Modern JavaScript Testing: Integration, Contract Testing & AI Tools - JSJ 692 の全文書き起こし

of testing AI trends, 78 00:04:05.280 --> 00:04:08.960 mostly the useful parts of AI. So yeah, okay, just 79 00:04:09.000 --> 00:04:09.639 to just to. 80 00:04:11.520 --> 00:04:15.639 Finish that particular point. The scary thing is afterwards, when 81 00:04:15.680 --> 00:04:18.720 you make changes and then the tests break, you tell it, 82 00:04:19.319 --> 00:04:22.199 Cursor fix the test for me, and you're not sure 83 00:04:22.240 --> 00:04:25.920 whether it's fixing your test to hide the bugs or 84 00:04:25.959 --> 00:04:29.120 fixing your code to fix the tests, or doing both 85 00:04:29.360 --> 00:04:33.199 or whatever it's It can be really interesting. 86 00:04:33.279 --> 00:04:36.959 That's this way, definitely. I mean, I think that one 87 00:04:36.959 --> 00:04:39.160 of the first things one should tune in Cursor or 88 00:04:39.199 --> 00:04:41.759 cloud code or whatever you're working is that you can 89 00:04:41.920 --> 00:04:44.040 let it work in yellow mode. But I think there 90 00:04:44.079 --> 00:04:48.759 is one exception any a human A human confirmation should 91 00:04:48.759 --> 00:04:51.759 be made anytime a test is changed. You may change 92 00:04:51.759 --> 00:04:54.399 my production code, I may inspect each and every nine 93 00:04:54.480 --> 00:04:57.000 or not, I don't know, but a test changed should 94 00:04:57.040 --> 00:04:58.839 supervise that that's. 95 00:04:58.720 --> 00:05:01.040 My That's a good take. I think. 96 00:05:01.839 --> 00:05:03.279 I can't think of any time I'm just going to 97 00:05:03.319 --> 00:05:05.720 say here writing some code and then committed or do 98 00:05:05.839 --> 00:05:08.839 whatever without looking over it and making sure it works. 99 00:05:09.279 --> 00:05:09.560 Well. 100 00:05:10.000 --> 00:05:14.600 Yeah, that's easy to say, and everybody starts with good intentions, 101 00:05:14.639 --> 00:05:18.720 but at a certain point in time, you know, especially 102 00:05:18.720 --> 00:05:21.079 when you're under the gun and certain things and things 103 00:05:22.680 --> 00:05:25.240 seem to work and you build a certain level of trust. 104 00:05:26.160 --> 00:05:30.480 I've known people to almost even engineers, almost get to 105 00:05:30.519 --> 00:05:33.759 the point of almost vibe coding their way through projects. 106 00:05:34.360 --> 00:05:40.240 And I think you Only's statement that if you've got 107 00:05:40.319 --> 00:05:46.199 good test coverage, then ensuring that the tests are properly 108 00:05:46.439 --> 00:05:51.360 maintained and reviewed is a good way to add a 109 00:05:51.399 --> 00:05:55.639 significant layer of trust and security around your vibe coding. 110 00:05:59.720 --> 00:06:03.079 Yeah, I agree with both. I think that people have 111 00:06:03.360 --> 00:06:07.680 various degrees of how much they supervise AI generated God, 112 00:06:09.199 --> 00:06:12.600 I'm just saying that the least conservative one can be 113 00:06:12.800 --> 00:06:15.000 is at least guard you're testing. You might do this 114 00:06:15.079 --> 00:06:19.319 also for the production cod but I think that letting 115 00:06:19.360 --> 00:06:22.040 AI also change tests has to change God. This is 116 00:06:22.040 --> 00:06:26.240 the highest amount of frizk one may take. 117 00:06:27.839 --> 00:06:32.199 Yes, Okay, then, so what are the latest and greatest 118 00:06:32.240 --> 00:06:35.839 trends when it comes to tests, aside from having AI 119 00:06:36.120 --> 00:06:38.000 do all the coding for us? 120 00:06:39.279 --> 00:06:44.000 Yeah, so I'll kind of mention some kind of super 121 00:06:44.120 --> 00:06:48.040 trends and are very impactful changes, and then we can 122 00:06:48.879 --> 00:06:52.040 dig dive deeper into each one of them, at least 123 00:06:52.079 --> 00:06:54.920 the ones that we found interesting. So I think there 124 00:06:54.959 --> 00:06:58.920 is always in testing one kind of everlasting progress in 125 00:06:58.959 --> 00:07:02.800 which we always we try to get our tests to 126 00:07:02.839 --> 00:07:07.360 be more production like without paying the price of So 127 00:07:08.759 --> 00:07:12.519 in the past two years we keep seeing this progress 128 00:07:13.120 --> 00:07:15.600 and in this in front and the front and world. 129 00:07:15.920 --> 00:07:18.079 For the first time, I think like we ad React 130 00:07:18.120 --> 00:07:20.720 testing library ruling this world for a lot of time 131 00:07:21.120 --> 00:07:23.360 tests that runs in a jazz things that's for the 132 00:07:23.399 --> 00:07:27.639 first time it's safe to say that tests don't run 133 00:07:27.680 --> 00:07:32.839 in the browser, that are not visual are almost out 134 00:07:32.879 --> 00:07:37.279 of scope. There are many new ways and frameworks to 135 00:07:37.399 --> 00:07:39.759 run your tests in the browser and get the full 136 00:07:39.920 --> 00:07:44.839 visual experience and the confidence, so it's more like production. 137 00:07:44.920 --> 00:07:47.600 But some of them run really fast, so you're not 138 00:07:47.720 --> 00:07:51.199 paying the price of like full full blown end to 139 00:07:51.360 --> 00:07:54.600 end framework on the back inside it means that you 140 00:07:54.600 --> 00:07:56.720 having a micro service, you can more easily now run 141 00:07:56.759 --> 00:08:00.000 it on your computer with all the layers and almost 142 00:08:00.079 --> 00:08:04.560 light production infrastructure, and again without paying a tremendous price. 143 00:08:05.360 --> 00:08:09.920 I think these are bold and the changes that we're seeing, 144 00:08:09.920 --> 00:08:12.279 and we I guess we will dig deeper into the 145 00:08:12.319 --> 00:08:13.600 specific tooling. 146 00:08:13.959 --> 00:08:19.720 And just to clarify, there are obviously layers to tests, 147 00:08:19.800 --> 00:08:23.720 so there I don't know if you should refer to 148 00:08:23.759 --> 00:08:27.480 them as higher or lower layers. It's just different positions 149 00:08:27.480 -->

00:08:30.759 in the testing cycle. So you've got obviously stuff like 150 00:08:30.920 --> 00:08:36.320 unit tests, which should be very relatively small, relatively straightforward, 151 00:08:36.360 --> 00:08:40.440 and certainly very very fast, and they usually don't need 152 00:08:40.919 --> 00:08:45.679 special wiring, although you know a certain degree or a 153 00:08:45.679 --> 00:08:48.960 certain amount of mocking is often involved. Then you've got 154 00:08:49.039 --> 00:08:52.720 integration tests, and you've got into end tests. It seemed 155 00:08:52.720 --> 00:08:55.480 to me that when you were talking about the improvements 156 00:08:55.519 --> 00:09:02.399 in our ability to do effective, efficient and performance testing, 157 00:09:02.480 --> 00:09:05.919 you were mostly talking about integration tests and end to 158 00:09:06.120 --> 00:09:08.759 end tests and less about unit tests. 159 00:09:08.799 --> 00:09:14.200 Correct yes and a little no. I mean, I think 160 00:09:14.200 --> 00:09:17.000 at first there is a trend. I'm not advocating, by 161 00:09:17.000 --> 00:09:19.759 the way, on anything. I'm just trying to cover what 162 00:09:20.039 --> 00:09:25.320 like trends that I observed. And I think another trend 163 00:09:25.440 --> 00:09:28.360 is seeing less unit tests and more kind of component 164 00:09:28.480 --> 00:09:32.039 page integration tests. But so called the testing diamond. It's 165 00:09:32.039 --> 00:09:34.840 a kind of an emerging concept where you should write 166 00:09:34.919 --> 00:09:39.320 more integration test but with the new tooling, even if 167 00:09:39.360 --> 00:09:42.039 you write unit tests like take a framework that we 168 00:09:42.080 --> 00:09:44.679 will cover with that. It's called VITAS browser mode. Even 169 00:09:44.759 --> 00:09:48.200 your unit tests run in the browser. Now think about it. 170 00:09:48.320 --> 00:09:50.000 Not all the unit tests in the world are pure. 171 00:09:50.240 --> 00:09:53.039 You might test the models that actually doing some stuff 172 00:09:53.080 --> 00:09:56.360 with the web APIs. So instead of running it in 173 00:09:56.559 --> 00:10:00.000 no JS with polyphils and trust and trust the similarity, 174 00:10:00.679 --> 00:10:03.080 it runs your test in a real browser and Chromium 175 00:10:03.159 --> 00:10:06.200 for example. So even your unit test gains more kind 176 00:10:06.240 --> 00:10:09.919 of production like experience. Does this make sense? 177 00:10:10.840 --> 00:10:14.240 Yes, it does, although it's I need to. 178 00:10:16.240 --> 00:10:20.440 Think about what I think about it. It's kind of 179 00:10:20.480 --> 00:10:27.000 a recursive statement because really, when I'm thinking about unit tests, 180 00:10:27.679 --> 00:10:31.039 I'm really trying in most cases to be as stateless 181 00:10:31.039 --> 00:10:37.120 as possible and consequently not be dependent on having access 182 00:10:37.320 --> 00:10:41.399 to a browser dome, which is kind of like persistent 183 00:10:41.559 --> 00:10:43.159 by the definition. 184 00:10:45.200 --> 00:10:47.840 And so I'm. 185 00:10:47.639 --> 00:10:52.000 Not sure that this is where I want my unit 186 00:10:52.039 --> 00:10:58.080 tests to go necessarily, but I certainly appreciate the fact 187 00:10:58.159 --> 00:11:03.960 that we don't need to mock browser a p I 188 00:11:04.200 --> 00:11:09.240 S domin p I S, which is always let's say, 189 00:11:09.600 --> 00:11:14.320 less than ideal. But why do you think that is 190 00:11:14.720 --> 00:11:18.440 what has changed that has made this What has changed 191 00:11:18.879 --> 00:11:20.559 that has made this possible? 192 00:11:20.679 --> 00:11:22.799 Now? That was not available before. 193 00:11:24.600 --> 00:11:27.799 Yeah, first forced margining that I fully agree. The real 194 00:11:27.879 --> 00:11:31.399 value is in is in integration tests and about page 195 00:11:31.399 --> 00:11:33.840 testing or component tests, also in end to end. In 196 00:11:34.399 --> 00:11:42.120 unit test, it's indeed less significant. So it's mostly if 197 00:11:42.279 --> 00:11:46.240 if we if we double double click into the front 198 00:11:46.320 --> 00:11:48.919 end word for a start. I think now there is 199 00:11:48.960 --> 00:11:52.639 a key up from decisions the team should make. There 200 00:11:52.679 --> 00:11:58.000 is kind of a triangle decision between three different frameworks 201 00:11:58.000 --> 00:12:03.360 for testing components and pages, and one of them is 202 00:12:03.840 --> 00:12:07.440 relatively One of them is is new. It's for the 203 00:12:07.480 --> 00:12:11.279 past two years, and it makes running visual test in 204 00:12:11.320 --> 00:12:14.600 the browser almost as fast as unit test. So there 205 00:12:14.639 --> 00:12:16.879 was a lot of people trying to avoid that kind 206 00:12:16.879 --> 00:12:20.440 of visual experience because of the hassle involved. But with 207 00:12:20.559 --> 00:12:23.360 some of the new frameworks like vitest browser mode and 208 00:12:23.639 --> 00:12:29.120 the new storybook model, it becomes easier faster, so you 209 00:12:29.159 --> 00:12:32.559 have less reasons not to just hey, let's test it 210 00:12:32.639 --> 00:12:36.720 like where the user lives. And maybe it's worth at 211 00:12:36.720 --> 00:12:40.399 some point explaining what is vitest browser mode, how does 212 00:12:40.440 --> 00:12:43.840 it stack with playwright go for it? 213 00:12:45.399 --> 00:12:45.559 Now? 214 00:12:45.639 --> 00:12:48.720 Is that the same as v test UI this browser mode? 215 00:12:50.360 --> 00:12:54.840 Ah, not exactly, because you can you can also have 216 00:12:54.879 --> 00:12:58.759 vites I with unit tests and browser mode is a 217 00:12:58.799 --> 00:13:03.360 new sub frameworks sort VI test and to zoom out 218 00:13:03.360 --> 00:13:05.559 for a second, I think that we have now three 219 00:13:05.960 --> 00:13:09.600 measure and good options in the front and word that 220

00:13:09.679 --> 00:13:12.240 one must make a decision up front. The first one 221 00:13:12.320 --> 00:13:18.840 is Playwright, the good all known Playwright, very reputable, very 222 00:13:18.879 --> 00:13:23.200 reputable frameworks that is being very popular, which allows you 223 00:13:23.240 --> 00:13:25.440 to test your system as is like you can spin 224 00:13:25.559 --> 00:13:28.120 up a whole page and you get a real page 225 00:13:28.759 --> 00:13:34.039 in a browser, but you get the all the you 226 00:13:34.120 --> 00:13:37.840 get the entire you get d their application with the 227 00:13:37.919 --> 00:13:40.759 router and the state and a lot of things beyond 228 00:13:40.840 --> 00:13:43.200 the page level or the component what you want to test. 229 00:13:43.600 --> 00:13:47.759 So the performance is like a little of concerning Playwright 230 00:13:47.840 --> 00:13:50.000 even when you mock the back. 231 00:13:50.840 --> 00:13:56.759 Correct me if I'm wrong. Playwright is essentially just a 232 00:13:56.840 --> 00:14:04.159 headless Chrome with api is to automate various tasks. That's 233 00:14:04.240 --> 00:14:05.519 more less it right. 234 00:14:07.200 --> 00:14:09.480 Yes, and maybe a little more because it can run 235 00:14:09.519 --> 00:14:11.960 your test in other browsers as well, like run the 236 00:14:12.000 --> 00:14:14.480 test once and get it running in Chrome or Safari 237 00:14:14.639 --> 00:14:15.440 or anything. 238 00:14:15.639 --> 00:14:18.919 Yeah, because they basically later on added support for the 239 00:14:18.960 --> 00:14:20.799 same APIs as I. 240 00:14:20.720 --> 00:14:24.519 Recall exactly they imported kind of a CDP protocol for 241 00:14:24.600 --> 00:14:28.320 other browsers exactly. But on top of this, they also 242 00:14:28.360 --> 00:14:33.679 have a very nice and stable API to make the 243 00:14:33.759 --> 00:14:38.519 experience of interacting with pages less flaky. So for example, 244 00:14:38.559 --> 00:14:40.879 if you're trying to assert on something, they will take 245 00:14:40.879 --> 00:14:45.320 care to retry in wait and until that condition is 246 00:14:45.360 --> 00:14:47.919 met on the page. So the overall testing experience is 247 00:14:48.759 --> 00:14:49.519 really great, and. 248 00:14:50.879 --> 00:14:55.960 It replaced all the frameworks like Cyprus and Selenium. 249 00:14:55.440 --> 00:14:59.039 Right exactly exactly. It is. 250 00:14:59.600 --> 00:15:03.320 Okay, So how is v test browser mode different from Playwright. 251 00:15:05.279 --> 00:15:08.000 Yeah, so first worldt mentioning mentioning is that the viittest 252 00:15:08.000 --> 00:15:12.279 browser mode is using Playwright under the hood, so you 253 00:15:12.440 --> 00:15:17.240 get a battle tested engine that you can trust. But 254 00:15:17.320 --> 00:15:19.320 what they try to do Playwright is a little bit 255 00:15:19.360 --> 00:15:23.799 disruptive to front and testing. So if you take traditional 256 00:15:23.840 --> 00:15:28.200 front and testing with people used to work with things 257 00:15:28.240 --> 00:15:31.519 like React Testing Library, the syntax of React Testing library, 258 00:15:31.879 --> 00:15:34.960 the concept of React Testing library supporting from there to 259 00:15:35.039 --> 00:15:41.159 Playwright was a little disruptive. Vitus Browser mode allows you 260 00:15:41.200 --> 00:15:44.720 to take your existing test or learn nothing new, use 261 00:15:44.759 --> 00:15:48.240 the same syntax, same concept, only instead of running it 262 00:15:48.279 --> 00:15:51.320 in nogs like React Testing Library, you run it in 263 00:15:51.480 --> 00:15:57.240 real browser, which obviously allows you to see. You probably 264 00:15:57.279 --> 00:16:00.559 know that that really nasty experience when you with React 265 00:16:00.559 --> 00:16:03.360 testing library and the test failed. What you see on 266 00:16:03.399 --> 00:16:07.120 the screen is three hundred lines of XML of HTML 267 00:16:07.440 --> 00:16:09.919 in a CLI in which you try to understand in 268 00:16:10.000 --> 00:16:12.519 this my component, how does it look like? What's wrong here? 269 00:16:12.559 --> 00:16:16.799 I mean nobody wants to travel shoot test in HTML 270 00:16:16.879 --> 00:16:20.000 over CLI. With this browser mode, you can see them 271 00:16:20.000 --> 00:16:22.559 in a browser and benefits from all of the Chrome 272 00:16:22.600 --> 00:16:29.039 Developer tool bar perks. So that's that's the premise. That's 273 00:16:29.080 --> 00:16:33.799 the obvious upside. They also make it like you know, 274 00:16:33.879 --> 00:16:37.600 vitt stands for fast in French, with the word wit 275 00:16:37.759 --> 00:16:39.720 is fast in French, and they they stand up to 276 00:16:39.759 --> 00:16:43.240 their reputation like I can I already use it in production. 277 00:16:43.639 --> 00:16:47.519 It's amazing. I can like test then entire page, some interactions, 278 00:16:47.519 --> 00:16:51.120 sometimes specific scenario in two hundred milliseconds. It's like blazing fast. 279 00:16:51.120 --> 00:16:54.000 You click on a button and you don't have enough 280 00:16:54.159 --> 00:16:55.960 chance to see what happened on the screen. So the 281 00:16:56.000 --> 00:17:02.679 performance is really awesome. With that, there are some cons 282 00:17:02.720 --> 00:17:07.359 that should be mentioned as well. It's very new, it's 283 00:17:07.400 --> 00:17:12.039 still tagged as experimental for like two years or so, 284 00:17:14.839 --> 00:17:18.559 and so's it's it didn't yet stand the battle of 285 00:17:18.640 --> 00:17:22.599 time like Playwright, And there are many other other decisions 286 00:17:22.640 --> 00:17:25.119 that were made that are that might appeal to some 287 00:17:25.200 --> 00:17:28.559 teams and not to others. But it's definitely a bold 288 00:17:28.680 --> 00:17:33.039 opportunity I think now in twenty twenty five for testing 289 00:17:33.079 --> 00:17:35.960 and I could not

finish that. I'm sorry. This is 290 00:17:36.039 --> 00:17:39.720 kind of a mini mini TED talk. But there is 291 00:17:39.759 --> 00:17:44.480 also Storybook, and Storybook now change their model and now 292 00:17:44.519 --> 00:17:48.519 they're using Vita's browser mode under the hood, which use Playwright. 293 00:17:49.160 --> 00:17:51.359 So you have three options. You know that Russian doll 294 00:17:51.440 --> 00:17:56.920 which is called I believe Matroyshka Babushka or matroys Career. 295 00:17:56.960 --> 00:17:59.680 It's kind of a nested doll. So what we have 296 00:17:59.799 --> 00:18:03.119 now the in the testing space, it's a troche car, 297 00:18:03.200 --> 00:18:05.359 so you have to play right or VI test brother 298 00:18:05.440 --> 00:18:07.559 mos that you just play right, or storybooks that use 299 00:18:08.720 --> 00:18:12.200 you got idea and you have to make you have 300 00:18:12.279 --> 00:18:13.279 to make the choice. 301 00:18:15.039 --> 00:18:20.599 So if you're starting a new front end project today, 302 00:18:20.680 --> 00:18:24.400 and let's say that you're using one of the two 303 00:18:24.480 --> 00:18:27.079 leading options, I don't know how familiar you are with them. 304 00:18:27.519 --> 00:18:31.720 Either React, which is approximately fifty percent of the market 305 00:18:32.480 --> 00:18:37.440 or view which is approximately twenty to thirty percent of 306 00:18:37.480 --> 00:18:41.720 the market, and you need to do and you want 307 00:18:41.759 --> 00:18:44.920 to do testing because we started with testing for the 308 00:18:44.960 --> 00:18:45.400 front end. 309 00:18:46.160 --> 00:18:47.759 What would you use? 310 00:18:51.240 --> 00:18:55.319 Yeah, that's that's first. Very I think that the choice 311 00:18:55.359 --> 00:18:57.400 is really hard today. I had to choose for a 312 00:18:57.440 --> 00:19:00.519 customer like two weeks ago, and the decision is really hard. 313 00:19:01.039 --> 00:19:04.880 So from one hand, Eileen Tower being conservative and say, hey, 314 00:19:04.920 --> 00:19:09.039 let's use play right. It works. It's like the most simple, 315 00:19:10.119 --> 00:19:13.240 it's like the last layer. It's simple, it's proven, let's 316 00:19:13.279 --> 00:19:14.920 just use it again. 317 00:19:15.119 --> 00:19:18.519 Playwright in this case is for mostly for the end 318 00:19:18.519 --> 00:19:19.599 to end tests. 319 00:19:19.359 --> 00:19:23.160 Right, or you can use it also for components, for 320 00:19:23.160 --> 00:19:26.359 for component or page testing. It it has great tooling 321 00:19:26.440 --> 00:19:28.599 to allow you to mock your bacond and I think 322 00:19:29.039 --> 00:19:32.400 I think that for the majority of the testing front 323 00:19:32.480 --> 00:19:36.599 and team want to mock the beacon. But yeah, you 324 00:19:36.640 --> 00:19:37.359 can also run in. 325 00:19:37.559 --> 00:19:40.960 Well, let's put it this way, if you're not mocking 326 00:19:41.039 --> 00:19:46.960 your back end, then you're by definition dependent. It's almost 327 00:19:47.000 --> 00:19:49.640 it kind of almost means that by definition your tests 328 00:19:49.640 --> 00:19:51.599 are going to be flaky at least some of the time. 329 00:19:52.920 --> 00:19:55.440 If you do involve if you do involve the back 330 00:19:57.640 --> 00:19:59.799 so so yes, if you want to go for a 331 00:20:00.000 --> 00:20:05.960 operations of concerns, you want to mock external services reduce 332 00:20:06.000 --> 00:20:09.039 the amount of dependencies that you have on external services 333 00:20:09.079 --> 00:20:10.599 and that includes your own back end. 334 00:20:12.920 --> 00:20:17.960 Totally agree, it's slower, it's more flaky. But my highest 335 00:20:17.960 --> 00:20:21.240 concern with end to end is actually that it's really 336 00:20:21.279 --> 00:20:25.279 hard to simulate things beyond heavy passes. So think about it. 337 00:20:25.359 --> 00:20:27.720 You want to test now that the user is disabled, 338 00:20:27.759 --> 00:20:31.200 it's a premium user. You get twenty records, all of 339 00:20:31.240 --> 00:20:34.559 the things that are beyond the simplest response. With real 340 00:20:34.599 --> 00:20:36.960 back end, it might be very hard to tune the 341 00:20:37.000 --> 00:20:39.839 backet into each and every scenarios that you want. However, 342 00:20:39.880 --> 00:20:42.279 when you mock it, you can simply just say, hey, 343 00:20:42.319 --> 00:20:44.200 I want this response, I want that response. 344 00:20:44.759 --> 00:20:45.000 Yeah. 345 00:20:45.079 --> 00:20:47.640 On the other hand, when you're mocking it, you're basically 346 00:20:47.680 --> 00:20:52.799 inventing the responses, so it's not necessarily how the system behaves. 347 00:20:52.839 --> 00:20:55.759 It's not really end to end if it's mocking. 348 00:20:56.400 --> 00:20:59.279 I mean when you normally like spin up like a 349 00:20:59.559 --> 00:21:02.720 maybe this is what you're talking about with mocking, but 350 00:21:02.839 --> 00:21:07.559 basically create a test database or a copy of your database, 351 00:21:07.599 --> 00:21:10.400 spin one upset it. You know, you got your data 352 00:21:10.440 --> 00:21:12.880 in it and then run your tests. Are actually using 353 00:21:12.920 --> 00:21:15.119 a database. It's not your Pride database obviously, but at 354 00:21:15.200 --> 00:21:17.160 least you have a clone and a real database that 355 00:21:17.200 --> 00:21:20.240 you're interacting with, because I know, speaking personal experience, a 356 00:21:20.279 --> 00:21:23.160 lot of times that's that's valuable testing. 357 00:21:23.960 --> 00:21:28.960 Yeah, yes, but I think that first before answering that 358 00:21:29.039 --> 00:21:31.319 I'd like to repeat to a dance. And I think 359 00:21:31.359 --> 00:21:35.400 it's like a key point here. If you're doing naive mocking, Okay, 360 00:21:35.440 --> 00:21:38.559 I'm just walking some Jason. Yeah, you are now at 361

00:21:38.599 --> 00:21:41.559 the risk of doing things that are not exactly the 362 00:21:41.559 --> 00:21:44.119 same in production, and this is why you also need 363 00:21:44.200 --> 00:21:47.039 contract testing. And I think this is definitely a topic 364 00:21:47.039 --> 00:21:49.279 we want to double click into later because there are 365 00:21:49.279 --> 00:21:53.119 a new interesting tooling in the JavaScript space related exactly 366 00:21:53.160 --> 00:21:57.400 with this problem. How do you create smart and aligned mogs? 367 00:21:58.000 --> 00:22:02.200 Now to this one. Yeah, in a simple case, it's 368 00:22:02.279 --> 00:22:04.559 just crawd applications, So you set up something in the 369 00:22:04.640 --> 00:22:08.640 database and you display them. But in many cases, creating 370 00:22:08.720 --> 00:22:11.480 the server response is much harder than this. Let's give 371 00:22:11.519 --> 00:22:15.119 you an example. One of my customers, is there a 372 00:22:15.599 --> 00:22:18.599 user entity and it might get disabled? How does it 373 00:22:18.640 --> 00:22:21.799 get disabled, there is some chron job running every minute 374 00:22:21.839 --> 00:22:24.799 and concluding that the user is disabled. So how would 375 00:22:24.920 --> 00:22:28.519 you create a disabled user from a test and make 376 00:22:28.559 --> 00:22:32.400 the back end run a chron job and tag it. 377 00:22:32.000 --> 00:22:36.839 It becomes much so much harder than just returning one line. Hey, 378 00:22:37.720 --> 00:22:41.319 just give me a disabled user, Jason, am I being 379 00:22:41.680 --> 00:22:42.480 clear on this point. 380 00:22:42.640 --> 00:22:47.799 No, it's It's obviously true like getting an application to 381 00:22:47.960 --> 00:22:53.519 do what you exactly what you want and can be challenging, 382 00:22:54.759 --> 00:23:00.960 and configuring a certain backend application to replicate certain scenarios 383 00:23:01.200 --> 00:23:05.319 is a problem. Just think about, you know, the scenario 384 00:23:05.359 --> 00:23:09.160 where a customer tells you that they're experiencing a certain problem, 385 00:23:09.240 --> 00:23:12.759 and maybe you even have traces that kind of show 386 00:23:12.799 --> 00:23:17.119 the problem the way that the customer is experiencing it, 387 00:23:17.240 --> 00:23:19.799 and now you're trying to replicate that scenario in the 388 00:23:19.880 --> 00:23:22.400 lab in order to debug it. We all know how 389 00:23:22.519 --> 00:23:28.079 challenging that can be. So I totally agree that replicating 390 00:23:29.599 --> 00:23:34.960 weird behaviors or edge certain edge behaviors that are legitimate 391 00:23:35.039 --> 00:23:40.079 and you want to test for in an actual environment 392 00:23:40.160 --> 00:23:44.119 that simulates production can be challenging, and it's often much 393 00:23:44.160 --> 00:23:50.480 easier to just play recordings of back end responses and 394 00:23:51.400 --> 00:23:54.839 you know, effectively mocking the back end it can be 395 00:23:55.039 --> 00:23:59.079 much much easier. Likewise, also mocking the front end. 396 00:23:59.160 --> 00:24:01.079 You know you want to. 397 00:24:02.599 --> 00:24:05.440 Test the back end, says system, then you might use 398 00:24:05.480 --> 00:24:07.880 some sort of I don't know, just post things out 399 00:24:07.920 --> 00:24:12.920 of Postman or whatever or playwright rather than actually having 400 00:24:12.960 --> 00:24:18.680 an actual front end that you know you create a 401 00:24:18.799 --> 00:24:21.799 process that clicks through. It can be much more challenging 402 00:24:21.880 --> 00:24:25.839 and more complicated than length fear. 403 00:24:28.240 --> 00:24:30.839 So is there a point, you know, Joanni, with that 404 00:24:30.920 --> 00:24:35.960 example you gave with the Crown job, is there a 405 00:24:35.960 --> 00:24:39.680 point where you have to say, Okay, we can't test 406 00:24:39.920 --> 00:24:42.519 everything in an automated test and maybe we have to 407 00:24:42.559 --> 00:24:45.880 do something manually or do you always work under the 408 00:24:45.880 --> 00:24:49.680 assumption that everything in Proude can be handled with an 409 00:24:49.680 --> 00:24:50.480 automated test. 410 00:24:52.799 --> 00:24:56.480 Yeah, so I'd say, based on my experience, as long 411 00:24:56.480 --> 00:25:00.160 as you stick to gray box testing. When I say 412 00:25:00.200 --> 00:25:03.119 gray box testing, it's a test that you don't test 413 00:25:03.119 --> 00:25:06.960 the entire system. You test one component. You test like 414 00:25:07.039 --> 00:25:09.640 the user like you just put things in and expect 415 00:25:09.640 --> 00:25:12.400 some outcomes. But you also run on the same process 416 00:25:12.759 --> 00:25:17.400 like that application. You can simulate anything I didn't encounter 417 00:25:17.480 --> 00:25:20.960 anything that is not simulated that it's hard to simulate 418 00:25:21.319 --> 00:25:24.680 as long as you are your test and the col 419 00:25:24.720 --> 00:25:29.559 under testing within the same process. In terms of ROI yeah, 420 00:25:29.599 --> 00:25:34.759 I would definitely prioritize and focus on first on the 421 00:25:34.759 --> 00:25:36.640 things that the user does and the things that are 422 00:25:36.799 --> 00:25:41.359 likely to happen and where the bugs are more severe. 423 00:25:42.160 --> 00:25:47.039 But in terms of the technical capability of simulating stuff, 424 00:25:47.160 --> 00:25:49.640 I think that today's in the jobs critical system, I 425 00:25:49.839 --> 00:25:53.440 rarely encounter the situation that something was not reproducible. 426 00:25:53.720 --> 00:25:57.759 And I'll add to that that manual testing is kind 427 00:25:57.799 --> 00:26:02.480 of like by definition anecdotal, and I hate to depend 428 00:26:02.559 --> 00:26:09.920 on anecdotal evidence of correctness. And so even if if 429 00:26:10.119 --> 00:26:13.880 if you do a certain amount of

manual testing, then 430 00:26:13.960 --> 00:26:17.480 record that testing and then turn it into an automated script, 431 00:26:18.200 --> 00:26:24.519 it doesn't need to stay manual forever. So a certain 432 00:26:24.559 --> 00:26:28.799 amount of manual testing, especially for I don't know complex systems, 433 00:26:28.960 --> 00:26:33.880 might be required, but again try to automate it going 434 00:26:33.920 --> 00:26:34.680 down the line. 435 00:26:36.200 --> 00:26:39.279 Definitely agree. I also have if someone is interested in articles, 436 00:26:39.279 --> 00:26:41.519 that's called testing the dark scenario of your back end, 437 00:26:41.559 --> 00:26:44.200 in which I've shown in an OGS application how each 438 00:26:44.599 --> 00:26:47.440 in seven lines tests no test there was more than 439 00:26:47.480 --> 00:26:52.640 seven lines. You can test really critical scenarios. It usually 440 00:26:52.720 --> 00:26:56.799 teams don't test, like, for example, test that when your bootstrap, 441 00:26:56.880 --> 00:27:02.440 your your beacon bootstop phase fails you then you get 442 00:27:02.440 --> 00:27:05.839 the right monitoring or test that when your live liveliness 443 00:27:06.400 --> 00:27:10.400 route is not behaving correctly, you get the right thing. 444 00:27:10.480 --> 00:27:12.960 Or when you know, very common in JavaScript word, when 445 00:27:12.960 --> 00:27:15.480 there is a process uncut exception or one hundred that 446 00:27:15.599 --> 00:27:19.119 rejection is happening in your application, do you do the 447 00:27:19.200 --> 00:27:21.680 right thing? This is a critical event. I mean your 448 00:27:21.720 --> 00:27:24.640 process might just crush. Do you test this? I showed 449 00:27:24.640 --> 00:27:26.680 out like in five lines of code you can also 450 00:27:26.759 --> 00:27:30.480 cover these advanced or endling passes. So yeah, I think that. 451 00:27:32.240 --> 00:27:35.160 So again to summarize what we've talked about so far, 452 00:27:35.240 --> 00:27:37.960 So basically what you're saying is that you're kind of 453 00:27:38.000 --> 00:27:46.559 promoting more sophisticated tests, usually component level tests or integration tests, 454 00:27:47.279 --> 00:27:53.039 less unit tests, and that using modern tooling that you 455 00:27:53.200 --> 00:27:58.279 mentioned like Playwright, like v test, browser mode, and like 456 00:28:00.680 --> 00:28:04.599 the storybook testing, it's much easier to do than it 457 00:28:04.720 --> 00:28:07.079 used to be, is basically what you're saying. If I 458 00:28:07.119 --> 00:28:08.519 understand correctly. 459 00:28:09.519 --> 00:28:12.720 Yeah, much easier than before. And i'd say, I mean 460 00:28:12.720 --> 00:28:15.279 I'm not against unit tests, but I think that for 461 00:28:15.400 --> 00:28:18.359 a start, your first step should be what the user 462 00:28:18.440 --> 00:28:21.839 is doing. And the user is not stretching functions or models, right, 463 00:28:21.839 --> 00:28:24.200 The user is visiting a page. So I think that 464 00:28:24.240 --> 00:28:27.599 the first things that you should test is your page, 465 00:28:28.119 --> 00:28:30.599 how the user interact, and then, as you call it, 466 00:28:30.680 --> 00:28:35.680 if you realize some complex, highly complex logics that you 467 00:28:35.720 --> 00:28:40.680 want to isolate, testing isolation for better whatever performance, then yeah, 468 00:28:40.799 --> 00:28:42.799 go for unit tests. But I think this is kind 469 00:28:42.799 --> 00:28:46.240 of contextual and optional. 470 00:28:46.720 --> 00:28:50.119 Well, to be fair, unit tests actually were kind of 471 00:28:50.160 --> 00:28:56.799 born out of TDD, So initially the concept of unit tests, 472 00:28:57.759 --> 00:29:02.759 at least for people who take that approach, is basically, 473 00:29:02.880 --> 00:29:06.119 we first write the tests, then we write the code. 474 00:29:07.039 --> 00:29:12.359 Now AI writes the tests and the code, so that's 475 00:29:12.960 --> 00:29:17.079 less that approach to be Also, to be fair, I've 476 00:29:17.200 --> 00:29:20.920 never been able to adopt the TDD model. Maybe I'm 477 00:29:20.920 --> 00:29:23.400 too much set in my ways. It just doesn't seem 478 00:29:23.440 --> 00:29:25.880 to be the way that I think. But I have 479 00:29:26.759 --> 00:29:32.039 gotten value from unit tests. The main advantage of unit 480 00:29:32.039 --> 00:29:36.640 tests is they're very lightweight, so they're almost effectively free 481 00:29:36.680 --> 00:29:40.000 to run, and they run, and I have them run 482 00:29:40.079 --> 00:29:46.200 constantly because you know, I don't care and if something break, 483 00:29:46.680 --> 00:29:51.079 it's there. They test very specific things so you get 484 00:29:51.119 --> 00:29:55.799 an instantaneous indication of where that problem is, so you 485 00:29:55.880 --> 00:29:59.119 don't need to think about Like with and let's say, 486 00:29:59.440 --> 00:30:03.559 especially the end to end tests. You know the test fails, Okay, 487 00:30:03.960 --> 00:30:08.400 why did it fail? It often requires pretty significant amount 488 00:30:08.480 --> 00:30:11.880 of investigation in order to understand why in end to 489 00:30:12.079 --> 00:30:16.119 end tests failed. And like we talked about before, it 490 00:30:16.160 --> 00:30:19.799 could also be related to certain things that to you know, 491 00:30:19.920 --> 00:30:24.279 flaky tests because you're dependent on external services. With unit tests, 492 00:30:24.319 --> 00:30:27.119 you're not dependent on everything and anything. You're just testing 493 00:30:27.160 --> 00:30:30.200 your own code. It runs very quickly. You're testing a 494 00:30:30.279 --> 00:30:32.839 very specific function. That's say, and. 495 00:30:32.839 --> 00:30:35.640 You know I've introduced I made a change that function. 496 00:30:35.720 --> 00:30:36.759 That's why it breaks. 497 00:30:37.240 --> 00:30:41.319 Couldn't agree

more. You know, I observed once something very 498 00:30:41.319 --> 00:30:44.119 interesting that can it back. For those of you are 499 00:30:44.119 --> 00:30:47.079 not familiar with, Can is kind of the father of DDD, 500 00:30:47.559 --> 00:30:50.480 A person who should really appreciate unit test and he 501 00:30:50.599 --> 00:30:53.680 said in one of his blog posts, unit tests give 502 00:30:53.759 --> 00:30:57.400 up predicting suitability to production and inspiring confidence. So it's 503 00:30:57.440 --> 00:31:00.559 not about product confidence at all. It's about being writeable, 504 00:31:00.720 --> 00:31:03.640 fast and specific. Now I think that most of the 505 00:31:03.640 --> 00:31:09.240 developers are looking in testing for confidence, and yes, it 506 00:31:09.279 --> 00:31:10.680 turns out it's not only about this. 507 00:31:11.880 --> 00:31:15.519 Yeah, there's that famous picture about the difference between unit 508 00:31:15.599 --> 00:31:19.440 tests and integration tests, about you know, building urinal in 509 00:31:19.519 --> 00:31:23.319 such a position that when you open the bathroom stall door. 510 00:31:23.200 --> 00:31:24.759 You can't actually get at it. 511 00:31:24.960 --> 00:31:28.519 So everything is fine at the unit level, at the 512 00:31:28.640 --> 00:31:31.359 unit level, but when you put it all together, it 513 00:31:31.400 --> 00:31:34.519 doesn't work. The only we've seem to have lost your picture. 514 00:31:34.599 --> 00:31:35.279 You're still with. 515 00:31:35.279 --> 00:31:38.279 Us, Yeah, I'm working on fixing the camera. The good 516 00:31:38.279 --> 00:31:41.759 news is that I'm still online. Yeah, okay, And I 517 00:31:41.759 --> 00:31:45.279 think it worth just revisiting for one last minute that 518 00:31:45.400 --> 00:31:51.759 that crucial decision between the three front end things. So 519 00:31:51.920 --> 00:31:57.240 we've mentioned playwright. I think that storybook and Viness browser 520 00:31:57.240 --> 00:32:00.640 mode are also great options. And we didn't story books 521 00:32:00.680 --> 00:32:04.119 or storybook if you have a design you asked me, 522 00:32:04.240 --> 00:32:07.039 what would I pick? So, if we're having a design 523 00:32:07.119 --> 00:32:10.000 system and rely a lot of small components that we 524 00:32:10.480 --> 00:32:14.599 tend to communicate and testing isolation, storybooks gives a very 525 00:32:14.599 --> 00:32:19.599 compelling package in which we get first a great workflow 526 00:32:20.039 --> 00:32:23.200 of visualizing our component and documenting them, but also a 527 00:32:23.240 --> 00:32:26.079 great new testing experience. So you write a test once 528 00:32:26.640 --> 00:32:29.759 and you get both accessibility testing. If you have some 529 00:32:29.799 --> 00:32:35.160 accessibility issue, you are covered visual regression testing without writing 530 00:32:35.240 --> 00:32:38.960 any other test out of the box, if you're if 531 00:32:39.000 --> 00:32:41.319 you are out of the box, and if you sharing 532 00:32:41.359 --> 00:32:47.759 your credit card, visual testing and also also functional testing 533 00:32:47.920 --> 00:32:50.799 using playwright under the hood. So you get where. 534 00:32:51.000 --> 00:32:55.799 Say visual testing just to verify that you understand your terminology. 535 00:32:56.279 --> 00:33:00.440 You basically mean that if I accidentally, I don't know, 536 00:33:00.680 --> 00:33:03.640 change the font or something, it will break my test 537 00:33:03.799 --> 00:33:08.240 because the content, the actual pixels on the screen are 538 00:33:08.279 --> 00:33:10.200 not the same as they're supposed to be. 539 00:33:11.200 --> 00:33:14.400 Exactly. It should be smarter than this, and it should 540 00:33:14.480 --> 00:33:21.440 realize and detect significant changes from like minor improvement, but 541 00:33:21.559 --> 00:33:24.279 exactly this. Since that functional tests can tell you about. 542 00:33:24.359 --> 00:33:30.599 So your side menu is crewed, something some I don't know, 543 00:33:31.359 --> 00:33:35.000 something is totally broken on the screen, a big change 544 00:33:35.039 --> 00:33:35.640 in the layout. 545 00:33:35.839 --> 00:33:39.039 You put a red background or something for testing purposes 546 00:33:39.079 --> 00:33:41.720 and then forgot to remove it. 547 00:33:42.559 --> 00:33:47.799 That's that's that's a nice example. Yeah, so storybook gives 548 00:33:47.799 --> 00:33:51.119 you all this together interesting option. On the other end, 549 00:33:51.519 --> 00:33:57.720 it's the more sophisticated option. It's it's all the storybook 550 00:33:58.000 --> 00:34:02.119 narratives and its browser modan hand play, right, So you 551 00:34:02.200 --> 00:34:04.799 have a lot of moving pieces to handle. 552 00:34:06.039 --> 00:34:08.599 Is it open source? Is it? Is it a paid 553 00:34:09.079 --> 00:34:10.760 product or service? 554 00:34:10.960 --> 00:34:12.239 What? What is it? 555 00:34:11.760 --> 00:34:15.719 It's open source, it's a free product, except I think 556 00:34:15.760 --> 00:34:18.639 this is how they make the revenues. Accept the visual 557 00:34:18.679 --> 00:34:23.559 regression test. What you just mentioned this one is is 558 00:34:23.559 --> 00:34:24.280 a paid service. 559 00:34:24.920 --> 00:34:25.639 Okay, cool? 560 00:34:26.159 --> 00:34:28.760 Hey, just to clarify terms. I know we've talked about 561 00:34:28.760 --> 00:34:32.880 different types of testing. We talked about unit testing, integration testing, 562 00:34:33.280 --> 00:34:35.400 and then testing visual testing. 563 00:34:36.239 --> 00:34:36.639 Uh. 564 00:34:37.159 --> 00:34:39.039 The one I always get hazy on is a difference 565 00:34:39.119 --> 00:34:42.760 between I guess what integration testing is how you find 566 00:34:42.760 --> 00:34:46.719 that versus unit tests versus and the s test? What 567 00:34:47.119 --> 00:34:50.199 excuse me, end to end testing? So what's

what's your 568 00:34:50.239 --> 00:34:54.519 definition of integration testing? Yea, because without understanding that, it's 569 00:34:54.519 --> 00:34:59.000 hard to get Dan's toilet joke about uh integration testing. 570 00:34:59.280 --> 00:35:05.239 So yeah, testing terminology is totally broken. I think that 571 00:35:05.280 --> 00:35:08.599 if when you say end to end testing, or or 572 00:35:08.639 --> 00:35:12.599 even unit testing and integrationion testing, different people will understand 573 00:35:13.079 --> 00:35:17.239 different things. And for this reason, maya suggestion in every 574 00:35:17.280 --> 00:35:20.440 company in your company, name a very specific name for 575 00:35:20.480 --> 00:35:24.559 your test which is not integration and to take for example, Uh, 576 00:35:24.840 --> 00:35:26.719 just one example. In one of my customers, we used 577 00:35:26.719 --> 00:35:29.320 to test pages. We take a front and page we 578 00:35:29.440 --> 00:35:32.800 isolated and the test cope is every time one page. 579 00:35:32.920 --> 00:35:36.440 We call it page testing, and every everyone understands what 580 00:35:36.480 --> 00:35:39.159 page testing means. It's a test of a page. Integration 581 00:35:39.280 --> 00:35:42.880 tests just mean might mean three different things two different people. 582 00:35:44.039 --> 00:35:47.400 If we try somehow to clarify the integration test is. 583 00:35:48.920 --> 00:35:52.159 I think that the most common definition of it that 584 00:35:52.519 --> 00:35:55.880 we test the user journey in a thing. We test 585 00:35:55.880 --> 00:35:59.320 the wholes user journey in a single tier of the system, 586 00:35:59.400 --> 00:36:02.960 not and then tire system. So it might be the 587 00:36:03.039 --> 00:36:06.559 user visiting some front end but without stretching the entire system. 588 00:36:06.840 --> 00:36:09.519 It might be a user visit a call, coding some 589 00:36:09.719 --> 00:36:13.960 backend API and expecting your response. This is our examples 590 00:36:14.000 --> 00:36:15.719 for integration tests. 591 00:36:15.440 --> 00:36:18.559 So correct me if I'm wrong. It also includes, for example, 592 00:36:18.639 --> 00:36:24.000 testing components in isolation. So let's say I implemented a 593 00:36:24.159 --> 00:36:28.920 date picker component that the date range picking component, because 594 00:36:29.000 --> 00:36:31.239 you know you don't get that out of the box 595 00:36:31.280 --> 00:36:34.679 in the browser, and I want to verify that it's 596 00:36:34.719 --> 00:36:41.760 working correctly, and I can host it within like independently 597 00:36:41.960 --> 00:36:46.639 and basically simulate interactions with it. It's not end to 598 00:36:46.800 --> 00:36:49.079 end because it's not done in the context of the 599 00:36:49.199 --> 00:36:52.360 entire application. So it's not for example, if you think 600 00:36:52.400 --> 00:36:55.599 about I don't know, let's say some sort of hotel 601 00:36:55.679

--> 00:36:59.719 booking application, it's not done in the context of actually 602 00:37:00.039 --> 00:37:05.760 looking stay at a hotel. It's done in isolation, but 603 00:37:05.920 --> 00:37:11.360 it's still done at the level of actual user interface 604 00:37:11.360 --> 00:37:17.079 and interactions, not just invoking functions and checking return values. 605 00:37:18.800 --> 00:37:22.039 Exactly. You're in other words, you are testing the user 606 00:37:22.119 --> 00:37:26.039 experience with this in its almost real environment. There is 607 00:37:26.079 --> 00:37:29.440 a browser here, There is your code. There is like 608 00:37:29.480 --> 00:37:32.960 you have multiple layers here orchestrating together. You have that 609 00:37:34.400 --> 00:37:38.280 UI kid usually the UI kids that you're using, and 610 00:37:38.320 --> 00:37:41.079 then you have your code that trapping it, and you 611 00:37:41.159 --> 00:37:44.639 have us sometimes your state manager, and there is the browser. 612 00:37:45.039 --> 00:37:49.159 All the moving pieces that exist in production are playing 613 00:37:49.159 --> 00:37:52.039 together in this test. In this regard, it's integration. 614 00:37:52.400 --> 00:37:55.480 And if you're talking about non front and stuff, for 615 00:37:55.519 --> 00:37:59.639 example a node system, it might be testing a particular 616 00:38:00.039 --> 00:38:03.800 specific API call. Again not in the context of an 617 00:38:03.840 --> 00:38:08.199 actual complete user flow, and potentially even with certain other 618 00:38:08.320 --> 00:38:11.800 parts of the system being mocked like we said before, 619 00:38:11.880 --> 00:38:14.880 and not actually working with let's say you're working with 620 00:38:15.000 --> 00:38:19.480 various external services. You're not actually working with these services, 621 00:38:19.519 --> 00:38:24.480 you're mocking them, and you're just testing effectively your particular 622 00:38:25.039 --> 00:38:28.880 let's say micro service or particular aspect of a microservice 623 00:38:28.920 --> 00:38:29.639 in isolation. 624 00:38:30.920 --> 00:38:34.159 Yes, exactly what I observe is it works great for 625 00:38:34.280 --> 00:38:38.239 backing teams is testing your entire micro services is put 626 00:38:38.599 --> 00:38:41.920 the infrastructure on local doctor composed, test all the layer 627 00:38:41.960 --> 00:38:45.039 of the micro service. If you deploy it, you test it, 628 00:38:45.119 --> 00:38:49.239 your data access there, your business logic, your API, but 629 00:38:49.639 --> 00:38:54.840 only intercept calls to other services. Otherwise it becomes end 630 00:38:54.840 --> 00:38:57.599 to end with all the downsides that we have mentioned. 631 00:38:58.320 --> 00:39:02.480 Yeah, I'm currently actually doing exactly that, and the portrait 632 00:39:02.559 --> 00:39:07.000 that I'm working on. Being able to use a doctor 633 00:39:07.079 --> 00:39:14.559 composed to effectively recreate the entire system

on your own 634 00:39:14.679 --> 00:39:20.719 particular on your local developer machine and then be able 635 00:39:20.840 --> 00:39:25.559 to dynamically deploy changes to the specific micro service that 636 00:39:25.599 --> 00:39:30.400 you're working on is effectively a godsend. It makes life 637 00:39:30.559 --> 00:39:34.920 so much easier when developing in such complex environments. 638 00:39:34.960 --> 00:39:38.199 Definitely, I mean's surprising how fast this test is. Like 639 00:39:38.559 --> 00:39:42.320 I think usually like a fifty proximately fifty test for 640 00:39:42.400 --> 00:39:46.920 micro service on a developer machine with real database last 641 00:39:47.000 --> 00:39:53.480 lesson twenty seconds. So yeah, it's a really efficient technique 642 00:39:53.480 --> 00:39:56.800 and maybe it's worth mentioning if we already discuss back end. 643 00:39:57.000 --> 00:40:00.400 There is now a build in test runners in o JS, 644 00:40:00.840 --> 00:40:04.119 mostly for back end, I think, and it's really gaining 645 00:40:04.119 --> 00:40:05.039 a lot of momentum. 646 00:40:06.079 --> 00:40:13.639 Yeah, it's it's interesting that it seems that between node 647 00:40:14.119 --> 00:40:18.079 implementing its own testing built in testing capabilities on the 648 00:40:18.119 --> 00:40:21.159 one hand, and v test on the other end, they're 649 00:40:21.239 --> 00:40:25.400 kind of cornering the testing market in a sense. And 650 00:40:25.559 --> 00:40:29.519 I recall, for example, seeing this post on x I 651 00:40:29.559 --> 00:40:34.800 think from KENC. Dodds talking about how happy is that 652 00:40:35.000 --> 00:40:38.800 the React testing library is no longer needed because effectively 653 00:40:38.840 --> 00:40:41.960 all this functionality is now built into v test. 654 00:40:42.039 --> 00:40:46.440 I think yes, And I think it's really interesting to 655 00:40:46.480 --> 00:40:54.719 see how people are contributing libraries and ideas together. So 656 00:40:54.960 --> 00:40:58.400 it reactesting Library was a library, but it was also 657 00:40:58.400 --> 00:41:01.480 a philosophy of how you test. Now the library is 658 00:41:02.039 --> 00:41:05.559 probably fading away or at least losing some momentum, but 659 00:41:05.920 --> 00:41:10.400 the concept that the concepts there, the techniques are being 660 00:41:10.920 --> 00:41:13.719 totally used in Vitas browser mode and in Playwright, and 661 00:41:13.800 --> 00:41:16.760 so many new testing framework are taking inspiration and building 662 00:41:16.760 --> 00:41:21.559 on this that It's very very interesting to see this 663 00:41:23.679 --> 00:41:25.119 evolving process. 664 00:41:27.960 --> 00:41:31.039 So in general, what would you say, is it still 665 00:41:31.320 --> 00:41:34.239 much more difficult to test the front end than the 666 00:41:34.320 --> 00:41:37.960 back end or are we achieving some sort of parity? 667 00:41:38.599 --> 00:41:43.079 M m yeah, I tell that I'd easily said that 668 00:41:43.239 --> 00:41:47.480 testing a back end is much much easier than testing 669 00:41:47.480 --> 00:41:55.199 a front end. Yeah. I mean, first, you can also 670 00:41:55.280 --> 00:41:58.119 judge it by the tooling. I means there are except 671 00:41:58.159 --> 00:42:01.280 that new test runners of NOJ that is actually bring 672 00:42:01.440 --> 00:42:04.519 less features. It's funny because the new test runners that 673 00:42:04.679 --> 00:42:07.320 a lot of back end teams are now using, it's 674 00:42:08.920 --> 00:42:11.960 us it brings much less testing features and what used 675 00:42:11.960 --> 00:42:15.320 to be like ingested people don't need even all the features, 676 00:42:15.360 --> 00:42:19.440 and there are very few like new tooling in the 677 00:42:19.480 --> 00:42:23.320 back and accept maybe contract testing tooling, which we will mention. However, 678 00:42:23.360 --> 00:42:25.320 you see that in the front and words there is 679 00:42:25.320 --> 00:42:29.559 a bloom and always constant change and struggling to build 680 00:42:29.679 --> 00:42:33.440 new tooling that are can catch up with the front 681 00:42:33.480 --> 00:42:35.199 and complexity so well. 682 00:42:35.400 --> 00:42:40.199 The fact that we actually need visual testing is an 683 00:42:40.199 --> 00:42:44.320 indication of how much more challenging front test and testing is. 684 00:42:44.360 --> 00:42:49.159 It's it's like almost like saying, let's let's test the 685 00:42:49.599 --> 00:42:54.719 bits flowing over some network connection and and basically try 686 00:42:54.760 --> 00:42:59.760 to analyze the system from that. If I'm using like 687 00:43:00.280 --> 00:43:05.039 silly analogy. So you were talking about you mentioned contract testing, 688 00:43:05.400 --> 00:43:06.719 Can you elaborate on that? 689 00:43:08.199 --> 00:43:10.039 Oh, definitely, I think that this is now one of 690 00:43:10.039 --> 00:43:15.760 the most interesting fields in testing in overall. I would 691 00:43:15.760 --> 00:43:18.320 even say that if in twenty twenty five you're not 692 00:43:18.400 --> 00:43:22.119 doing your your coding against API without any explicit contract, 693 00:43:22.920 --> 00:43:25.360 it shouldn't be an option on the table. First, let 694 00:43:25.440 --> 00:43:28.400 me explain there is. So we are all consuming API, 695 00:43:28.599 --> 00:43:30.639 whether front and to back and back and to backan 696 00:43:31.280 --> 00:43:38.519 and every system is distributed. Now, if as a consumer 697 00:43:38.559 --> 00:43:41.760 of an API, the payloads that I'm sending and receiving 698 00:43:41.840 --> 00:43:44.880 for me are just kind of unknown jason, some kind 699 00:43:44.880 --> 00:43:49.119 of Jason that I can control and predict, then imate 700 00:43:49.280 --> 00:43:53.920 risk of receiving a lot of breaking changes. So say 701 00:43:53.960 --> 00:43:57.440 for example, things work, but then my the APIs and 702 00:43:57.519 --> 00:44:02.639

un consuming change something. My test pass, but production will fail. 703 00:44:03.519 --> 00:44:06.320 And we are at this point we are really good 704 00:44:06.320 --> 00:44:12.840 at testing our components. But but the ocean, that ocean 705 00:44:12.880 --> 00:44:17.880 between components, we have much less testing ways and techniques 706 00:44:17.920 --> 00:44:21.320 to ensure the two different parties are aligned on the 707 00:44:21.400 --> 00:44:24.039 on the API between the two and now there is 708 00:44:24.079 --> 00:44:27.599 a there is a bloom of new techniques and tooling 709 00:44:27.760 --> 00:44:31.760 to cover exactly this as API consumer, how do I know? 710 00:44:31.880 --> 00:44:35.920 How do I get a guarantee that the payloads are 711 00:44:36.039 --> 00:44:38.400 changing and I should be prepared for that. 712 00:44:40.400 --> 00:44:43.320 So what are the solutions in this space? 713 00:44:44.559 --> 00:44:48.639 Yeah, so there is a spectrum of tooling and options. Uh, 714 00:44:48.840 --> 00:44:52.519 the simplest one just in a moder repot, just creates 715 00:44:52.599 --> 00:44:55.360 OD schemas, share them between the back end and the 716 00:44:55.360 --> 00:44:59.639 front end, and you get a solid protection, right, I mean, 717 00:45:00.239 --> 00:45:03.920 whenever the back end changes the schemas, your front and 718 00:45:04.039 --> 00:45:07.360 tests or it can be also your back end will 719 00:45:07.360 --> 00:45:12.239 fail because you have now you can generate typing from this. 720 00:45:13.360 --> 00:45:16.639 From these schemas, you can also do runtime validation and 721 00:45:16.679 --> 00:45:19.199 you're protected. But it doesn't cover a lot of frisk. 722 00:45:19.280 --> 00:45:22.880 On the other sound side of the spectrum, the more 723 00:45:23.039 --> 00:45:27.519 fancy tooling are mocking servers that will let you know 724 00:45:27.599 --> 00:45:31.800 that we'll provide you with a very sophisticated mechanism to 725 00:45:32.400 --> 00:45:36.039 synchronize the two parties. But there is the middle. I 726 00:45:36.079 --> 00:45:40.599 think that most of the solutions that are somewhere located 727 00:45:40.599 --> 00:45:43.480 in the middle between that are not too complex and 728 00:45:43.760 --> 00:45:47.320 not too simple and are not simplistic. Use the following. 729 00:45:47.880 --> 00:45:53.400 The API provider generates open API, usually it comes for free, 730 00:45:53.559 --> 00:45:56.519 and it somehow shared that open API with the consumer. 731 00:45:57.239 --> 00:46:00.920 Now the consumer all the consumers have tooling to generate 732 00:46:01.039 --> 00:46:05.280 both types, but also run time validation out of open API. 733 00:46:06.519 --> 00:46:10.440 So now as a fronten for example, whenever my back 734 00:46:10.559 --> 00:46:14.039 end changes the open API, if my types I say 735 00:46:14.119 --> 00:46:17.840 that's a new field was added, my code will generate 736 00:46:17.920 --> 00:46:21.159

types types based on this open API. And if I'm 737 00:46:21.199 --> 00:46:24.280 trying to access some fields it doesn't exist anymore or 738 00:46:24.400 --> 00:46:28.239 is the wrong type, I will get compile time or 739 00:46:28.320 --> 00:46:31.440 testing failure right away. So it's kind of me and 740 00:46:31.719 --> 00:46:35.840 back end are always aligned without spinning up a complex 741 00:46:35.880 --> 00:46:37.519 and fancy edit end environment. 742 00:46:38.440 --> 00:46:41.920 Yeah, but like you said, it almost kind of necessitates 743 00:46:41.960 --> 00:46:47.119 in many cases working in a mono repo. Otherwise, First 744 00:46:47.119 --> 00:46:49.519 of all, I hesitate to call it tests. I mean, 745 00:46:49.760 --> 00:46:54.639 it's it's basically you know, one of the benefits of 746 00:46:54.800 --> 00:47:01.400 using static typing is that you get certain bugs become 747 00:47:02.159 --> 00:47:05.840 impossible because of the type system. But I don't really 748 00:47:05.920 --> 00:47:10.440 consider it to be tests. It's basically that's it might 749 00:47:11.519 --> 00:47:14.480 You could call it another step in your build that 750 00:47:14.639 --> 00:47:18.119 might fail, but I hesitate to call it tests. 751 00:47:18.960 --> 00:47:22.239 Maybe it is. I don't know. That's an interesting philosophical question. 752 00:47:22.719 --> 00:47:25.800 Some of the fancy solutions are going beyond static typing. 753 00:47:25.880 --> 00:47:28.199 So what they do is that if you have an 754 00:47:28.199 --> 00:47:32.039 open API, it's not all about typing. Some kind of 755 00:47:32.079 --> 00:47:36.400 the assertions there are can be asserted only on run time. Say, 756 00:47:36.440 --> 00:47:39.159 for example, you have a rejects some of the fields 757 00:47:39.199 --> 00:47:43.360 type is a rejics. You can't assert this in typing Typescript. Also, 758 00:47:43.480 --> 00:47:46.320 Typescript is now adding some kind of tragic support. So 759 00:47:46.360 --> 00:47:50.119 what these tooling are doing during testing? Whenever you are 760 00:47:50.159 --> 00:47:53.079 sending to your mock some on payload, they are very 761 00:47:53.119 --> 00:47:58.199 fine whether the received testing payload is indeed conforming to 762 00:47:58.239 --> 00:48:02.800 the open API definition. So in that regard they can 763 00:48:03.119 --> 00:48:06.639 make the test fail if what you're using, the pedal 764 00:48:06.719 --> 00:48:12.719 you're using is not what the Beckend API defined. Of 765 00:48:12.800 --> 00:48:15.320 course for this you need like real testing. And maybe 766 00:48:15.360 --> 00:48:18.280 it's worth mentioning one work very cool features. I guess 767 00:48:18.320 --> 00:48:22.079 you both know ms W. It's a very popular JavaScript 768 00:48:22.159 --> 00:48:25.239 library for mocking intercepting HTP requests. 769 00:48:26.119 --> 00:48:27.079 No, sorry, don't. 770 00:48:27.519 --> 00:48:31.079 Oh okay, So let me introduce it. Ms w H 771 00:48:32.039 --> 00:48:36.079 is a library that allows by the way both also 772 00:48:36.079 --> 00:48:38.880 know jes also back in. It allows you to kind 773 00:48:38.880 --> 00:48:43.079 of define in your in your environment, kind of in 774 00:48:43.280 --> 00:48:46.440 kind of intercept network requests and say, hey, whenever I'm 775 00:48:46.480 --> 00:48:51.760 approaching another API, say the beckend, intercept it in browser. 776 00:48:51.800 --> 00:48:54.400 It's using by the way service workers, so you will 777 00:48:54.440 --> 00:48:57.400 see the request in the network tub and Chrome. But 778 00:48:57.559 --> 00:49:01.239 instead of going out outside your browser, MSW will return 779 00:49:01.320 --> 00:49:06.159 the response on behalf of your API, so you get 780 00:49:06.159 --> 00:49:11.599 a much more a much faster, and more predictable working environment. Now, 781 00:49:11.719 --> 00:49:15.400 ms W added a new feature this year, and we're 782 00:49:15.440 --> 00:49:20.800 discussing recent changes that instead of you defining what will 783 00:49:20.840 --> 00:49:23.400 be the API responses and you might be wrong there 784 00:49:23.480 --> 00:49:25.719 right what you think that there might be threefills in 785 00:49:25.719 --> 00:49:29.039 this payload, the beck end in fact, in production will 786 00:49:29.079 --> 00:49:33.400 return four or two. Ms W generates the responses from 787 00:49:33.440 --> 00:49:36.920 the beckend open API. So you provided, hey, this is 788 00:49:36.920 --> 00:49:39.760 what the API provided us the real world, the real 789 00:49:40.079 --> 00:49:43.639 scheme as it was generated, and it takes and generates 790 00:49:43.679 --> 00:49:47.079 responses based on this. So it's kind of you are 791 00:49:47.239 --> 00:49:51.920 you and the API that you consume are aligned, are 792 00:49:51.960 --> 00:49:54.360 always aligned? Does it make sense? 793 00:49:54.639 --> 00:49:55.599 Yes, as it does. 794 00:49:56.679 --> 00:49:59.519 Yeah, MSW stands from Mark Service Worker in case that 795 00:49:59.559 --> 00:50:02.800 wasn't clear already. So before we wrap up, is there 796 00:50:02.840 --> 00:50:05.039 anything else you wanted to talk about real quick that 797 00:50:05.039 --> 00:50:06.760 we have not yet covered in this world? 798 00:50:09.400 --> 00:50:11.960 Let me see, I've prepared a table of content here 799 00:50:12.559 --> 00:50:17.639 on the side. We can discuss a little bit more 800 00:50:17.679 --> 00:50:20.719 about testing with it. I also have we have covered 801 00:50:20.719 --> 00:50:28.039 this nothing significant syd with We have touched the measure 802 00:50:28.159 --> 00:50:33.119 points and advancement in testing, all right. 803 00:50:34.199 --> 00:50:39.440 So if before we I do have a final question though, 804 00:50:39.480 --> 00:50:43.559 in this context, So when you usually are when you're 805 00:50:43.599 --> 00:50:49.119 brought into a new project and and you want to 806 00:50:49.639 -->

00:50:52.559 before making any changes, I assume you want to make 807 00:50:52.599 --> 00:50:55.800 sure that you have good test coverage, because who wants 808 00:50:55.800 --> 00:50:58.559 to start making changes in a project before we have 809 00:50:58.639 --> 00:51:03.280 good testing coverage. What's the main thing that you're looking 810 00:51:03.400 --> 00:51:07.000 for in order to say this project has good test 811 00:51:07.079 --> 00:51:10.320 coverage versus another project that you might say this project 812 00:51:10.400 --> 00:51:12.239 does not have good test coverage. 813 00:51:15.079 --> 00:51:21.840 Yeah, that's that's the million dollar question. I'm even struggling 814 00:51:21.880 --> 00:51:30.960 to enter it. It's so it's so deep. But I 815 00:51:31.000 --> 00:51:34.440 think that I'm using two I'm using two aroistics for this. 816 00:51:37.000 --> 00:51:41.880 The first one is unfortunately not not in period. It's 817 00:51:41.880 --> 00:51:45.280 not measured. Uh, if people are moving, if people are 818 00:51:45.320 --> 00:51:50.719 really are having confidence to deploy fast, deploying deploying fast 819 00:51:50.760 --> 00:51:53.519 and just feeling confident to do it, you see it, 820 00:51:53.639 --> 00:51:56.960 I mean you you feel it, they trust are testing. 821 00:51:57.559 --> 00:52:03.320 That's I think the ultimate eventually the ultimate test. On 822 00:52:03.400 --> 00:52:04.679 top of this, I'm. 823 00:52:04.760 --> 00:52:07.760 It's literally just to just to pause you for a second, 824 00:52:08.239 --> 00:52:13.280 it's literally that subjective feeling that you test. You can 825 00:52:13.360 --> 00:52:18.199 ask people working on the project, how well do you 826 00:52:18.440 --> 00:52:23.239 trust your tests? If you if you make change and 827 00:52:23.280 --> 00:52:26.079 it makes its way all the way to to to 828 00:52:26.199 --> 00:52:31.039 being deployed to production, how secure do you feel about 829 00:52:31.119 --> 00:52:35.119 it working properly and not breaking any of the things 830 00:52:35.239 --> 00:52:39.679 or introducing any form of regression. It's it's essentially that 831 00:52:39.679 --> 00:52:42.239 that level of subjective feeling in the team. 832 00:52:42.719 --> 00:52:45.519 Yeah, that's the bottom line. I'm also using test coverage. 833 00:52:46.039 --> 00:52:48.440 I'm using branch coverage. Yeah, I know that it's not 834 00:52:48.480 --> 00:52:52.039 a perfect metric, but it's very simple metric and for 835 00:52:52.119 --> 00:52:54.519 the majority of teams, they are not like they're not 836 00:52:54.679 --> 00:52:58.360 fooling themselves. If if if system has a very high 837 00:52:58.440 --> 00:53:03.400 branch coverage. It's means that the testing are indeed testing 838 00:53:03.559 --> 00:53:06.000 most of that ninety coverage, not all of them, but 839 00:53:06.119 --> 00:53:09.599 a high percentage of this at least it tells me 840 00:53:10.199 --> 00:53:14.159 that most of the system areas are are trying to 841 00:53:14.199 --> 00:53:16.320 be tested. On top of this, I usually add some 842 00:53:16.400 --> 00:53:22.159 kind of mutation testing. So mutation testing means that it's 843 00:53:22.199 --> 00:53:25.719 a framework that plant bugs in your code, Like it 844 00:53:25.760 --> 00:53:28.480 can plant a bug in each and every line and 845 00:53:28.519 --> 00:53:30.639 then it runs your test and then true that they fail. 846 00:53:31.840 --> 00:53:36.079 And if yeah, now, because I mean it's it's an 847 00:53:36.519 --> 00:53:42.719 it's amazingly efficient technique to realize testing efficiency. But on 848 00:53:42.760 --> 00:53:45.280 the other end, it's really slow and cumbersome. So what 849 00:53:45.400 --> 00:53:48.840 I usually do I run it like I run it 850 00:53:48.880 --> 00:53:53.119 occasionally just to get a sense of what is the 851 00:53:53.239 --> 00:53:57.280 ratio between the COD coverage and the mutations. So in 852 00:53:57.320 --> 00:54:00.400 other words, if I see that I have COD coverage 853 00:54:00.559 --> 00:54:03.320 of ninety percent in a file, say I find file 854 00:54:03.360 --> 00:54:07.760 as ninety percent coverage and the mutation score is very low, 855 00:54:08.400 --> 00:54:10.639 it tells me that this is a nonsense coverage. You 856 00:54:10.679 --> 00:54:12.760 know that that's just visits amer coat off it. But 857 00:54:12.800 --> 00:54:15.800 it's not asserting one anything On the other hand, if 858 00:54:15.840 --> 00:54:19.079 I seek correlation between coverage and mutations in file, I 859 00:54:19.159 --> 00:54:24.480 know that this coverage is authentic. It's like trustworthy. So 860 00:54:24.599 --> 00:54:29.000 this is the empiric side of measuring that the test. 861 00:54:29.559 --> 00:54:32.000 Cool, Okay, thank you for that. 862 00:54:34.280 --> 00:54:39.480 Elucidation is that how you put elucidations. 863 00:54:39.639 --> 00:54:40.079 That's the word. 864 00:54:40.079 --> 00:54:43.880 I'll I'll google that word right after elucidate to explain 865 00:54:44.000 --> 00:54:46.119 something clearly is basically what it means. 866 00:54:46.159 --> 00:54:50.280 So that is the word of the day, elucidate already. 867 00:54:50.559 --> 00:54:53.119 So without move on the picks, picture that part of 868 00:54:53.119 --> 00:54:55.320 the show where we get to talk about anything we 869 00:54:55.360 --> 00:55:00.920 want to within reason of course, board, games, books, food, movie, tech, 870 00:55:01.000 --> 00:55:04.920 if you want to, I'll go first and my pick. 871 00:55:05.480 --> 00:55:07.039 I do have a pick before I get to the 872 00:55:07.159 --> 00:55:08.960 high point of the Dad jokes of the week, and 873 00:55:09.000 --> 00:55:11.639 it has to do with dad jokes. So at the 874 00:55:11.719 --> 00:55:14.639 end of our podcast episode that Dan and I recorded 875 00:55:14.719 --> 00:55:19.000 with Ryan Carneato and Tanner Lindsley, I found out something 876 00:55:19.039 -->

00:55:20.960 that I wasn't aware of that now Crosby, who was 877 00:55:20.960 --> 00:55:25.039 the founder of a G Grid, had died and apparently 878 00:55:25.079 --> 00:55:26.760 it is some type of helicopter accident. 879 00:55:27.039 --> 00:55:30.599 Well, it's to be fair, it's not new right. 880 00:55:30.440 --> 00:55:32.280 I know it is the first I'd found about it, 881 00:55:32.280 --> 00:55:34.679 you know. So yeah, that just tells you how out 882 00:55:34.679 --> 00:55:37.880 there in the news I am every day, and that 883 00:55:37.920 --> 00:55:41.119 brings back memories for me. So four years ago we 884 00:55:41.159 --> 00:55:45.719 did a Jobascript Jobber episode with Nile, myself and Ag 885 00:55:46.000 --> 00:55:49.960 and just a j O'Neal and Chuck and the Reason 886 00:55:50.000 --> 00:55:52.360 that I was. Episode five oh four came out in 887 00:55:52.440 --> 00:55:55.119 October of twenty twenty one, and this was one of 888 00:55:55.159 --> 00:55:57.639 my favorite episodes, as I mentioned before, because when it 889 00:55:57.679 --> 00:55:59.480 got to the end, Nile just jumped right in and 890 00:55:59.519 --> 00:56:04.880 throughout a whole string of geek dad jokes that I 891 00:56:05.000 --> 00:56:07.280 had to do with going to a no sequel bar 892 00:56:07.360 --> 00:56:09.000 and there's no table, So then he went to a 893 00:56:09.039 --> 00:56:11.599 sequel bar that had tables, but he couldn't join anybody, 894 00:56:11.639 --> 00:56:12.760 and it was awesome. 895 00:56:13.000 --> 00:56:14.320 I was giving him a standing ovation. 896 00:56:15.039 --> 00:56:18.000 So that's my pick of the for the week is 897 00:56:18.360 --> 00:56:21.719 you can find it on Top Endos episode five o 898 00:56:21.880 --> 00:56:24.039 four and if you want to look at the transcript 899 00:56:24.159 --> 00:56:26.760 or listen to it, it's it's one of my faves. 900 00:56:27.440 --> 00:56:30.480 So with that, we'll get to the dad jokes of 901 00:56:30.519 --> 00:56:35.000 the week. So question, how come nobody laugh at the 902 00:56:35.079 --> 00:56:36.800 joke about the fault DG eighteen. 903 00:56:38.360 --> 00:56:39.440 It was poor execution. 904 00:56:43.159 --> 00:56:44.400 That's good, all right. 905 00:56:44.639 --> 00:56:46.840 My wife told me the other day, I'm really getting 906 00:56:46.880 --> 00:56:49.639 sick of you over using contractions, and I said, it's 907 00:56:49.639 --> 00:56:56.760 what it is. And then, uh, someone threw a can 908 00:56:56.800 --> 00:56:58.760 of soda at me today or i'd say pop on 909 00:56:58.800 --> 00:56:59.440 the West coast. 910 00:56:59.440 --> 00:57:05.360 But I'm all right. It was a soft drink. And 911 00:57:05.400 --> 00:57:06.519 those are the dad jokes of. 912 00:57:06.480 --> 00:57:09.000 The week, and those who are definitely dad jokes. 913 00:57:08.800 --> 00:57:09.440 Yes they were. 914 00:57:09.760 --> 00:57:12.239 I'm embarrassed to admit that I understood two of the three. 915 00:57:13.039 --> 00:57:16.320 I promised to to rewatch that section and catch up. 916 00:57:16.800 --> 00:57:18.440 Okay, yeah, I can explain it. Yeah, it sort of 917 00:57:18.480 --> 00:57:20.639 kills a joke and you have to explain it. 918 00:57:21.079 --> 00:57:21.880 That is. 919 00:57:24.039 --> 00:57:25.559 Dan. What do you got for us? 920 00:57:26.440 --> 00:57:29.119 Okay, I've got a couple of things. So first of all, 921 00:57:29.199 --> 00:57:32.119 I posted it on x but I also want to 922 00:57:32.159 --> 00:57:35.800 shout it out a big thanks to THEO also known 923 00:57:35.840 --> 00:57:38.840 as THEO three GG, who puts out a lot of 924 00:57:38.880 --> 00:57:44.800 content online. I'm working on a project that involves Stripe 925 00:57:44.800 --> 00:57:49.239 integration for one of our services, and I've not worked 926 00:57:49.239 --> 00:57:54.639 with Stripe significantly, I think, ever, so I was kind 927 00:57:54.639 --> 00:57:57.199 of learning a lot about the stripe APIs and how 928 00:57:57.239 --> 00:58:03.119 it works, and they seem but Theo's video was essentially 929 00:58:04.480 --> 00:58:09.039 all the stripe gotchas that he encountered while working on 930 00:58:09.119 --> 00:58:16.119 various projects and how to overcome them, and it was very, 931 00:58:16.480 --> 00:58:19.480 very helpful. It came along at just the right time 932 00:58:20.679 --> 00:58:26.400 and undoubtedly saved me from a lot of pain down 933 00:58:26.480 --> 00:58:29.480 the line. So thank you for that, CEOs, Thank you 934 00:58:29.559 --> 00:58:32.599 for sharing this information. This is what sharing on the 935 00:58:32.599 --> 00:58:36.400 web is all about. So that's one thing I wanted 936 00:58:36.440 --> 00:58:40.920 to mention. Another is that we did this kind of 937 00:58:41.480 --> 00:58:45.360 watch along party at work. It's something that we tried 938 00:58:45.400 --> 00:58:48.280 for the first time and got a lot of good responses. 939 00:58:49.239 --> 00:58:53.719 It's great when you can have somebody at work giving 940 00:58:53.760 --> 00:58:57.639 a talk on a particular topic, but it's not always possible. Obviously, 941 00:58:58.119 --> 00:59:01.679 you may not have experts in how on everything. So 942 00:59:01.760 --> 00:59:04.119 what we actually did is we wanted to learn more 943 00:59:04.159 --> 00:59:08.239 about MCP and Jack Harrington, who we've had on the 944 00:59:08.280 --> 00:59:10.480 show a couple of times. He was also one of 945 00:59:10.480 --> 00:59:19.079 the hosts on the React podcast, and he created a 946 00:59:19.159 --> 00:59:24.199 series of videos about MCP that are just excellent. So 947 00:59:24.559 --> 00:59:27.639 we watched several of them in a row. They're actually 948 00:59:27.679 --> 00:59:30.719 fairly short each They're like ten to twenty minutes each 949 00:59:32.079 --> 00:59:35.519 and he explains things so well. So we basically watched 950 00:59:35.519 --> 00:59:38.039 a video and then had a brief discussion about

it, 951 00:59:38.119 --> 00:59:42.079 and then watched the next video and so forth, and 952 00:59:42.400 --> 00:59:47.119 it was really great and I highly recommend this if 953 00:59:47.199 --> 00:59:51.960 you want to educate people at work. It's a great approach, 954 00:59:52.039 --> 00:59:55.199 I think. And the final thing I want to shout 955 00:59:55.239 --> 01:00:02.519 out is that I recently countered this video. This guy 956 01:00:02.679 --> 01:00:06.920 does videos about old hardware, and he released a video 957 01:00:07.000 --> 01:00:14.239 about a personal computer from the early eighties really called 958 01:00:14.280 --> 01:00:18.320 the Ti ninety nine four A. And the reason that 959 01:00:18.639 --> 01:00:22.639 this hit close to my heart is that that was 960 01:00:22.800 --> 01:00:26.920 my first personal computer when I was a kid. It 961 01:00:27.159 --> 01:00:32.519 was a very weird and wacky machine. And he literally 962 01:00:32.880 --> 01:00:36.320 kind of opened the screws and showed the insight of 963 01:00:36.400 --> 01:00:39.559 how that machine worked, to explain why it was so 964 01:00:39.679 --> 01:00:42.639 weird and wacky, and why it worked the way that 965 01:00:42.719 --> 01:00:44.719 it did, and what was good about it, and a 966 01:00:44.719 --> 01:00:47.440 lot of things that weren't so good about it. And 967 01:00:48.760 --> 01:00:52.119 you know, it brought back a lot of memories. So 968 01:00:53.039 --> 01:00:56.920 I'll share that link to that video. Let's just say 969 01:00:57.079 --> 01:01:01.119 that it was really interesting to probe for a system 970 01:01:01.559 --> 01:01:03.519 that had all of four. 971 01:01:03.400 --> 01:01:04.559 K of RAM. 972 01:01:04.880 --> 01:01:08.679 You know, in these days of gigabytes, you know, this 973 01:01:08.760 --> 01:01:09.599 is all we had. 974 01:01:11.920 --> 01:01:12.760 Yeah, that's it. 975 01:01:12.840 --> 01:01:16.800 Four kfram And within that four kf RAM, I implemented 976 01:01:16.840 --> 01:01:20.599 a Donkey Kong like game that had two screens. And 977 01:01:20.639 --> 01:01:24.880 I was like thirteen at the time, so yeah, it. 978 01:01:24.679 --> 01:01:26.760 Was which language did you use? 979 01:01:26.960 --> 01:01:34.760 I'm curious, basic, because that's what it had. It was. 980 01:01:36.760 --> 01:01:42.079 I joked that it simultaneously got me into programming and 981 01:01:42.400 --> 01:01:44.559 kind of also scarred me for life. 982 01:01:45.679 --> 01:01:47.119 So my first. 983 01:01:46.880 --> 01:01:48.800 Job in tech was doing tech support, and we were 984 01:01:48.840 --> 01:01:51.119 dealing with Win just three point one with our software, 985 01:01:51.159 --> 01:01:54.639 and I was constantly having to help users with memory 986 01:01:54.679 --> 01:01:57.679 issues and upper memory and lower memory and shifting things 987 01:01:57.679 --> 01:02:00.480 around so that things could run. And and then when 988 01:02:00.519 --> 01:02:02.000 I was ninety five came along and a lot of 989 01:02:02.000 --> 01:02:02.679 that went bye bye. 990 01:02:02.760 --> 01:02:05.920 I have a whole story about that, but we don't 991 01:02:05.920 --> 01:02:08.599 have the time. We could probably do an entire episode 992 01:02:08.639 --> 01:02:12.440 on reminiscing. So those would be my picture today. 993 01:02:13.119 --> 01:02:18.920 All right, Johanni, or to you, Yeah, I've recently found 994 01:02:18.920 --> 01:02:25.679 myself I'm interested in quality coffee, So I realized in 995 01:02:25.719 --> 01:02:28.159 mostly in the coffee story, So I really I always 996 01:02:28.199 --> 01:02:31.039 thought that like an expensive coffee is like expensive wine, 997 01:02:31.079 --> 01:02:34.079 you just pay for you know, for that brand or 998 01:02:34.079 --> 01:02:37.360 for that weird test. But then I when I started 999 01:02:37.400 --> 01:02:39.960 reading more and more and see videos about testing, I 1000 01:02:40.000 --> 01:02:43.880 realized that like between the farm and your cap, they 1001 01:02:43.920 --> 01:02:49.559 are like ten mistakes, like like dozens of mistakes. Most 1002 01:02:49.599 --> 01:02:54.480 of them are intentional, intentional to just to just cut 1003 01:02:54.559 --> 01:03:00.039 coast and make the coffee horrible. Take for example, in 1004 01:03:00.079 --> 01:03:04.960 any commodity farms, when they when they pick the cherry, 1005 01:03:05.480 --> 01:03:08.960 when they when they collect the cherry picks, some of 1006 01:03:09.000 --> 01:03:13.639 them are already just mature and some are unriped, some 1007 01:03:13.679 --> 01:03:16.519 are red, some are green. Uh so, but you know 1008 01:03:16.599 --> 01:03:18.880 they have to be efficient, so so just take the 1009 01:03:19.079 --> 01:03:21.960 entire tree down. You get coffee bins that are right 1010 01:03:22.119 --> 01:03:26.199 unripe together, some are really sour, some are really really sweet. 1011 01:03:26.559 --> 01:03:30.880 And and that's just one thing that the coffee coffee 1012 01:03:30.920 --> 01:03:34.559 production is doing in order to bring your coffee with 1013 01:03:34.719 --> 01:03:39.800 lower costs and more revenues for them, and really getting 1014 01:03:39.800 --> 01:03:43.360 a great cup of coffee demands a lot of a 1015 01:03:43.360 --> 01:03:49.320 lot of uh decisions and optimization all over the all 1016 01:03:49.360 --> 01:03:54.360 over that process, and it's it's there's no point here, 1017 01:03:54.559 --> 01:04:01.440 just a very fascinating journey that involves biology, nature people. Yeah. 1018 01:04:01.519 --> 01:04:04.440 Other than that, I realized that we are getting more 1019 01:04:04.440 --> 01:04:08.320 and more dependent on lms, right, I mean we're about 1020 01:04:08.320 --> 01:04:10.280 to use it for ten percent off coding or maybe 1021 01:04:10.360 --> 01:04:16.199 ninety percent. I mean it's objective, and I felt like 1022 01:04:16.280 --> 01:04:18.360 it's too much of a black box for me. I 1023 01:04:18.400 --> 01:04:21.199 can't understand the mechanic them. So say, for example, I'm 1024 01:04:21.199 --> 01:04:25.760 providing it with fifteen instructions, it's respecting thirty. Why it 1025 01:04:26.119 --> 01:04:28.400 fits inside the context within the Why did you ignore 1026 01:04:28.400 --> 01:04:31.320 twenty of them? So I bought a book which is 1027 01:04:31.360 --> 01:04:36.559 called I wanted to understand this box from inside. I 1028 01:04:36.599 --> 01:04:39.159 bought a book that's called Building an LM from Scratch. 1029 01:04:39.639 --> 01:04:44.000 It's just teach you build all the layers of an LM. 1030 01:04:44.039 --> 01:04:46.519 Of course a very simplistic one, but you get idea 1031 01:04:46.639 --> 01:04:50.239 of all the moving pieces. I'm now reading it, coding it, 1032 01:04:51.760 --> 01:04:56.280 and I think that I, after processing four chapters, I 1033 01:04:56.320 --> 01:04:57.679 can surely recommend it. 1034 01:04:59.800 --> 01:05:01.760 Right, to go back to your first pick, I would. 1035 01:05:01.840 --> 01:05:04.239 All I will say is that the term quality copy 1036 01:05:04.280 --> 01:05:09.840 is an oxymoron, so there is no such thing, and 1037 01:05:09.880 --> 01:05:12.760 I'll leave bit at that. So all right, Well, thank you, 1038 01:05:12.800 --> 01:05:15.400 Johanni for joining us today and talk to you about testing. 1039 01:05:15.480 --> 01:05:18.519 I appreciate it you coming here before we go. 1040 01:05:19.119 --> 01:05:20.840 If people want to get hold of you and give 1041 01:05:20.840 --> 01:05:24.360 you money and and do all that kind of stuff, 1042 01:05:24.400 --> 01:05:26.880 where is the best place to track your genius? 1043 01:05:29.000 --> 01:05:32.679 Thanks for mentioning this. Well, my gietime page has a 1044 01:05:32.800 --> 01:05:35.239 contacts There is also goldberg Oni dot com. 1045 01:05:36.639 --> 01:05:39.960 And you said that you work as a consultant, so 1046 01:05:40.039 --> 01:05:44.559 I assume that you give assistance training and help organizations 1047 01:05:44.599 --> 01:05:48.039 improve their testing and test coverage and quality of tests 1048 01:05:48.039 --> 01:05:48.599 and whatnot. 1049 01:05:49.599 --> 01:05:56.639 Yes, workshops ends on, ends on, and holding any any 1050 01:05:56.679 --> 01:05:59.280 any kind of assistant and training. 1051 01:06:00.760 --> 01:06:01.079 And that's it. 1052 01:06:01.199 --> 01:06:07.880 Joannie Goldberg Goldberg Goldberg, Yohnnie excuse me? Alrighty and Yonie 1053 01:06:07.880 --> 01:06:09.719 is y o n I just. 1054 01:06:10.239 --> 01:06:15.039 Yeah or gita dot com slash Goldberg Yonnie. 1055 01:06:15.519 --> 01:06:19.920 Okay, alrighty, Well that is it. Another thrilling episode is 1056 01:06:19.920 --> 01:06:21.800 in the books. Thank you for joining us and we 1057 01:06:21.880 --> 01:06:24.280 will talk to you next time on job script Jabber