the new features 518 00:46:51,039 --> 00:46:54,480 that we have. By the way, the sixth and seventh chapter that we 519 00:46:54,719 --> 00:47:00,480 drop soon are all about signals, which is the idea hottest topic right now. 520 00:47:01,519 --> 00:47:08,559 And for this episode, I actually got a special promo code from Manning 521 00:47:08,719 --> 00:47:15,519 Publications, so you can use the promo code pod Angler twenty four. Let's 522 00:47:15,559 --> 00:47:22,800 also put that in the description to get forty five percent of any Manning product. 523 00:47:22,719 --> 00:47:29,000 I hope you spent it on my book, but yeah so, and 524 00:47:29,119 --> 00:47:35,079 also yeah, thanks to many Publications for giving me that promo code specifically for 525 00:47:35,320 --> 00:47:40,320 our episodes. So if anyone of the listeners is interesting, take a look 526 00:47:40,320 --> 00:47:45,760 at the description. I will. Let's also put the link to the book 527 00:47:45,840 --> 00:47:50,960 there. So yeah, that's it. That's my promotion. Gonna be the 528 00:47:51,000 --> 00:47:55,079 same for a while. Yeah, Aaran, can you send it on the 529 00:47:55,280 --> 00:48:00,159 internal chat so that I can put it in the comments section for anyone in 530 00:48:00,239 --> 00:48:05,239 the livestream. Yeah, forty five percent discount? Is that what you said? 531 00:48:06,119 --> 00:48:12,239 Yep, ye, that's a huge, huge discount, so awesome, 532 00:48:12,480 --> 00:48:16,000 awesome, thanks Man. And I think the fact that you have two chapters 533 00:48:16,079 --> 00:48:25,840 about signals in angular is really interesting because it's uncharted territory, right like in 534 00:48:25,960 --> 00:48:30,440 terms of conventions and things like that. So I think this is the kind 535 00:48:30,480 --> 00:48:36,559 of thing that is relevant even for those that are already experienced, like you 536 00:48:36,639 --> 00:48:39,480 might be thinking, I don't need Arman's book, like I'm already experienced in 537 00:48:39,519 --> 00:48:44,679 Angler, there's no need for that. And even Signals. Yeah, like 538 00:48:44,719 --> 00:48:47,679 I'm comfortable with Signals. I know that they were released with whatever, but 539 00:48:50,559 --> 00:48:53,800 I'm pretty sure there's something that you can learn from those two chapters specific to 540 00:48:53,920 --> 00:49:00,199 Signals. So I would still highly recommend buying Arman's book, even if you 541 00:49:00,239 --> 00:49:05,239 feel extremely comfortable with Angler already. So yeah, coupon is in the show 542 00:49:05,280 --> 00:49:09,280 notes. Again, it's pod pod Angular twenty four altogether, no spaces, 543 00:49:09,360 --> 00:49:15,760 no dashes. All right, thanks thanks Man for having coo cool book. 544 00:49:15,920 --> 00:49:20,880 Just check it out. Thank you. All right, So, guys, 545 00:49:21,000 --> 00:49:24,960 thank you so much. Also putting the link to Arman's book in the comment 546 00:49:25,039 --> 00:49:29,719 section, so you have the coupon and the book link. There no reason 547 00:49:29,800 --> 00:49:34,239 to not not buy it. But yeah, again, guys, thank you 548 00:49:34,360 --> 00:49:37,119 so much. Dean. Thanks for being on the show, for not just 549 00:49:37,559 --> 00:49:44,199 all the time that you spent in here, but also the effort to build 550 00:49:44,679 --> 00:49:51,679 this amazing library that is available for everyone in open source. So that's really 551 00:49:52,199 --> 00:49:55,159 that's really a major effort. Thank you for that. And yeah, at 552 00:49:55,280 --> 00:50:00,440 least that's it for me. Yeah, thanks everyone. You're in the next 553 00:50:00,519 --> 00:50:02,880 one here in the next year.