1 00:00:00,040 --> 00:00:02,920 Architecture is, as we say, it's on the cost of being rejected 2 00:00:02,960 --> 00:00:05,800 wholesale by the industry. It has been a tough ten years 3 00:00:05,800 --> 00:00:07,520 for the concept of enterprise architecture. 4 00:00:07,880 --> 00:00:10,520 We have a bunch of really old ideas that seem a little bit 5 00:00:10,520 --> 00:00:12,840 outdated, and our developers are telling us that they're very 6 00:00:12,840 --> 00:00:14,680 outdated and they don't want to work like this anymore, with 7 00:00:14,680 --> 00:00:16,960 lines and boxes, with schedules and repositories full of 8 00:00:16,960 --> 00:00:18,880 diagrams. At one point they said to me to 9 00:00:18,880 --> 00:00:20,080 just teach people to do what you do. 10 00:00:20,200 --> 00:00:22,080 And I had to answer them and say, you know, I don't know what 11 00:00:22,080 --> 00:00:23,160 I do. None of us know what we do. 12 00:00:23,480 --> 00:00:25,160 So I asked a whole bunch of architects, the senior 13 00:00:25,160 --> 00:00:26,640 architects that I work with, what do you do? 14 00:00:26,680 --> 00:00:29,040 We don't know either. So many people don't have the 15 00:00:29,040 --> 00:00:30,640 same definition, and sometimes they also don't know how to 16 00:00:30,640 --> 00:00:32,280 describe it. But in your definition, what is 17 00:00:32,280 --> 00:00:35,360 your definition of architecture? My definition of architecture is 18 00:00:35,400 --> 00:00:37,280 decision making in the face of ignorance. 19 00:00:37,680 --> 00:00:39,760 As architects, we've got to work with concepts like customer and 20 00:00:39,760 --> 00:00:42,240 order and markets and demand and all of this stuff that none of 21 00:00:42,240 --> 00:00:44,200 it stays still. The only thing we focused on is 22 00:00:44,200 --> 00:00:46,200 logic. And so we try to logically 23 00:00:46,200 --> 00:00:49,200 describe how the business works. We try to logically make our 24 00:00:49,200 --> 00:00:51,440 stakeholders talk about requirements in a way that 25 00:00:51,480 --> 00:00:53,640 satisfies our needs as logicians. 26 00:00:54,000 --> 00:00:57,120 And it completely collapses at that point because you cannot 27 00:00:57,160 --> 00:01:00,360 grasp the the complexity of a social organization with logic. 28 00:01:00,360 --> 00:01:02,520 And this causes frustration for our stakeholders. 29 00:01:02,560 --> 00:01:04,599 They don't like it. We have this obsession with 30 00:01:04,599 --> 00:01:06,120 these abstractions, these essences. 31 00:01:06,440 --> 00:01:09,000 What we don't have sometimes is the humility that these 32 00:01:09,000 --> 00:01:12,200 abstractions die very quickly. The problem we have is that to 33 00:01:12,200 --> 00:01:13,760 build software, you need it to be still. 34 00:01:13,760 --> 00:01:16,880 I taught this to 600 people last year, and it turns out that they 35 00:01:16,880 --> 00:01:18,880 loved it. There's a real hunger out there 36 00:01:18,960 --> 00:01:22,440 for new thinking, new ideas, because everyone sees the gap, 37 00:01:22,480 --> 00:01:24,200 or there's a gap between programming and architecture. 38 00:01:24,320 --> 00:01:26,080 And the goal of a programmer is correctness. 39 00:01:26,120 --> 00:01:27,320 We have to write code that's correct. 40 00:01:27,320 --> 00:01:43,840 For an architect, the goal is. Hello everyone, welcome back to 41 00:01:43,840 --> 00:01:46,160 the new episode of the Packly Journal Podcast. 42 00:01:46,160 --> 00:01:48,080 Today. I'm very excited to meet Barry 43 00:01:48,080 --> 00:01:50,600 O'Reilly. He's kind of like someone who is 44 00:01:50,640 --> 00:01:52,920 into architecture, a lot of experience. 45 00:01:53,080 --> 00:01:54,720 Barry is also partly a researcher. 46 00:01:54,720 --> 00:01:56,960 So Barry, welcome to the show and thank you for this 47 00:01:56,960 --> 00:01:59,080 opportunity. Thank you, Henry. 48 00:01:59,640 --> 00:02:02,640 Good to be here. So Barry, before we start diving 49 00:02:02,640 --> 00:02:04,760 into architecture and all that, maybe if you can share a little 50 00:02:04,760 --> 00:02:08,000 bit about yourself to telling us about your curriculum points 51 00:02:08,000 --> 00:02:09,520 that you think we all can learn from you. 52 00:02:10,479 --> 00:02:15,400 Yeah, so I, I've been in the business for a very long time. 53 00:02:15,600 --> 00:02:18,120 I was 26 years or something like that. 54 00:02:18,480 --> 00:02:21,760 I'm originally a pure mathematician who graduated at 55 00:02:21,760 --> 00:02:25,560 the end of the last century. So it was finance or computing, 56 00:02:25,960 --> 00:02:30,120 and luckily I chose computing. I've worked very early on in my 57 00:02:30,120 --> 00:02:33,840 career, started to work as an architect just by luck, which is 58 00:02:33,840 --> 00:02:36,080 a big part of an architectural career. 59 00:02:36,080 --> 00:02:40,320 So in one of the very early projects I worked in, the lead 60 00:02:40,320 --> 00:02:43,640 architects needed to take some time off and I had to take over 61 00:02:43,920 --> 00:02:46,000 mostly within the Microsoft stack. 62 00:02:46,080 --> 00:02:49,640 And then several years later I went to to work for Microsoft, 63 00:02:50,160 --> 00:02:53,720 was eventually a chief architect on the European level, did a lot 64 00:02:53,720 --> 00:02:58,080 of cool stuff with IoT and large distributed systems in the 65 00:02:58,080 --> 00:03:01,880 cloud. And when I was there, I led the 66 00:03:01,880 --> 00:03:05,240 global architecture solution architecture community. 67 00:03:05,240 --> 00:03:08,600 So I spent a lot of time teaching and thinking, and what 68 00:03:08,600 --> 00:03:10,600 is architecture? How do we move this to the next 69 00:03:10,600 --> 00:03:12,760 level? And, and most importantly, how 70 00:03:12,760 --> 00:03:16,160 do we adapt architecture for this world of cloud which we've 71 00:03:16,160 --> 00:03:17,800 created? But we don't really have the 72 00:03:17,800 --> 00:03:20,360 answer to those questions. If you go back 15 years, what 73 00:03:20,760 --> 00:03:25,280 does it mean to be an architect in the cloud and nobody knew? 74 00:03:25,880 --> 00:03:28,480 And so a real turning point in the work. 75 00:03:28,480 --> 00:03:30,640 I do know I was asking the question, you know, how do we 76 00:03:30,640 --> 00:03:32,880 create architects? What do we teach them? 77 00:03:32,920 --> 00:03:36,040 What do they need to know? And we have a bunch of really 78 00:03:36,040 --> 00:03:38,960 old ideas that seem a little bit outdated. 79 00:03:39,120 --> 00:03:41,960 And our developers are telling us that they're very outdated 80 00:03:41,960 --> 00:03:44,160 and they don't want to work like this anymore with lines and 81 00:03:44,160 --> 00:03:47,200 boxes, with schedules and repositories full of diagrams. 82 00:03:47,760 --> 00:03:50,320 And that was a real turning point, I think, for me. 83 00:03:50,400 --> 00:03:54,000 I started to ask the question, what is this architecture thing? 84 00:03:54,000 --> 00:03:56,840 What is it that I do? At one point they said to me to 85 00:03:56,840 --> 00:03:58,440 just teach people to do what you do. 86 00:03:58,520 --> 00:04:01,040 And they had to answer them and say, you know, I don't know what 87 00:04:01,040 --> 00:04:02,440 I do. None of us know what we do. 88 00:04:02,880 --> 00:04:05,000 So I asked a whole bunch of architects, the senior 89 00:04:05,000 --> 00:04:06,680 architects that I work with, what do you do? 90 00:04:06,720 --> 00:04:09,960 I know we don't know either. So I had to do a lot of digging 91 00:04:10,040 --> 00:04:13,080 to figure that out. And I started to realize that 92 00:04:13,080 --> 00:04:15,520 the thing that made the difference between us, the 93 00:04:15,720 --> 00:04:19,160 senior architects who were actually doing well, delivering, 94 00:04:19,440 --> 00:04:21,560 seem to know what they were doing, even if they couldn't 95 00:04:21,560 --> 00:04:23,280 describe it. That the difference was 96 00:04:23,280 --> 00:04:27,520 uncertainty, that we had a much easier time dealing with 97 00:04:28,040 --> 00:04:30,960 situations, business situations where no one really knew what 98 00:04:30,960 --> 00:04:33,360 was happening and, and stakeholders would come and say, 99 00:04:33,360 --> 00:04:35,360 look, we, we don't really have requirements. 100 00:04:35,360 --> 00:04:36,960 We don't know what we want. We're not even sure what this 101 00:04:36,960 --> 00:04:39,480 problem is. And this would cause in that 102 00:04:39,480 --> 00:04:42,200 case, a lot of developers would say, Hey, we're not doing 103 00:04:42,200 --> 00:04:43,400 anything. You, you haven't told us what 104 00:04:43,400 --> 00:04:44,480 you want. You haven't told us what to 105 00:04:44,480 --> 00:04:45,680 build. You don't know what your, your 106 00:04:45,680 --> 00:04:47,880 features you want or what problem you're trying to solve. 107 00:04:47,880 --> 00:04:50,600 So you'll go away and come back when you know what you want. 108 00:04:50,600 --> 00:04:52,840 Whereas the architect will step in and say, right, we can still 109 00:04:52,840 --> 00:04:54,720 do something here. We can do something with this, 110 00:04:55,280 --> 00:04:58,960 this information. And that took me into a, a 111 00:04:58,960 --> 00:05:01,680 discussion about what is uncertainty and how do we deal 112 00:05:01,680 --> 00:05:03,640 with it? And is there a scientific basis 113 00:05:03,640 --> 00:05:07,080 for dealing with uncertainty? As you step into that as a 114 00:05:07,080 --> 00:05:10,480 mathematician, you immediately turn to probability and base and 115 00:05:10,480 --> 00:05:12,240 all of that. And that doesn't work. 116 00:05:12,240 --> 00:05:13,960 None of it works. If, if that worked, we'd have 117 00:05:13,960 --> 00:05:16,600 figured it out decades ago. And you end up realizing that 118 00:05:16,600 --> 00:05:18,160 there's something else going on here. 119 00:05:18,360 --> 00:05:22,000 And that took me into a completely different discipline, 120 00:05:22,000 --> 00:05:25,360 which is the the complexity sciences, which, you know, has a 121 00:05:25,360 --> 00:05:28,040 very strong relationship to mathematics. 122 00:05:28,040 --> 00:05:30,720 So it was, it was an easy place to start looking. 123 00:05:31,120 --> 00:05:33,680 And what I started to do was I started to read up and those 124 00:05:33,680 --> 00:05:36,520 ideas, I started to pull them in and describe my own practice in 125 00:05:36,520 --> 00:05:38,920 those terms. And that caught the attention of 126 00:05:38,920 --> 00:05:40,000 a lot of people. A lot of people thought, hey, 127 00:05:40,000 --> 00:05:41,800 this is cool. Can you talk more about this? 128 00:05:42,240 --> 00:05:45,360 I was sort of nervous about it because I didn't want to get up 129 00:05:45,360 --> 00:05:48,800 on a stage and say, hey, look, do this and everything will be 130 00:05:48,800 --> 00:05:50,040 fine. Because I thought that that 131 00:05:50,040 --> 00:05:53,960 doesn't feel right. In our industry, it's perfectly 132 00:05:53,960 --> 00:05:56,520 acceptable to get up on a stage and say, do this and everything 133 00:05:56,520 --> 00:05:58,240 will be fine. People will just nod and they 134 00:05:58,240 --> 00:05:59,760 believe you and they do whatever you say. 135 00:06:00,160 --> 00:06:02,080 But I wanted to take it a bit further than that. 136 00:06:02,080 --> 00:06:05,240 So I talked to a whole bunch of people, eventually went back to 137 00:06:05,240 --> 00:06:07,320 university. And when I went back to, you 138 00:06:07,320 --> 00:06:10,320 know, I went and I decided to do a PhD and I have these ideas and 139 00:06:10,880 --> 00:06:14,160 I want proof, either proof that they work or proof that they 140 00:06:14,160 --> 00:06:16,080 don't work. And if they don't work, then 141 00:06:16,080 --> 00:06:17,480 why? Why don't they work? 142 00:06:17,480 --> 00:06:20,560 And I developed a bunch of ideas around how we should be doing 143 00:06:20,560 --> 00:06:23,360 architecture, which we'll talk about, I'm sure about the ideas 144 00:06:23,360 --> 00:06:27,120 like random simulation and stuff that I kind of borrowed and 145 00:06:27,120 --> 00:06:30,040 reformed from the complexity sciences to suit our world, 146 00:06:30,040 --> 00:06:33,040 which it turns out is very, very different than the world of 147 00:06:33,040 --> 00:06:36,760 other engineers. And I just wanted to see, is 148 00:06:36,760 --> 00:06:40,080 this real or is it just a bunch of things that sound good? 149 00:06:40,200 --> 00:06:41,960 Because there's an awful lot of things in our industry that 150 00:06:41,960 --> 00:06:44,360 sound good. And the ideas about architecture 151 00:06:44,360 --> 00:06:47,200 that I and the rest of the industry were in the process of 152 00:06:47,200 --> 00:06:50,360 abandoning at that point, they all sounded good at one point, 153 00:06:50,360 --> 00:06:53,240 but no one done any work to prove that they actually worked. 154 00:06:54,040 --> 00:06:56,600 I didn't want to just do another loop of the hamster wheel. 155 00:06:57,000 --> 00:07:01,240 So I I decided to go back and prove that this actually held or 156 00:07:01,240 --> 00:07:02,840 didn't. And that's the point I've come 157 00:07:02,840 --> 00:07:06,360 to now. That journey has been fantastic 158 00:07:06,360 --> 00:07:08,760 and I've done it out loud The whole time I've been working, 159 00:07:08,760 --> 00:07:11,000 I've been getting up on stages at conferences to say, hey, 160 00:07:11,000 --> 00:07:13,480 look, this seems to be the right direction. 161 00:07:13,480 --> 00:07:16,160 What do you think of this? And I've taught a lot of 162 00:07:16,160 --> 00:07:18,040 architects along the way how to do this. 163 00:07:18,280 --> 00:07:20,720 The result of that has been in the last year, I've released two 164 00:07:20,720 --> 00:07:24,840 books which are kind of side quests from my PhD, which I 165 00:07:24,840 --> 00:07:27,600 should have finished by now, but I keep writing books instead of 166 00:07:27,600 --> 00:07:31,400 the actual dissertation. And that's brought us up to the 167 00:07:31,400 --> 00:07:34,680 present day where these ideas are kind of starting to really 168 00:07:34,960 --> 00:07:37,600 get noticed in the industry, and a lot of people are excited 169 00:07:37,600 --> 00:07:39,800 about them. And so it's a pretty cool time 170 00:07:39,880 --> 00:07:42,520 right now. This episode is brought to you 171 00:07:42,520 --> 00:07:47,240 by Swim IO, and I'm excited to have its CTO and cofounder Omer 172 00:07:47,240 --> 00:07:50,320 Rosenbaum with me today to tell you more about Swim. 173 00:07:50,440 --> 00:07:53,040 Hi Henry, very nice to meet you and thank you for having me. 174 00:07:53,480 --> 00:07:56,160 So tell us a little bit more, what is swim.com? 175 00:07:56,320 --> 00:08:00,040 At Swim, we want to help companies understand their code 176 00:08:00,040 --> 00:08:03,000 bases. We combine static code analysis 177 00:08:03,040 --> 00:08:06,960 with generative AI to create comprehensive documents that 178 00:08:06,960 --> 00:08:10,120 help you navigate the code base. As an engineer myself, I 179 00:08:10,120 --> 00:08:14,080 wouldn't want engineers to spend so much time understanding 180 00:08:14,080 --> 00:08:16,480 existing code. I would want them to spend time 181 00:08:16,680 --> 00:08:20,280 creating and building new stuff. When you have code that has 182 00:08:20,280 --> 00:08:24,920 accumulated over decades, and especially in legacy languages 183 00:08:25,280 --> 00:08:30,640 that not many people are adapted nowadays, then the problem is 184 00:08:30,640 --> 00:08:34,080 even bigger. Swim dot IO is specializing into 185 00:08:34,080 --> 00:08:36,600 helping mainframe developers to understand their code base. 186 00:08:36,720 --> 00:08:39,600 Why mainframes? We actually didn't start there. 187 00:08:39,720 --> 00:08:45,320 Kabul had been by some people obsolete for a few years, and I 188 00:08:45,320 --> 00:08:48,720 discovered that it's not really obsolete, not at all. 189 00:08:48,720 --> 00:08:52,840 There are more than 800 billion lines of cobalt code that are in 190 00:08:52,840 --> 00:08:55,680 production and they drive lots of the business in the world. 191 00:08:56,240 --> 00:08:59,160 And we got more and more requests. 192 00:08:59,320 --> 00:09:02,200 From. Customers to help them 193 00:09:02,200 --> 00:09:05,560 understand the legacy code basis that was written decades ago and 194 00:09:05,680 --> 00:09:08,320 get accumulated over a very long period of time. 195 00:09:08,520 --> 00:09:11,880 So from your customers so far, what are the some of the success 196 00:09:11,880 --> 00:09:15,320 stories that you can share? So we worked with an analyst who 197 00:09:15,560 --> 00:09:19,960 shared with us that it took them a year to document a single 198 00:09:19,960 --> 00:09:23,200 mainframe application, and using SWIM they were able to document 199 00:09:23,200 --> 00:09:25,800 a similar application in a matter of hours. 200 00:09:26,160 --> 00:09:30,160 So saving that amount of time enables them to focus on other 201 00:09:30,360 --> 00:09:33,280 tasks. Thanks, Amir for sharing with us 202 00:09:33,280 --> 00:09:36,040 about SWIM today. To learn more about SWIM, check 203 00:09:36,040 --> 00:09:38,080 out their website at swim dot IO. 204 00:09:39,640 --> 00:09:41,200 Thank you for sharing your story, right? 205 00:09:41,200 --> 00:09:43,840 So in preparation of this conversation, actually, I did 206 00:09:43,840 --> 00:09:46,240 have a look at your previous talks, right, in the 207 00:09:46,240 --> 00:09:48,080 conferences. I also read your book. 208 00:09:48,280 --> 00:09:51,280 I must say that it is really fascinating reading this new 209 00:09:51,280 --> 00:09:53,160 paradigm, I would say about architecture, right? 210 00:09:53,480 --> 00:09:55,560 And I think it's quite interesting if you just follow 211 00:09:55,560 --> 00:09:57,000 the way you explain things, right? 212 00:09:57,000 --> 00:10:00,000 I find it quite novel and looking forward to actually have 213 00:10:00,000 --> 00:10:01,600 a good discussion with you today. 214 00:10:02,080 --> 00:10:05,440 And you said that you published 2 books, one is the Architect's 215 00:10:05,440 --> 00:10:08,600 paradox and the other one is the residual theory, right? 216 00:10:08,920 --> 00:10:11,080 So maybe let's start with the Architect's paradox. 217 00:10:11,240 --> 00:10:13,240 When you mention about paradox, what do you mean? 218 00:10:13,360 --> 00:10:17,000 Can you explain a little bit? Yeah, so the architect's paradox 219 00:10:17,000 --> 00:10:20,160 is this unfortunate trap that was built for us. 220 00:10:20,760 --> 00:10:23,600 To be honest, we built it ourselves, but we didn't realize 221 00:10:23,600 --> 00:10:26,480 it was there. And the paradox is this, that in 222 00:10:26,480 --> 00:10:30,760 order to invent a computer and to invent software, you have to 223 00:10:30,760 --> 00:10:34,640 think in a very, very rigid, logical way. 224 00:10:34,880 --> 00:10:37,760 And you have to be really consistent in that logic. 225 00:10:37,760 --> 00:10:39,400 You have to really know what you're doing. 226 00:10:39,400 --> 00:10:44,520 So you get people like Alan Turing, von Neumann, Alfonso 227 00:10:44,520 --> 00:10:46,320 Church, right? They really knew their logic. 228 00:10:46,320 --> 00:10:48,960 They were really focused on it, and they had to be to create 229 00:10:48,960 --> 00:10:51,560 software. And when you really focus in on 230 00:10:51,560 --> 00:10:55,520 logic, you accidentally adapt a way of looking at the world, 231 00:10:55,520 --> 00:10:57,680 what you try to see logic in everything. 232 00:10:58,040 --> 00:11:01,040 We are working with computers and chips and software and 233 00:11:01,040 --> 00:11:03,560 registers and all of these things and functions, and we're 234 00:11:03,560 --> 00:11:07,200 thinking pure logic. And then eventually the machines 235 00:11:07,200 --> 00:11:10,840 get so powerful that we realized we can solve a whole bunch of 236 00:11:10,840 --> 00:11:14,080 human problems and business problems with these systems. 237 00:11:14,360 --> 00:11:17,520 And you look out into the real world, away from the machine, 238 00:11:18,080 --> 00:11:20,560 and you see that there's people and they're doing stuff and 239 00:11:20,560 --> 00:11:21,600 they're trying to get stuff done. 240 00:11:21,600 --> 00:11:24,560 And they have all these things called emotions and they have 241 00:11:24,560 --> 00:11:26,560 social groups, they're trying to achieve something. 242 00:11:26,560 --> 00:11:28,240 There's markets, everything's moving. 243 00:11:28,480 --> 00:11:30,720 And we look at this and we try to analyse it. 244 00:11:30,760 --> 00:11:34,040 And the only tool we have, the only thing we focused on is 245 00:11:34,040 --> 00:11:36,680 logic. And so we try to logically 246 00:11:36,680 --> 00:11:40,600 describe how the business works. We try to logically make our 247 00:11:40,600 --> 00:11:43,400 stakeholders talk about requirements in a way that's 248 00:11:43,400 --> 00:11:47,160 consistent, in a way that satisfies our needs as 249 00:11:47,160 --> 00:11:50,440 logicians. And it completely collapses at 250 00:11:50,440 --> 00:11:54,920 that point because you cannot grasp the complexity of a social 251 00:11:54,920 --> 00:11:57,520 organization with logic. You can only reduce it. 252 00:11:57,520 --> 00:12:00,240 You can only make it smaller. And everything that we've done 253 00:12:00,240 --> 00:12:03,200 in the history of our industry, from requirements engineering to 254 00:12:03,200 --> 00:12:06,320 user stories, to the mapping and modelling of domains, it's all 255 00:12:06,320 --> 00:12:09,920 been about reducing things to a point where we can access them 256 00:12:09,920 --> 00:12:13,160 with logic. And this causes frustration for 257 00:12:13,160 --> 00:12:15,000 our stakeholders. They don't like it. 258 00:12:15,120 --> 00:12:17,560 So that's why they run away when they see us in the corridor. 259 00:12:17,560 --> 00:12:19,440 They're, Oh no, I'm going to be forced into another one of those 260 00:12:19,440 --> 00:12:21,800 difficult conversations where they make me feel stupid because 261 00:12:21,800 --> 00:12:24,440 I can't answer questions in a logical way. 262 00:12:24,760 --> 00:12:27,120 And it limits our ability to accept things. 263 00:12:27,400 --> 00:12:30,280 And it means that what happens is that we have the logic down. 264 00:12:30,280 --> 00:12:32,480 We can build the machines, we can make the machine do exactly 265 00:12:32,480 --> 00:12:36,080 what we wanted to do, but we can't translate between the real 266 00:12:36,080 --> 00:12:39,680 business world and the machine because we're applying logic in 267 00:12:39,680 --> 00:12:42,680 the wrong place. And this was a problem. 268 00:12:42,880 --> 00:12:45,520 This had been a problem since the beginning, well, since 269 00:12:45,520 --> 00:12:48,720 computers got powerful enough to start being used in complex 270 00:12:48,720 --> 00:12:52,040 human systems like large businesses, things like that. 271 00:12:52,200 --> 00:12:55,880 And our approach to this has always been to force our 272 00:12:55,880 --> 00:12:59,480 stakeholders to force that complexity to be still and say, 273 00:12:59,480 --> 00:13:02,520 look, no, don't talk, just tell us requirements. 274 00:13:02,800 --> 00:13:04,320 Those requirements have to be smart. 275 00:13:04,320 --> 00:13:06,080 They have to be atomic. They have to be sharp. 276 00:13:06,080 --> 00:13:07,880 They have to not contradict each other. 277 00:13:07,920 --> 00:13:10,720 If you describe your world in a way that follows the laws of our 278 00:13:10,720 --> 00:13:12,760 world, then we can make your computer work for you. 279 00:13:12,760 --> 00:13:16,080 It keeps collapsing. And we know from the statistics, 280 00:13:16,080 --> 00:13:19,000 from the standard chaos report and all of those other reports 281 00:13:19,000 --> 00:13:20,760 that we're not doing this particularly well. 282 00:13:20,760 --> 00:13:23,040 And we knew since the late 60s that we weren't doing this 283 00:13:23,440 --> 00:13:25,800 particularly well. But no one has ever raised the 284 00:13:25,800 --> 00:13:28,360 question before, hey, is it the way we think? 285 00:13:28,840 --> 00:13:31,760 Is it the way we think? That's the problem Now the agile 286 00:13:31,760 --> 00:13:35,920 movement to its credits 25 years ago started making some noise 287 00:13:35,920 --> 00:13:38,680 and said, hey, you know, this rigid, stiff way of looking at 288 00:13:38,680 --> 00:13:41,440 the world is hurting us more than it's helping us and we need 289 00:13:41,440 --> 00:13:43,920 to get rid of this. They got rid of all the 290 00:13:43,920 --> 00:13:47,560 immediate aspects of this, like the steering, the the idea that 291 00:13:47,560 --> 00:13:49,600 there's a waterfall process that's repeatable. 292 00:13:49,600 --> 00:13:51,720 That's all part of the architects paradox. 293 00:13:52,200 --> 00:13:55,200 What the agile movement missed was the fact that when we build 294 00:13:55,200 --> 00:13:58,800 conceptual models of the business, the same paradoxes 295 00:13:58,800 --> 00:14:02,160 there, we're still using logic. So the the problem with the 296 00:14:02,160 --> 00:14:04,960 agile movement was that it just wasn't extreme enough. 297 00:14:05,040 --> 00:14:07,520 It should have thrown the baby out with the bathwater, but it 298 00:14:07,520 --> 00:14:10,000 didn't. It tried to keep the baby. 299 00:14:10,320 --> 00:14:13,520 The baby got upset. And that's where we we are 300 00:14:13,640 --> 00:14:16,400 today. It, it's a very, very difficult 301 00:14:16,400 --> 00:14:19,240 thing to separate ourselves. For some it's a difficult 302 00:14:19,240 --> 00:14:22,840 conversation to have because we're so entrenched in this 303 00:14:22,840 --> 00:14:26,240 logical way of thinking about the world that to move away from 304 00:14:26,240 --> 00:14:29,320 it, to think a different thought seems impossible. 305 00:14:29,760 --> 00:14:32,880 One of the bits of feedback I got in the book was from a very 306 00:14:32,880 --> 00:14:35,680 senior developer in the UK who was one of the sort of pre 307 00:14:35,680 --> 00:14:38,920 readers and he said this is great stuff, but the risk is 308 00:14:38,920 --> 00:14:42,000 that you're kicking away the crutches from developers. 309 00:14:42,000 --> 00:14:44,680 They're going to feel like they have nothing left, nothing to 310 00:14:44,680 --> 00:14:47,360 hang on to, and that's kind of dangerous. 311 00:14:47,360 --> 00:14:50,560 It can make people feel a little bit nervous and so I guess 312 00:14:50,560 --> 00:14:53,560 that's the price of progress is that we will have to feel a 313 00:14:53,560 --> 00:14:56,760 little bit nervous, a little bit out to sea while we have these 314 00:14:56,760 --> 00:14:58,720 discussions. But I think it's important 315 00:14:58,960 --> 00:15:02,680 because architecture is, as we say, it's on the cusp of being 316 00:15:02,680 --> 00:15:04,840 rejected wholesale by the industry. 317 00:15:04,840 --> 00:15:08,560 It has been a tough ten years for the concept of enterprise 318 00:15:08,560 --> 00:15:11,240 architecture. There needs to be a new we have 319 00:15:11,240 --> 00:15:14,640 a new way of thinking around these things and that means sort 320 00:15:14,640 --> 00:15:17,840 of looking inwards and maybe reflecting a little bit, facing 321 00:15:17,840 --> 00:15:20,240 up to some hard truths. Yeah. 322 00:15:20,560 --> 00:15:23,840 So the way you explain that, I think people here who listen, 323 00:15:23,840 --> 00:15:26,600 right understand the paradox where you're coming from, right. 324 00:15:26,600 --> 00:15:29,760 So I think most of the software developers, we are trained to 325 00:15:29,760 --> 00:15:32,360 think in logic, right? Even like for example, given a 326 00:15:32,360 --> 00:15:34,880 certain input, it always comes out with the with the same 327 00:15:34,880 --> 00:15:36,160 output, right? Those kind of stuff. 328 00:15:36,440 --> 00:15:39,400 But we know that every time we build software, there's always 329 00:15:39,560 --> 00:15:43,640 something that make the software fragile, not easy to evolve, 330 00:15:43,640 --> 00:15:46,520 right? And for that reason, right, many 331 00:15:46,520 --> 00:15:49,320 software projects actually statistically considered fail 332 00:15:49,720 --> 00:15:52,680 either they cannot be delivered on time, the scope keeps 333 00:15:52,680 --> 00:15:54,520 increasing, right, and all things like that. 334 00:15:54,800 --> 00:15:57,000 So I think it's really interesting the way you brought 335 00:15:57,000 --> 00:16:01,200 it up as a paradox and I think before we go into more details 336 00:16:01,200 --> 00:16:04,520 about the inner things, right, maybe in your definition, what 337 00:16:04,520 --> 00:16:06,880 is architecture? Because when you mention in the 338 00:16:06,880 --> 00:16:09,760 beginning you asked so many architects, what do you do? 339 00:16:10,000 --> 00:16:12,120 So many people don't have the same definition. 340 00:16:12,120 --> 00:16:14,320 I guess sometimes they also don't know how to describe it. 341 00:16:14,680 --> 00:16:18,440 But in your definition, maybe if you can, what is your definition 342 00:16:18,440 --> 00:16:21,880 of architecture? So my definition of architecture 343 00:16:21,880 --> 00:16:24,800 is decision making in the face of ignorance. 344 00:16:25,360 --> 00:16:28,920 And for us, what what that means is that as an architect, you 345 00:16:28,920 --> 00:16:31,600 know, you're expected to make decisions about the structure of 346 00:16:31,600 --> 00:16:34,480 a software application. That's a definition in itself 347 00:16:35,000 --> 00:16:37,960 that some maybe older enterprise architects don't agree with 348 00:16:37,960 --> 00:16:39,720 because they think that they should move away from the 349 00:16:39,720 --> 00:16:41,800 technology and just deal with business problems. 350 00:16:41,800 --> 00:16:44,520 And I don't share that view of architecture at all. 351 00:16:45,080 --> 00:16:47,400 Architecture needs to be very close to the technology. 352 00:16:47,520 --> 00:16:50,880 And so an architect is a very advanced programmer in in my 353 00:16:50,880 --> 00:16:53,560 world, what we have to do is make decisions about the 354 00:16:53,560 --> 00:16:56,360 structure of the software, but we don't know what's going to 355 00:16:56,360 --> 00:16:59,040 happen in the future. And the function of that 356 00:16:59,040 --> 00:17:02,640 structure is to make sure that the application survives into an 357 00:17:02,640 --> 00:17:04,720 unknown future. So we have to make decisions 358 00:17:04,720 --> 00:17:07,000 about things that we don't yet know about. 359 00:17:07,319 --> 00:17:09,960 And that is what architecture is. 360 00:17:09,960 --> 00:17:12,680 And that's what leads us into a lot of problems. 361 00:17:13,119 --> 00:17:14,680 So that how do you solve that problem? 362 00:17:14,760 --> 00:17:17,079 Well, one way to solve that problem is to get better at 363 00:17:17,079 --> 00:17:20,040 knowing what the future is. And that leads to a whole lot of 364 00:17:20,079 --> 00:17:22,280 snake oil, to be honest. People saying, I can tell you 365 00:17:22,280 --> 00:17:23,960 the future, I know what's going to happen. 366 00:17:23,960 --> 00:17:27,440 I've seen this before. Or it can lead to some people 367 00:17:27,440 --> 00:17:29,880 giving up and saying that there is no point in architecture 368 00:17:29,880 --> 00:17:31,840 because we don't know what we're going to happen in the future. 369 00:17:31,840 --> 00:17:34,400 So let's just wait for the future to happen and then RIP 370 00:17:34,400 --> 00:17:37,240 everything down and rebuild it. And that historically hasn't 371 00:17:37,240 --> 00:17:39,720 given us good architecture that gives us bad architecture. 372 00:17:40,480 --> 00:17:42,960 It might give us confidence that it doesn't give us good 373 00:17:42,960 --> 00:17:46,080 architecture, good structures. And so a lot of the challenges 374 00:17:46,080 --> 00:17:48,480 in how do we handle this fact that we don't know what's going 375 00:17:48,480 --> 00:17:51,600 to happen in the future and what we've tried to do. 376 00:17:51,680 --> 00:17:54,320 So a bunch of very smart people got together in in a place 377 00:17:54,320 --> 00:17:57,840 called Garmisch in Germany, 1968, these NATO conferences. 378 00:17:57,840 --> 00:17:59,920 And they said, hey, why are we, why do we suck at this? 379 00:17:59,920 --> 00:18:01,320 Why does everything keep going wrong? 380 00:18:01,560 --> 00:18:04,240 And they said it's because we're not good enough what we're doing 381 00:18:04,240 --> 00:18:06,520 right now, which is we're sort of trying to copy other 382 00:18:06,520 --> 00:18:09,280 engineering disciplines because we don't have an identity yet as 383 00:18:09,280 --> 00:18:11,360 software people and we still don't. 384 00:18:11,720 --> 00:18:15,680 And So what we try to do is we try to bring in that engineering 385 00:18:15,680 --> 00:18:19,480 paradigm into software. And every time we field, we 386 00:18:19,480 --> 00:18:21,920 said, well, we must not be good enough at engineering. 387 00:18:21,920 --> 00:18:24,520 So we have to engineer harder. And that means we have to be 388 00:18:24,840 --> 00:18:27,080 more explicit about requirements. 389 00:18:27,080 --> 00:18:31,280 We have to be more accurate with risk predictions and all of 390 00:18:31,280 --> 00:18:33,520 these things. And we're trying to copy or 391 00:18:33,520 --> 00:18:36,200 steal the clothes of of real engineers. 392 00:18:36,720 --> 00:18:39,240 And what we've never really had the conversation, what if what 393 00:18:39,240 --> 00:18:41,560 we're doing isn't really engineering? 394 00:18:41,560 --> 00:18:45,520 What if it's something way, way harder than what engineers have 395 00:18:45,520 --> 00:18:47,520 to do usually? Say, you know, the people who 396 00:18:47,520 --> 00:18:50,960 build cars, the architects who design cars, they have it easy, 397 00:18:50,960 --> 00:18:52,000 right? Because they're working with 398 00:18:52,000 --> 00:18:55,920 aluminium and gravity and rubber and fuel, and the physical 399 00:18:55,920 --> 00:18:58,520 properties of those things don't keep changing, right? 400 00:18:58,520 --> 00:19:01,480 They're stable and static and everything is still. 401 00:19:01,480 --> 00:19:04,080 And they can work safe in the knowledge that aluminium will 402 00:19:04,080 --> 00:19:06,560 behave in 10 years from now in much the same way as it behaves 403 00:19:06,560 --> 00:19:08,800 today. As architects, we've got to work 404 00:19:08,800 --> 00:19:12,400 with concepts like customer and order and markets and demand and 405 00:19:12,400 --> 00:19:14,160 all of this stuff, but none of it stays still. 406 00:19:14,160 --> 00:19:17,200 And so that would be trying to design a car in an environment 407 00:19:17,200 --> 00:19:19,880 with variable gravity and variable properties of aluminium 408 00:19:19,880 --> 00:19:22,920 and different types of fuel. And, and it's much, much more 409 00:19:22,920 --> 00:19:24,880 difficult what we're expected to do. 410 00:19:25,320 --> 00:19:28,320 And when you try to then apply the logic of the paradox to that 411 00:19:28,320 --> 00:19:30,920 environment, it all just gets very, very messy. 412 00:19:31,200 --> 00:19:33,720 And that leads to all kinds of speculation. 413 00:19:33,720 --> 00:19:37,680 But eventually it ends up at this rigid waterfall like thing 414 00:19:37,680 --> 00:19:40,320 that we call enterprise architecture, which eventually 415 00:19:40,320 --> 00:19:43,840 the industry starts to push back on because we we know and we see 416 00:19:43,840 --> 00:19:46,400 and we feel in our projects every day that this approach 417 00:19:46,720 --> 00:19:48,320 doesn't work. It doesn't give us the results 418 00:19:48,320 --> 00:19:50,200 we need. Yeah. 419 00:19:50,480 --> 00:19:52,800 So I think a lot of people also mentioned about this, right? 420 00:19:52,800 --> 00:19:56,160 We borrow the discipline, maybe the practices from other 421 00:19:56,160 --> 00:19:58,680 engineering, right. But most of those engineering, 422 00:19:58,720 --> 00:20:01,680 you know, it's more like a physical thing, be it the civil 423 00:20:01,680 --> 00:20:03,640 engineering machinery and things like that. 424 00:20:03,680 --> 00:20:06,400 But software has its own unique properties, I guess, right. 425 00:20:06,400 --> 00:20:09,120 So one good thing is like it's very easy to start. 426 00:20:09,120 --> 00:20:12,160 You can mold it as and when you like, you can deliver it really 427 00:20:12,160 --> 00:20:13,960 quickly. At the same time, those 428 00:20:13,960 --> 00:20:17,040 properties probably is also the challenge how we deal with them, 429 00:20:17,040 --> 00:20:18,360 right? Because, for example, if you 430 00:20:18,360 --> 00:20:19,880 want to change your evolve your software. 431 00:20:20,040 --> 00:20:21,480 Sometimes it's very tricky, right? 432 00:20:21,480 --> 00:20:25,520 Because the business complexity changes and you mentioned that 433 00:20:25,560 --> 00:20:29,000 also architecture typically don't survive a few things, 434 00:20:29,120 --> 00:20:30,640 right? So you mentioned just now about 435 00:20:30,640 --> 00:20:32,480 uncertainty, right? We don't know what the future 436 00:20:32,480 --> 00:20:34,640 will look like. The business keeps changing. 437 00:20:34,720 --> 00:20:36,440 The concepts may be changing as well. 438 00:20:36,440 --> 00:20:40,200 Like you might have customer today defined as this, but maybe 439 00:20:40,440 --> 00:20:43,560 in the next month or so, the definition might change, new 440 00:20:43,560 --> 00:20:45,640 business line appears and things like that. 441 00:20:45,960 --> 00:20:48,840 And also there are two other properties, which is time and 442 00:20:48,840 --> 00:20:51,720 the change. If you can explain us about 443 00:20:51,720 --> 00:20:55,320 these few things right, Why they make architecture so much more 444 00:20:55,320 --> 00:20:58,200 challenging? Yeah, the the big problem we 445 00:20:58,200 --> 00:21:01,640 have in in architecture and that you'll see as soon as you ask an 446 00:21:01,640 --> 00:21:04,040 architect, show me your architecture, they'll show you 447 00:21:04,280 --> 00:21:07,760 something very static as a set of lines and boxes in a nice 448 00:21:07,760 --> 00:21:10,520 diagram or scribbled on a white board, One of those two. 449 00:21:11,120 --> 00:21:13,160 What you'll see is that it's a snapshot. 450 00:21:13,160 --> 00:21:15,680 It's a picture of the architecture in one particular 451 00:21:15,680 --> 00:21:17,920 moment. And everyone knows, if you've 452 00:21:17,920 --> 00:21:20,360 had more than a few years of experience, you know that this 453 00:21:20,360 --> 00:21:23,200 architecture will not look like this one year from now. 454 00:21:23,240 --> 00:21:25,240 It will not look like this two years from now. 455 00:21:25,240 --> 00:21:27,480 It will definitely not look because five years from now. 456 00:21:27,760 --> 00:21:30,720 That's something that's very curious, isn't it? 457 00:21:30,720 --> 00:21:33,000 Because why don't we have that? Why is that not part of our 458 00:21:33,000 --> 00:21:34,880 diagram? Why is it not part of our plan? 459 00:21:35,600 --> 00:21:38,280 And the reason that time is relevant is because things 460 00:21:38,280 --> 00:21:40,760 change over time. And the reason that our 461 00:21:40,840 --> 00:21:43,400 architectural tools and our diagramming and our way of 462 00:21:43,400 --> 00:21:46,440 thinking doesn't embrace that change over time is because of 463 00:21:46,440 --> 00:21:48,280 the uncertainty. It's because we don't know 464 00:21:48,680 --> 00:21:51,200 what's going to happen. And So what we kind of do is 465 00:21:51,200 --> 00:21:53,320 sweep it under the rug. And so we say, look, I'll give 466 00:21:53,320 --> 00:21:56,640 you a lovely 2 dimensional UML diagram of what's going to 467 00:21:56,640 --> 00:21:58,480 happen when we put it into production. 468 00:21:59,280 --> 00:22:00,760 But we won't talk about the change. 469 00:22:00,760 --> 00:22:03,360 We won't talk about the fact that everything's going to move. 470 00:22:03,720 --> 00:22:06,920 We won't talk about the fact that today we have identified a 471 00:22:06,920 --> 00:22:09,840 particular kind of customer with a particular set of properties. 472 00:22:09,840 --> 00:22:13,240 And in five years we'll have dozens of different kinds of 473 00:22:13,240 --> 00:22:15,520 customer and they'll all have different properties and they'll 474 00:22:15,520 --> 00:22:16,920 all react in different way. They'll be in different 475 00:22:16,920 --> 00:22:19,240 countries with different languages and different ways of 476 00:22:19,240 --> 00:22:22,800 identifying them in different legal frameworks around them. 477 00:22:22,800 --> 00:22:24,960 That makes us need to know other properties about that. 478 00:22:25,640 --> 00:22:30,360 And these things are swept under the rug because it's just too 479 00:22:30,360 --> 00:22:32,560 difficult. It's not that it's too difficult 480 00:22:32,560 --> 00:22:36,880 because if I'm building a business without software, it's 481 00:22:36,880 --> 00:22:38,480 easy for me to think about these things. 482 00:22:38,480 --> 00:22:40,600 It's very easy for business people to say, hey, you, you 483 00:22:40,600 --> 00:22:42,680 know, we have a customer today because we're in this one 484 00:22:42,680 --> 00:22:43,880 country. There's one jurisdiction, 485 00:22:43,880 --> 00:22:46,160 there's one product, but the future is going to look 486 00:22:46,160 --> 00:22:48,480 different. How do we make sure that we're 487 00:22:48,480 --> 00:22:51,240 able to deal with that? And the business is very good at 488 00:22:51,240 --> 00:22:53,840 making sure that we can step this up, We can make sure that 489 00:22:53,840 --> 00:22:56,320 we don't close, that we don't need to be ready for this 490 00:22:56,320 --> 00:22:59,680 massive global multi customer world tomorrow, but we need to 491 00:22:59,680 --> 00:23:01,640 make sure that we don't close the doors to that. 492 00:23:01,840 --> 00:23:03,240 So for business people, it's easy. 493 00:23:03,320 --> 00:23:05,880 At least this is natural thinking for them. 494 00:23:06,120 --> 00:23:08,000 For us, it's very, very difficult because we think so 495 00:23:08,000 --> 00:23:10,480 rigidly and so logically. And we what we do is we close 496 00:23:10,480 --> 00:23:13,160 doors with our software. We say this is the customer 497 00:23:13,160 --> 00:23:15,040 object, this is what it looks like forever. 498 00:23:15,320 --> 00:23:17,720 And if it changes, then we're going to have to RIP everything 499 00:23:17,760 --> 00:23:19,080 out and put it all back together. 500 00:23:19,080 --> 00:23:20,960 And it becomes really, really difficult for us. 501 00:23:21,400 --> 00:23:24,000 And so one of the things that software does for a business is 502 00:23:24,000 --> 00:23:25,520 it actually slows the business down. 503 00:23:25,520 --> 00:23:28,960 It makes it much more rigid that you have to codify what's 504 00:23:28,960 --> 00:23:30,840 actually happening. And that's why we have so much 505 00:23:30,840 --> 00:23:33,640 frustration sometimes between business stakeholders and and 506 00:23:33,640 --> 00:23:38,120 said to people because we can't move at their speed if we think 507 00:23:38,120 --> 00:23:42,040 in this particular. Way, I think that explanation is 508 00:23:42,040 --> 00:23:44,360 quite common, right? Maybe in a start up scale up 509 00:23:44,400 --> 00:23:46,840 those kind of situation, right? You will tend to hear that 510 00:23:46,840 --> 00:23:48,440 software becomes a bottleneck, right? 511 00:23:48,440 --> 00:23:51,800 The team cannot, right? The software cannot cope with 512 00:23:51,800 --> 00:23:53,440 the business changes and things like that. 513 00:23:53,680 --> 00:23:56,160 And I think when you explain about that, right, I kind of 514 00:23:56,160 --> 00:23:59,040 think about the way software engineer or architect solves 515 00:23:59,040 --> 00:24:01,600 this problem is to create modeling and abstraction, right? 516 00:24:01,760 --> 00:24:04,720 We try to distill the requirement that we hear from 517 00:24:04,720 --> 00:24:07,880 the business and try to come up with some models, the boxes, the 518 00:24:07,880 --> 00:24:10,400 diagrams and things like that. And I know in your book you 519 00:24:10,400 --> 00:24:14,360 explain this is probably one of the danger, the risk of trying 520 00:24:14,360 --> 00:24:17,600 to come up with something that is too rigid structure and 521 00:24:17,600 --> 00:24:20,480 things like that. Maybe tell us a little bit why 522 00:24:20,480 --> 00:24:23,720 this kind of paradigm where we want to build more abstraction 523 00:24:23,720 --> 00:24:26,080 and modeling doesn't actually survive? 524 00:24:27,320 --> 00:24:30,440 Yeah, so this is something that comes from mathematics. 525 00:24:30,440 --> 00:24:33,880 And so I came out of university as a naive pure mathematician, 526 00:24:33,880 --> 00:24:38,120 and abstraction, when I studied, your ability to work with 527 00:24:38,120 --> 00:24:41,920 abstraction was praised, rewarded, encouraged. 528 00:24:41,920 --> 00:24:45,320 And when you come into a business, then you're looking at 529 00:24:45,320 --> 00:24:46,840 all the things that are moving around. 530 00:24:46,840 --> 00:24:48,880 You try to put a label on some things. 531 00:24:49,320 --> 00:24:51,480 And so, well, that's a customer, and that's a customer, and 532 00:24:51,480 --> 00:24:54,440 that's a product. And, you know, that's a product. 533 00:24:54,440 --> 00:24:56,840 If we squint a little bit, you know, it might be closer to a 534 00:24:56,840 --> 00:24:59,680 service or, you know, but wait, let's call it a product because 535 00:24:59,680 --> 00:25:02,240 then we only have two words to deal with instead of three. 536 00:25:02,760 --> 00:25:04,840 And abstraction is good. And we want to make the world 537 00:25:04,840 --> 00:25:06,600 smaller and we want to hold it still. 538 00:25:07,120 --> 00:25:10,160 And if you study mathematics or computer science, you'll get 539 00:25:10,160 --> 00:25:12,600 extra points for your ability to build these beautiful 540 00:25:12,680 --> 00:25:15,680 abstractions. And we can fall in love with our 541 00:25:15,680 --> 00:25:17,760 abstractions. And one thing that you'll see 542 00:25:17,760 --> 00:25:20,400 very common with developers and architects across the world is 543 00:25:20,400 --> 00:25:21,920 they come up with an abstraction. 544 00:25:22,360 --> 00:25:24,240 They love it. They fall in love with it, and 545 00:25:24,240 --> 00:25:26,160 they'll fight you if you try to change it. 546 00:25:26,160 --> 00:25:27,760 They're not. No, we will not change. 547 00:25:27,760 --> 00:25:31,440 This is a customer object, or this is what we mean by message. 548 00:25:31,840 --> 00:25:35,600 And what happens is that we we're trying to isolate what is 549 00:25:35,640 --> 00:25:38,080 a customer. And this goes right back to 550 00:25:38,120 --> 00:25:42,280 Plato and Aristotle where they talk about trying to identify 551 00:25:42,280 --> 00:25:45,040 the essence of something. What is the essence of customer? 552 00:25:45,040 --> 00:25:47,680 Can we capture this essence? Can we hold it still? 553 00:25:47,960 --> 00:25:51,400 Plato, of course, he believed that there's this heavenly realm 554 00:25:51,560 --> 00:25:53,360 that exists outside of the earth. 555 00:25:53,360 --> 00:25:57,040 We can't see it, and it contains these perfect forms. 556 00:25:57,160 --> 00:26:00,000 And somewhere in that heavenly realm there is a perfect 557 00:26:00,000 --> 00:26:02,000 customer. And the way to find out what the 558 00:26:02,000 --> 00:26:06,040 perfect customer is is to just think about it rationally and 559 00:26:06,040 --> 00:26:07,640 you will eventually come up with a perfect customer. 560 00:26:07,640 --> 00:26:11,920 When you do that, you're wise. This is rejected of course, by 561 00:26:11,920 --> 00:26:15,400 his own student Aristotle, but we carry that with us. 562 00:26:15,400 --> 00:26:18,320 And when we have this concept within object orientation of the 563 00:26:18,320 --> 00:26:21,240 class which we define, that's that perfect heavenly realm 564 00:26:21,240 --> 00:26:23,680 which doesn't actually exist. That's the perfect customer 565 00:26:23,720 --> 00:26:25,800 customer class. And then you have customer 566 00:26:25,800 --> 00:26:29,440 objects, which are the, the real world copies of that running in 567 00:26:29,440 --> 00:26:31,800 memory. And so this behavior kind of 568 00:26:31,800 --> 00:26:36,320 carries over into and this also leads us to try and, and 569 00:26:36,320 --> 00:26:39,840 conceptualize and abstract over things and make it efficient all 570 00:26:39,840 --> 00:26:42,240 the time. That leads to us thinking things 571 00:26:42,240 --> 00:26:45,800 like security is something that is a property of every project. 572 00:26:46,200 --> 00:26:48,480 And that leads us to thinking that there is some magical 573 00:26:48,480 --> 00:26:51,000 heavenly security. That's the CM in every single 574 00:26:51,000 --> 00:26:53,080 project. And so you end up with an 575 00:26:53,080 --> 00:26:55,880 industry that thinks that security is the CM. 576 00:26:56,000 --> 00:26:59,280 Whether you're building an app to share pictures of your lunch 577 00:26:59,760 --> 00:27:02,840 or an app that runs inside a nuclear power station, they both 578 00:27:02,840 --> 00:27:04,800 have security. It's the same word, it's the 579 00:27:04,800 --> 00:27:06,960 same thing. It's absolutely not the same 580 00:27:06,960 --> 00:27:09,360 thing. And these abstractions are an 581 00:27:09,360 --> 00:27:12,760 attempt to make it easier for us, but they actually end up 582 00:27:12,760 --> 00:27:14,480 making things much, much more difficult. 583 00:27:15,040 --> 00:27:17,880 Human history, or the history of knowledge, tends to be a search 584 00:27:17,880 --> 00:27:20,480 for abstractions where we stumble around getting things 585 00:27:20,480 --> 00:27:23,440 wrong until one day we suddenly find an abstraction like Newton 586 00:27:23,440 --> 00:27:27,080 did, which really works for us, and then we get hung up on it 587 00:27:27,080 --> 00:27:29,400 for a couple of centuries. We haven't really got there in 588 00:27:29,400 --> 00:27:31,680 software. The abstractions that we use are 589 00:27:31,680 --> 00:27:34,800 still they're very dangerous to us in very volatile 590 00:27:34,800 --> 00:27:37,360 environments. There is no solution to this 591 00:27:37,360 --> 00:27:40,400 problem, right, because humans have to think an abstraction. 592 00:27:40,400 --> 00:27:44,360 It's because we can't. You have to have the word cat in 593 00:27:44,360 --> 00:27:47,160 your head. You can't memorize the name of 594 00:27:47,160 --> 00:27:49,960 every single cat that you meet on earth and refer to them as 595 00:27:49,960 --> 00:27:52,000 individuals. You have to be able to say cat, 596 00:27:52,240 --> 00:27:56,000 but you also have to be prepared if you go to the zoo as a kid 597 00:27:56,000 --> 00:27:59,240 for the first time and see a lion, you have to be prepared to 598 00:27:59,240 --> 00:28:02,240 rapidly adjust your idea what is a cat? 599 00:28:02,560 --> 00:28:04,640 Because it can be very, very different all of a sudden. 600 00:28:05,000 --> 00:28:08,200 And in software, we have this obsession with these 601 00:28:08,200 --> 00:28:11,880 abstractions, these essences. What we don't have sometimes is 602 00:28:11,880 --> 00:28:15,760 the humility that these abstractions die very quickly. 603 00:28:15,840 --> 00:28:17,720 And so an abstraction is never still. 604 00:28:17,960 --> 00:28:21,240 The problem we have is that to build software, you need it to 605 00:28:21,240 --> 00:28:22,960 be still. And so we kind of convince 606 00:28:22,960 --> 00:28:25,320 ourselves that this is customer now, we've arrived, this is what 607 00:28:25,320 --> 00:28:27,400 it's going to be. What's needed really is a 608 00:28:27,400 --> 00:28:30,880 humility that this abstraction that I've come up with will not 609 00:28:30,880 --> 00:28:33,800 survive in the wild. Occasionally some abstractions 610 00:28:33,800 --> 00:28:36,640 do survive for a very long time, but we don't know which ones 611 00:28:36,640 --> 00:28:39,560 from the outset. And you know, you see these 612 00:28:39,560 --> 00:28:43,080 problems all the time with these huge kind of ERP installations 613 00:28:43,120 --> 00:28:47,080 from large companies. That's a set of abstractions 614 00:28:47,080 --> 00:28:49,560 from somewhere else. And someone said these 615 00:28:49,560 --> 00:28:52,000 abstractions will definitely work for you and it will cost 616 00:28:52,000 --> 00:28:54,840 this much. And then you buy the platform 617 00:28:54,840 --> 00:28:56,480 and you say what we need to change a few of these 618 00:28:56,480 --> 00:28:59,680 abstractions and suddenly it's a five year project that's 500 619 00:28:59,680 --> 00:29:03,000 million and no one's happy. And that's the constant failure 620 00:29:03,000 --> 00:29:06,480 of the abstraction that leads to this kind of misery within 621 00:29:06,480 --> 00:29:08,880 software architecture. And there's no solution to it. 622 00:29:08,920 --> 00:29:11,720 We have to think like this. We're kind of pre programmed to 623 00:29:11,720 --> 00:29:15,240 think like this, but all that's needed is the realization that 624 00:29:15,240 --> 00:29:19,800 this will crumble constantly and there are some things we can do 625 00:29:19,800 --> 00:29:22,720 to protect ourselves against that inevitable crumbling. 626 00:29:22,720 --> 00:29:26,360 We can't get it to be still and that's the big issue for us 627 00:29:26,360 --> 00:29:28,760 because we've got to make the still static software for this 628 00:29:28,760 --> 00:29:32,080 constantly moving world. So I think I left when you 629 00:29:32,080 --> 00:29:34,440 mentioned about, you know, we tried to think of the perfect 630 00:29:34,440 --> 00:29:36,600 abstraction and we fight with each other, right? 631 00:29:36,840 --> 00:29:39,880 And I think you mentioned also a few famous philosophers, right? 632 00:29:39,880 --> 00:29:43,120 The scientists also, right? And I think it's a good segue to 633 00:29:43,160 --> 00:29:46,240 go into this philosophy of architecture because in your 634 00:29:46,240 --> 00:29:49,480 talks, in your books, right, you mentioned about this study of 635 00:29:49,480 --> 00:29:53,280 philosophy that you did and how to actually apply that into 636 00:29:53,400 --> 00:29:55,000 software engineering and architecture, right? 637 00:29:55,280 --> 00:29:58,600 So why exactly philosophy? And I think you also kind of 638 00:29:58,600 --> 00:30:01,320 explained that we all actually have a philosophy in the way we 639 00:30:01,320 --> 00:30:03,640 build software, but we just don't know about it or 640 00:30:03,640 --> 00:30:06,920 unconsciously know about it. So tell us the importance of 641 00:30:06,920 --> 00:30:09,160 understanding this philosophy. Yeah. 642 00:30:09,200 --> 00:30:12,880 So as part of a PhD, which I've been doing, you have to write a 643 00:30:12,880 --> 00:30:16,520 chapter that says, this is my philosophical view of the world, 644 00:30:16,520 --> 00:30:19,280 this is how I believe the world works. 645 00:30:19,760 --> 00:30:21,960 And you have to do this because you're going to design an 646 00:30:21,960 --> 00:30:23,480 experiment that's going to have a result. 647 00:30:23,480 --> 00:30:25,880 You're going to interpret those results in a particular way, and 648 00:30:26,200 --> 00:30:29,200 it has to be consistent. So you have to interpret those. 649 00:30:29,360 --> 00:30:31,560 You design your experiment, you run your experiment, you 650 00:30:31,560 --> 00:30:34,440 interpret your results all through a philosophical lens. 651 00:30:34,440 --> 00:30:36,840 There's no escape in that. Everyone has to do that. 652 00:30:36,840 --> 00:30:39,400 And so it's very important upfront to say, this is how I 653 00:30:39,400 --> 00:30:41,320 think, this is how I look at the world. 654 00:30:41,960 --> 00:30:45,280 And so I was given this project and Oh my God, what is this 655 00:30:45,280 --> 00:30:48,120 nonsense? And my wife, who's a scholar, 656 00:30:48,120 --> 00:30:51,120 she said, ah, you need to read this and then this and then 657 00:30:51,120 --> 00:30:52,440 this. And I thought, oh, this is 658 00:30:52,480 --> 00:30:55,360 overwhelming. And I got really interested in 659 00:30:55,360 --> 00:30:57,920 it in the end of it, a little bit too interested in it and 660 00:30:57,920 --> 00:31:00,280 went a bit too far. They say, like, this has to be 7 661 00:31:00,280 --> 00:31:02,600 pages. I got to like 200 pages. 662 00:31:02,600 --> 00:31:06,440 I thought maybe I've gone too far again with this idea and I 663 00:31:06,440 --> 00:31:09,760 started to realize that once I started to get what it gave me 664 00:31:09,760 --> 00:31:12,360 was a vocabulary, I started to understand that, hey, a whole 665 00:31:12,360 --> 00:31:15,200 lot of these discussions that we've been having in software 666 00:31:15,360 --> 00:31:18,440 for the the entire length of my career are philosophical in 667 00:31:18,440 --> 00:31:20,000 nature. We think we're on the cutting 668 00:31:20,000 --> 00:31:21,960 edge. We think we're the cool hip kids 669 00:31:21,960 --> 00:31:25,720 doing all this new stuff. But there was 2 old man in a 670 00:31:25,720 --> 00:31:29,560 square in Athens had this exact argument 2 1/2 thousand years 671 00:31:29,560 --> 00:31:31,160 ago. And they were smart enough to 672 00:31:31,160 --> 00:31:34,120 say we'd better just leave it. We we haven't really got there 673 00:31:34,120 --> 00:31:36,720 yet. And I started to realize was 674 00:31:36,720 --> 00:31:41,560 that I had a philosophy and that philosophy wasn't coming from a 675 00:31:41,560 --> 00:31:43,840 study of philosophy. It was coming from my pure 676 00:31:43,840 --> 00:31:45,960 mathematics, my computer science education. 677 00:31:46,400 --> 00:31:48,400 And that philosophy was coming from my professors. 678 00:31:48,400 --> 00:31:50,640 And they didn't know where their philosophy was coming from 679 00:31:50,640 --> 00:31:52,200 either. The vast majority of them didn't 680 00:31:52,200 --> 00:31:54,680 know where they got it from. And they were inheriting their 681 00:31:54,680 --> 00:31:58,400 philosophy from a number of thinkers at the turn of the last 682 00:31:58,400 --> 00:32:01,080 century. So people like Gottlob Figgia in 683 00:32:01,080 --> 00:32:04,480 Germany, logicians, Bertrand Russell, Ludwig Wittgenstein, 684 00:32:05,080 --> 00:32:08,640 people who were very, very focused on logic as being at the 685 00:32:08,640 --> 00:32:11,200 center of everything. And so these philosophers, they 686 00:32:11,200 --> 00:32:15,360 created the platform for Turing and von Neumann to take the 687 00:32:15,360 --> 00:32:19,440 steps they needed to take. And so we have this invisible 688 00:32:19,560 --> 00:32:22,120 philosophy that comes from a bunch of different thinkers, a 689 00:32:22,120 --> 00:32:27,120 bunch of different trends across the millennia that causes us as 690 00:32:27,440 --> 00:32:29,920 the computer scientists to software engineers to think in a 691 00:32:29,920 --> 00:32:32,800 very particular way. And what I started to realize 692 00:32:32,800 --> 00:32:35,480 was that when I talked to engineers and I scraped at the 693 00:32:35,480 --> 00:32:39,520 surface a little bit and asked them some awkward questions, and 694 00:32:39,520 --> 00:32:41,560 the ones who didn't get upset and actually answered the 695 00:32:41,560 --> 00:32:46,320 questions would reveal there's a similar set of ideas in all of 696 00:32:46,320 --> 00:32:48,800 us. We've all inherited the same way 697 00:32:48,800 --> 00:32:51,400 of thinking. This default philosophy that 698 00:32:51,400 --> 00:32:55,480 believes in this heavenly customer and constantly seeks 699 00:32:55,480 --> 00:32:59,720 it, that believes that there's an if A then B relationship 700 00:33:00,000 --> 00:33:03,120 constantly everywhere. That believes at the end of the 701 00:33:03,120 --> 00:33:07,080 day, that an organization of people connected to each other 702 00:33:07,120 --> 00:33:10,920 doing very human things to which we call a business. 703 00:33:11,200 --> 00:33:13,480 That it can be reduced and modelled. 704 00:33:13,480 --> 00:33:15,920 In the same way that we talk about circuit boards, in the 705 00:33:15,920 --> 00:33:18,280 same way that we talk about classes and objects and all of 706 00:33:18,280 --> 00:33:22,280 these things and the way that we act and the way that we behave 707 00:33:22,280 --> 00:33:25,080 around our stakeholders, the way that we form conceptual models 708 00:33:25,080 --> 00:33:29,320 and software is a reflection of that inbuilt philosophy that we 709 00:33:29,320 --> 00:33:32,080 all carry. And I realized that as I was 710 00:33:32,080 --> 00:33:34,880 teaching the ideas in residuality, which are kind of 711 00:33:35,320 --> 00:33:38,760 radical, they ask you to throw out a whole lot of all that ways 712 00:33:38,880 --> 00:33:41,800 thinking. As I was asking people to do 713 00:33:41,800 --> 00:33:44,360 that, I noticed that they would struggle sometimes. 714 00:33:44,360 --> 00:33:46,080 There's things they didn't want to let go off. 715 00:33:46,080 --> 00:33:49,720 And I said this is deeper than just a tool or knowledge of how 716 00:33:49,720 --> 00:33:52,080 something works. This is philosophical. 717 00:33:52,080 --> 00:33:54,920 So I'm asking them to change philosophically how they look at 718 00:33:54,920 --> 00:33:57,000 the world. But the vast majority of 719 00:33:57,000 --> 00:34:00,600 developers can't even describe, and I couldn't how they look at 720 00:34:00,600 --> 00:34:02,800 the world philosophically. So you're asking someone to 721 00:34:02,800 --> 00:34:04,760 change something that they don't know is there. 722 00:34:05,360 --> 00:34:07,400 And that is incredibly difficult. 723 00:34:07,440 --> 00:34:11,480 And so I gave a talk about this because I find this fascinating. 724 00:34:11,480 --> 00:34:14,639 I got in the book, I say like I'm a software engineer who got 725 00:34:14,639 --> 00:34:16,320 lost in the wrong part of the library. 726 00:34:16,760 --> 00:34:20,280 And I spent a couple of years there and really got into these 727 00:34:20,280 --> 00:34:23,719 ideas and these thoughts. And eventually I started to talk 728 00:34:23,719 --> 00:34:25,199 about it. And they talked about it at an 729 00:34:25,199 --> 00:34:28,040 architecture conference. At the architecture conference, 730 00:34:28,040 --> 00:34:29,840 they said they were really interested. 731 00:34:29,920 --> 00:34:32,080 They thought this is really interesting stuff. 732 00:34:32,080 --> 00:34:34,560 And there was a computer science professor who came on directly 733 00:34:34,560 --> 00:34:37,760 after me and he was absolutely shell shocked is actually, I 734 00:34:37,800 --> 00:34:40,520 never heard anything like this. How dare you RIP up the rule 735 00:34:40,520 --> 00:34:43,719 book. And then I went to, I spoken to 736 00:34:43,719 --> 00:34:47,159 developer conference, not a huge developer conference, a small 1. 737 00:34:47,199 --> 00:34:49,520 And I thought, this is a big risk where they're going to 738 00:34:49,520 --> 00:34:51,400 throw chairs at me, They're going to hit this, this is 739 00:34:51,400 --> 00:34:53,600 boring. Because I found that they were 740 00:34:53,600 --> 00:34:56,719 really interested as well. And then they did an even bigger 741 00:34:56,719 --> 00:35:00,080 developer conference, NDC and Oslo last year. 742 00:35:00,480 --> 00:35:02,640 And I thought they said, can you come and give a talk? 743 00:35:02,640 --> 00:35:04,760 I said, yeah, OK, I'll give my philosophy talk. 744 00:35:04,760 --> 00:35:07,600 And I figured they'll put me in a room far at the back. 745 00:35:07,600 --> 00:35:09,960 There'll be 4 mostly older people there. 746 00:35:09,960 --> 00:35:12,480 No one will care. You know, it'll be fun. 747 00:35:13,160 --> 00:35:16,240 I got there and they put me in the big room, the keynote room. 748 00:35:16,360 --> 00:35:18,600 So there's these four people are going to be in. 749 00:35:18,600 --> 00:35:20,240 There's like 400 chairs of poor people. 750 00:35:20,240 --> 00:35:22,680 This is going to be a disaster. And the room filled and there 751 00:35:22,680 --> 00:35:25,120 were people standing at the back and standing in the aisles and I 752 00:35:25,120 --> 00:35:28,400 was absolutely terrified. But this is too many developers 753 00:35:28,400 --> 00:35:30,560 to hear about coming to listen to philosophy. 754 00:35:30,880 --> 00:35:33,880 Most of them are going to be drastically disappointed that 755 00:35:33,880 --> 00:35:35,920 this has got nothing to do with code really. 756 00:35:36,520 --> 00:35:38,520 And it turns out that they they loved it. 757 00:35:38,560 --> 00:35:42,480 And I thought, well, there's a real hunger out there for new 758 00:35:42,480 --> 00:35:45,960 thinking, new ideas, because everyone sees the gap where 759 00:35:45,960 --> 00:35:48,120 there's a gap between programming and architecture. 760 00:35:48,480 --> 00:35:50,520 We know that the old ways don't fill it. 761 00:35:50,960 --> 00:35:52,560 We don't have anything to fill it with. 762 00:35:52,560 --> 00:35:55,920 The only way to fill that gap is to just throw yourself in at the 763 00:35:55,920 --> 00:35:58,640 deep end and suffer for a few years until you've sort of 764 00:35:58,640 --> 00:36:02,000 figured it out. And so there's a real hunger for 765 00:36:02,000 --> 00:36:04,400 developers who want to take that step up to architecture. 766 00:36:04,520 --> 00:36:07,320 How do we do this? And they're very open to the 767 00:36:07,320 --> 00:36:10,720 idea that we just have to kind of change how you think and how 768 00:36:10,720 --> 00:36:13,680 you look at things a little bit and suddenly you can solve new 769 00:36:13,680 --> 00:36:16,040 problems. The, the response of that 770 00:36:16,040 --> 00:36:19,760 conference then gave me the idea that, hey, those 200 extra PhDs 771 00:36:19,760 --> 00:36:21,720 that I have lying around, maybe that should be a book. 772 00:36:21,920 --> 00:36:24,200 And so that's what I've been doing for the last couple of 773 00:36:24,200 --> 00:36:26,480 months instead of the work that I'm actually meant to be doing. 774 00:36:27,080 --> 00:36:30,880 So that that's how I arrived at the fact that philosophy is a 775 00:36:30,880 --> 00:36:35,720 very, very important question for us as software architects 776 00:36:36,240 --> 00:36:39,920 and that it's a very good starting point for someone who 777 00:36:39,920 --> 00:36:43,360 wants to take that journey from, especially if you're a very 778 00:36:43,360 --> 00:36:46,080 senior developer and which I guess is your audience. 779 00:36:46,480 --> 00:36:47,920 And you're saying, what's the next step? 780 00:36:47,920 --> 00:36:49,600 And the next step isn't more algorithms. 781 00:36:49,600 --> 00:36:52,000 It isn't getting better at leaked, it isn't getting better 782 00:36:52,320 --> 00:36:54,840 at Kubernetes. It's a much more inward 783 00:36:54,840 --> 00:36:58,640 reflective journey which will allow you to solve problems that 784 00:36:58,640 --> 00:37:01,720 are actually a little bit outside the machine, but still 785 00:37:01,720 --> 00:37:04,680 very much related to the code on the compiler. 786 00:37:05,520 --> 00:37:07,680 So I think when you mentioned about both philosophy, I think 787 00:37:07,680 --> 00:37:11,360 we all have, it may be inherited from the study, from the books, 788 00:37:11,360 --> 00:37:14,920 right, from the logical thinking that we thought ourselves maybe 789 00:37:14,920 --> 00:37:17,120 as well. I also follow your talks, right? 790 00:37:17,120 --> 00:37:20,800 So I think it's quite refreshing the way you kind of think about 791 00:37:20,880 --> 00:37:22,840 architecture. And I think it's not something 792 00:37:22,840 --> 00:37:25,080 that straightforward naturally we think about, right. 793 00:37:25,080 --> 00:37:28,560 So I think this is also a good discussion in terms of what you 794 00:37:28,560 --> 00:37:31,560 think as the unit of architecture, because many 795 00:37:31,560 --> 00:37:34,720 people, if you ask architecture, they have different answers, 796 00:37:34,720 --> 00:37:36,040 right? They could relate to like 797 00:37:36,040 --> 00:37:38,760 monolith micro service, they could relate to something like 798 00:37:38,760 --> 00:37:41,320 domain driven design, they could relate to something like 799 00:37:41,360 --> 00:37:44,680 evolutionary architecture, event driven architecture, So many 800 00:37:44,680 --> 00:37:45,800 different architectures out there. 801 00:37:46,120 --> 00:37:48,520 But the way you actually describe in your philosophy, 802 00:37:48,520 --> 00:37:51,880 architecture, maybe architecture has a different kind of unit. 803 00:37:52,040 --> 00:37:55,640 So maybe now it's a good time to actually introduce us to this 804 00:37:55,640 --> 00:37:58,520 concept of residuality. Yeah. 805 00:37:58,520 --> 00:38:01,880 So if we go back to the ancient Greeks, which is where I always 806 00:38:01,880 --> 00:38:03,520 start this story, does it make more sense? 807 00:38:03,840 --> 00:38:07,440 They looked at the world. And for them, you know, this is 808 00:38:07,720 --> 00:38:09,920 2 1/2 thousand years ago, there were a lot of mysteries. 809 00:38:09,920 --> 00:38:11,920 There were a lot of things we know today that they didn't 810 00:38:11,920 --> 00:38:14,320 know. And they would look at things 811 00:38:14,800 --> 00:38:16,400 and ask questions. What is it? 812 00:38:16,400 --> 00:38:18,840 What is it made of? And they would look at these 813 00:38:18,840 --> 00:38:21,320 things and they'd say, well, what if it's made of water? 814 00:38:21,440 --> 00:38:24,320 One of the sort of renowned as one of the first Greek 815 00:38:24,320 --> 00:38:28,240 philosophers called Talas, who came up with this idea, it's 816 00:38:28,240 --> 00:38:29,800 what if everything is made of water? 817 00:38:30,360 --> 00:38:32,640 That's obviously ridiculous. So if we look at that now, 818 00:38:32,640 --> 00:38:33,720 everything is not made from water. 819 00:38:33,720 --> 00:38:36,240 What a stupid idea. But it's not the idea. 820 00:38:36,240 --> 00:38:38,720 It's not the conclusion that he came to that's important. 821 00:38:38,720 --> 00:38:42,120 It's the thought that what if we look at the world as if it's all 822 00:38:42,120 --> 00:38:45,560 made of one thing, as if there's some kind of arcade, they called 823 00:38:45,560 --> 00:38:48,200 it, some kind of building block that everything is made of. 824 00:38:48,960 --> 00:38:50,920 And he came up with water and he was obviously wrong. 825 00:38:50,920 --> 00:38:54,400 But that instinct started a way of thinking that changed the 826 00:38:54,400 --> 00:38:57,680 world for us in the West. There were other thinkers. 827 00:38:57,680 --> 00:39:00,360 We will come back to one called Heraclitus, and he said 828 00:39:00,360 --> 00:39:02,600 everything is made of fire, everything's on fire, 829 00:39:02,600 --> 00:39:04,600 everything's moving and changing all the time. 830 00:39:04,800 --> 00:39:07,880 That's very much influenced with ideas from the East, which is 831 00:39:07,880 --> 00:39:10,400 going to turn out to be much more useful to us as architects. 832 00:39:10,920 --> 00:39:14,320 And eventually then they have this one guy, the Democritus, 833 00:39:14,320 --> 00:39:16,280 who comes along and says, what if everything's made of tiny, 834 00:39:16,280 --> 00:39:18,280 tiny little particles called atoms? 835 00:39:18,800 --> 00:39:22,880 And that's an idea that stuck around for thousands of years. 836 00:39:22,880 --> 00:39:24,400 And we've all learned that in high school. 837 00:39:24,840 --> 00:39:28,560 And what we get from that journey is that when we all 838 00:39:28,560 --> 00:39:32,560 assemble and agree temporarily that there is this one thing 839 00:39:32,560 --> 00:39:35,240 that everything else is made out of, it makes a discipline 840 00:39:35,240 --> 00:39:37,160 possible. It makes a science possible. 841 00:39:37,520 --> 00:39:40,640 And so as soon as we come up with the idea of the atom and we 842 00:39:40,640 --> 00:39:42,920 coalesce around that idea, we get physics. 843 00:39:43,120 --> 00:39:44,640 And then you have the idea of the cell. 844 00:39:44,640 --> 00:39:47,880 Suddenly you have biology and the molecule gives us chemistry 845 00:39:48,120 --> 00:39:51,640 and then in software we have this constantly shifting 846 00:39:51,840 --> 00:39:55,040 picture, right O someone says it's the machine, it's the UNCH 847 00:39:55,040 --> 00:39:59,160 card, it's this function, it's the procedure, it's the module, 848 00:39:59,320 --> 00:40:02,960 it's this actor, the service, the object, the the micro 849 00:40:02,960 --> 00:40:04,960 service. So it gets very tiring. 850 00:40:05,200 --> 00:40:08,200 It changes every couple of years and we're so we're very much in 851 00:40:08,200 --> 00:40:11,800 that sort of earth, fire, wind, water, fears of what's 852 00:40:11,800 --> 00:40:13,680 everything made of. We don't know, let's make 853 00:40:13,680 --> 00:40:15,560 something up and see what happens. 854 00:40:15,680 --> 00:40:18,640 And we still haven't really got to agree up to result there. 855 00:40:18,640 --> 00:40:21,040 I mean, the service has been revolutionary for us. 856 00:40:21,040 --> 00:40:22,600 I think it's given us cloud computing. 857 00:40:22,600 --> 00:40:26,360 It's given us network computing. But we're not quite settled yet. 858 00:40:26,360 --> 00:40:29,680 And in terms of architecture, then we, we have to ask the 859 00:40:29,680 --> 00:40:32,320 question, what is the unit of architecture? 860 00:40:32,680 --> 00:40:35,680 And what I found is that all of these units restrict our 861 00:40:35,680 --> 00:40:38,640 conversations about architecture because if you combine it with 862 00:40:38,640 --> 00:40:42,400 this need to conceptualize, what we try to do then is identify 863 00:40:42,400 --> 00:40:46,120 these magic objects in the world, customer order, product, 864 00:40:46,120 --> 00:40:48,440 all of these things. And we immediately try to map it 865 00:40:48,440 --> 00:40:53,200 to some kind of base unit, a service, an object deemisly 866 00:40:53,200 --> 00:40:56,400 which we wasted decades on mapping all of these things. 867 00:40:56,840 --> 00:40:59,760 And for architecture, I find that this can actually limit us, 868 00:40:59,760 --> 00:41:01,360 even if it's very, very necessary. 869 00:41:01,720 --> 00:41:05,840 And So what I managed to do was to create a new kind of unit, 870 00:41:06,040 --> 00:41:10,480 which I call the residue. And the residue isn't a single 871 00:41:10,480 --> 00:41:12,200 thing. It's a little collection of 872 00:41:12,200 --> 00:41:13,760 things. And it's a delta. 873 00:41:13,760 --> 00:41:16,680 It's a little bit different than all of the other residues around 874 00:41:16,680 --> 00:41:19,280 it. So the thing that demarcates is 875 00:41:19,280 --> 00:41:23,280 difference. And these residues are basically 876 00:41:23,280 --> 00:41:25,640 our architecture as it collapses. 877 00:41:26,080 --> 00:41:28,840 And So what we're interested in is describing our architecture 878 00:41:28,840 --> 00:41:33,760 not as this static 2 dimensional picture, but as a series of 879 00:41:33,840 --> 00:41:37,520 collapses. So we assume with humility that 880 00:41:37,520 --> 00:41:40,440 the architecture I've designed, the concepts that I've used, are 881 00:41:40,440 --> 00:41:42,240 going to collapse no matter what happens. 882 00:41:42,280 --> 00:41:44,360 And there's uncertainty as we move into the future. 883 00:41:44,560 --> 00:41:46,840 My architecture collapses when it collapses. 884 00:41:46,840 --> 00:41:49,880 The important question for me as an architect is it's not to go, 885 00:41:49,960 --> 00:41:51,080 ah, dying. I was wrong. 886 00:41:51,080 --> 00:41:52,800 That should have been a better concept. 887 00:41:52,800 --> 00:41:55,040 Let's start again. The question for me as an 888 00:41:55,040 --> 00:41:57,160 architect is when it collapses, what do we do? 889 00:41:57,280 --> 00:41:59,640 How do we get out of the hole? How do we keep running? 890 00:41:59,880 --> 00:42:04,280 And So what we do is we randomly simulate our architecture with 891 00:42:04,280 --> 00:42:06,120 stress and we say make it collapse. 892 00:42:06,160 --> 00:42:09,240 And when it collapses, we observe it and we say, if we'd 893 00:42:09,240 --> 00:42:12,160 added this little thing, if we'd done this from the start, it 894 00:42:12,160 --> 00:42:14,200 wouldn't have collapsed. But we haven't built anything 895 00:42:14,200 --> 00:42:16,320 yet, so let's add it. Noi that we know. 896 00:42:16,320 --> 00:42:19,760 And what we do is we try to provoke constant collapse in our 897 00:42:19,760 --> 00:42:22,640 conceptual model of the world before we commit to an 898 00:42:22,640 --> 00:42:25,280 architecture. And so we take our ideas. 899 00:42:25,280 --> 00:42:28,080 So I have an idea in my head after I speak to a customer for 900 00:42:28,080 --> 00:42:29,680 a few minutes. You have a customer that looks 901 00:42:29,680 --> 00:42:31,040 like this. You have orders that look like 902 00:42:31,040 --> 00:42:32,200 this. You have products that look like 903 00:42:32,200 --> 00:42:35,400 this, you have billing and everything that looks like this. 904 00:42:35,640 --> 00:42:39,040 Let's try and break every single one of those concepts in as many 905 00:42:39,040 --> 00:42:41,600 ways as we can, randomly breaking them. 906 00:42:42,000 --> 00:42:45,120 And as the models collapse, every time the model collapses, 907 00:42:45,120 --> 00:42:47,440 we have to find a way to make sure that that collapse, if it 908 00:42:47,440 --> 00:42:49,960 happened in real life, would be comfortable for us. 909 00:42:50,480 --> 00:42:53,280 And every time we discover something that we can do that 910 00:42:53,280 --> 00:42:56,880 makes that collapse comfortable, we assemble that in a residue 911 00:42:56,880 --> 00:42:59,880 and we set it in a little pile. And at the end of our work, we 912 00:42:59,880 --> 00:43:03,640 have maybe hundreds of these residues and that describe how 913 00:43:03,680 --> 00:43:05,920 our architecture was going to look, how it's going to collapse 914 00:43:05,920 --> 00:43:08,280 in different trouser and eggs in a future that we're not quite 915 00:43:08,280 --> 00:43:11,200 certain about. What this does is because of the 916 00:43:11,440 --> 00:43:14,880 internal mathematics of complex systems and the rigidity of 917 00:43:14,880 --> 00:43:17,760 software. So there's a big gap between the 918 00:43:17,760 --> 00:43:20,480 number of possible states in a software application and the 919 00:43:20,480 --> 00:43:22,640 number of possible states in a human system. 920 00:43:22,640 --> 00:43:25,240 The human system is orders of magnitude, many orders of 921 00:43:25,240 --> 00:43:27,640 magnitude bigger. And we can leverage that 922 00:43:27,720 --> 00:43:31,560 distance between these two things to make sure that because 923 00:43:31,560 --> 00:43:34,720 the human system will keep returning to the same points 924 00:43:34,720 --> 00:43:37,520 over and over again, something called attractors in complexity 925 00:43:37,520 --> 00:43:41,080 science, we can actually manipulate the attractors in our 926 00:43:41,080 --> 00:43:44,360 software through these residues to get a system that we can 927 00:43:44,360 --> 00:43:48,040 prove is more likely going to survive in unknown conditions. 928 00:43:48,440 --> 00:43:50,400 And this is a huge step for architecture. 929 00:43:50,960 --> 00:43:52,640 It's an entirely different way of thinking. 930 00:43:52,640 --> 00:43:54,880 There's a lot of concepts there that probably just flew over the 931 00:43:54,880 --> 00:43:56,960 heads of most people. It's too many concepts. 932 00:43:57,440 --> 00:43:59,720 It's something that you need to take a bit of time, read and 933 00:43:59,720 --> 00:44:02,160 listen to the talks. The ideas that I'm talking about 934 00:44:02,160 --> 00:44:06,600 here are easily as awful as object orientation, right? 935 00:44:06,600 --> 00:44:09,200 And we take a whole university term to suffer through that. 936 00:44:09,200 --> 00:44:11,680 So this is about the same level of intensity. 937 00:44:12,120 --> 00:44:15,520 But what happens is that we can actually mathematically show 938 00:44:15,560 --> 00:44:18,160 that I've designed an architecture more likely to 939 00:44:18,160 --> 00:44:21,880 survive unknown unknowns in the future than what we originally 940 00:44:21,880 --> 00:44:24,520 started with. Just by stressing my model, just 941 00:44:24,520 --> 00:44:27,920 by stressing my concepts through random simulation. 942 00:44:28,280 --> 00:44:30,760 And once we get to the end of that journey, we realize that 943 00:44:30,760 --> 00:44:33,480 what we have is this arcade, this foundational building block 944 00:44:33,480 --> 00:44:35,280 of an architecture. It's called a residue. 945 00:44:35,280 --> 00:44:37,680 It doesn't look like a service or an object or anything like 946 00:44:37,680 --> 00:44:39,640 that. It's more about a related number 947 00:44:39,640 --> 00:44:42,440 of components and how they act in a particular situation. 948 00:44:43,000 --> 00:44:45,120 So it's a collection of little deltas. 949 00:44:45,400 --> 00:44:46,760 And if you read the books, you'll see there's very 950 00:44:46,800 --> 00:44:50,360 philosophical reasons why that all fits together very nicely. 951 00:44:51,040 --> 00:44:53,000 So it's it's an enormous subject. 952 00:44:53,000 --> 00:44:55,040 I think one of the things I've heard is someone said that, you 953 00:44:55,040 --> 00:44:57,880 know, the only way you're going to get these ideas to succeed is 954 00:44:57,880 --> 00:45:02,000 if you can explain it like I'm 5 and I said to them that this is 955 00:45:02,040 --> 00:45:05,040 absolutely not something that five year olds should be doing. 956 00:45:05,120 --> 00:45:07,080 So I don't think that's the right way to go. 957 00:45:07,240 --> 00:45:09,520 A 5 year olds can't do object orientation either. 958 00:45:09,760 --> 00:45:11,640 Yeah. So it's, there's a lot going on 959 00:45:11,640 --> 00:45:12,480 there. I understand that. 960 00:45:12,480 --> 00:45:14,640 It's quite confusing. I've managed to put it together 961 00:45:14,640 --> 00:45:17,360 in the talks on YouTube in a way that's quite logical, which 962 00:45:17,680 --> 00:45:20,160 brings 1 concept in after the other. 963 00:45:20,160 --> 00:45:22,440 I don't think I succeeded in the last three minutes of doing 964 00:45:22,440 --> 00:45:24,960 that, so to be honest. Yeah, definitely. 965 00:45:24,960 --> 00:45:28,240 It's a fascinating thing right there so many concepts and like 966 00:45:28,240 --> 00:45:30,480 embed a lot of philosophical thinking as well. 967 00:45:30,640 --> 00:45:35,320 So I think the way that you explain it in a long talks is 968 00:45:35,320 --> 00:45:38,320 much better definitely, But I think it's still invite us to 969 00:45:38,320 --> 00:45:41,280 actually dig deeper, right? I think the thing that I can 970 00:45:41,280 --> 00:45:43,960 summarize maybe from the way you explain just now, right, instead 971 00:45:43,960 --> 00:45:47,440 of thinking architecture is A1 solution, right, You should 972 00:45:47,440 --> 00:45:50,360 think in terms of multiple kind of options, which you call 973 00:45:50,360 --> 00:45:53,600 residue, right? And each residue is a result of 974 00:45:53,600 --> 00:45:57,040 some things that change in which you mentioned architecture 975 00:45:57,040 --> 00:45:59,440 collapse, right? And yeah, collapse because of a 976 00:45:59,440 --> 00:46:02,520 certain stressor, right? It could be business change, it 977 00:46:02,520 --> 00:46:05,200 could be some kind of cloud computing suddenly exists, 978 00:46:05,200 --> 00:46:07,440 right? So certainly the way you design 979 00:46:07,440 --> 00:46:10,080 software is different. So maybe there are lots more 980 00:46:10,080 --> 00:46:13,040 paradigm that comes. And that's why maybe architect 981 00:46:13,040 --> 00:46:15,560 needs to think in terms of multiple residues, right? 982 00:46:16,000 --> 00:46:19,360 Series of residues such that we can come up with maybe an 983 00:46:19,360 --> 00:46:22,920 evolution or a plan in which if something changes, you can kind 984 00:46:22,920 --> 00:46:25,080 of like pick the residue that you have thought about. 985 00:46:25,160 --> 00:46:27,560 And maybe that architecture is not so bad, right? 986 00:46:27,800 --> 00:46:30,680 It can still survive whatever changes that you have thought 987 00:46:30,680 --> 00:46:33,000 through before. So definitely for people who are 988 00:46:33,000 --> 00:46:36,000 interested in this kind of thinking, I invite you to 989 00:46:36,440 --> 00:46:39,800 Barry's book and also watch maybe the past talks and 990 00:46:39,800 --> 00:46:42,760 definitely hearing about this. What do you think are some 991 00:46:42,760 --> 00:46:45,480 takeaways for us? Definitely learn more about it, 992 00:46:45,480 --> 00:46:47,920 but what are some other practical things that you think 993 00:46:47,920 --> 00:46:50,400 we can do? So there are two books out 994 00:46:50,400 --> 00:46:52,560 there. No one is called Residues, and 995 00:46:52,560 --> 00:46:55,680 that's very much based on how you practically attack an 996 00:46:55,680 --> 00:46:59,280 architecture with these ideas. And that involves stressing your 997 00:46:59,280 --> 00:47:01,280 architecture. So you start off with a naive 998 00:47:01,280 --> 00:47:04,920 architecture, which we all do. So by the time a customer gets 999 00:47:04,920 --> 00:47:07,640 to the end of their second sentence, we've already designed 1000 00:47:07,640 --> 00:47:10,160 something in our heads, and you just start stressing, You start 1001 00:47:10,160 --> 00:47:12,360 asking questions. What if a customer isn't doesn't 1002 00:47:12,360 --> 00:47:14,360 always look like this? What if an order changes? 1003 00:47:14,360 --> 00:47:17,680 What if we have new products? What happens if my assumptions 1004 00:47:17,680 --> 00:47:20,640 about the world are wrong and they're constantly changing? 1005 00:47:21,040 --> 00:47:23,240 And what you naturally get out of that is lots and lots of 1006 00:47:23,240 --> 00:47:25,720 these residues, these architectures that are slightly 1007 00:47:25,720 --> 00:47:28,000 different and you're very, very interested in the high 1008 00:47:28,000 --> 00:47:31,360 architecture collapses. Not so much in what it's made 1009 00:47:31,360 --> 00:47:34,600 of, not so much in what's the perfect way to do this, Not so 1010 00:47:34,600 --> 00:47:38,280 much in solving a particular problem, but how does this thing 1011 00:47:38,280 --> 00:47:40,040 collapse? How does it break? 1012 00:47:40,040 --> 00:47:42,280 Where are the crumple zones that I need to put into this 1013 00:47:42,280 --> 00:47:45,680 architecture to make it stand up, to make it survive over a 1014 00:47:45,680 --> 00:47:48,880 longer period of time on a high, very high level? 1015 00:47:49,200 --> 00:47:51,640 Humility, You're not going to get this right ever. 1016 00:47:51,640 --> 00:47:55,400 It's always going to be wrong. And with that pessimism, OK, 1017 00:47:55,440 --> 00:47:57,120 this is wrong, it's going to go wrong. 1018 00:47:57,120 --> 00:47:59,720 How is it going to go wrong? What ways can I make this thing 1019 00:47:59,720 --> 00:48:02,840 break? And that will allow you to 1020 00:48:03,080 --> 00:48:06,880 formulate and capture what's happening in your head as an 1021 00:48:06,880 --> 00:48:09,000 architect. The really interesting thing 1022 00:48:09,000 --> 00:48:10,760 about these ideas is I've taught this. 1023 00:48:10,760 --> 00:48:13,160 I mean, I taught this to 600 people last year. 1024 00:48:13,760 --> 00:48:16,000 When I talk to really senior architects, people who've been 1025 00:48:16,000 --> 00:48:19,040 around for a while and, you know, have the scars to prove it 1026 00:48:19,040 --> 00:48:22,040 and have built some big stuff, they will say, look, this, what 1027 00:48:22,040 --> 00:48:24,600 you're describing here is exactly what we do. 1028 00:48:24,600 --> 00:48:27,440 This is where you arrive at when you've had a few decades, when 1029 00:48:27,440 --> 00:48:30,240 you've seen a few things. This is how you start to think. 1030 00:48:30,640 --> 00:48:33,680 Now, we all think like this, but we all have different words for 1031 00:48:33,680 --> 00:48:35,920 what we do. And when someone asks what is it 1032 00:48:35,920 --> 00:48:38,560 that you do, exactly the question I got, We can't explain 1033 00:48:38,560 --> 00:48:40,240 it, we don't know, but this is it. 1034 00:48:40,280 --> 00:48:43,080 This is what we do. Is this random stressing that 1035 00:48:43,080 --> 00:48:45,880 doesn't have to come from the mouth of a stakeholder, doesn't 1036 00:48:45,880 --> 00:48:48,240 have to be related to probability or to proper 1037 00:48:48,240 --> 00:48:50,520 engineering. It's more about the ability to 1038 00:48:50,640 --> 00:48:54,160 consistently stress your conceptual model until you feel 1039 00:48:54,160 --> 00:48:57,040 happy with it. And in the case of residuality, 1040 00:48:57,240 --> 00:48:59,760 you stress it until you get to a point where you can't break it 1041 00:48:59,760 --> 00:49:02,800 anymore, and that's when you're ready to start implementing. 1042 00:49:03,480 --> 00:49:06,040 So I, if I get right, you mention it as a criticality, 1043 00:49:06,040 --> 00:49:07,440 right, the point where you think? 1044 00:49:07,520 --> 00:49:10,640 Yeah, yeah. So what what I normally say is 1045 00:49:10,640 --> 00:49:12,920 that the goal of a programmer is correctness. 1046 00:49:12,960 --> 00:49:14,520 We have to write code that's correct. 1047 00:49:14,520 --> 00:49:16,360 It executes the same way every single time. 1048 00:49:16,360 --> 00:49:18,760 It gives us the same results, the same inputs every single 1049 00:49:18,760 --> 00:49:21,520 time. For an architect, the goal is 1050 00:49:21,520 --> 00:49:25,160 criticality, as you say, and criticality is the ability of a 1051 00:49:25,160 --> 00:49:29,560 system to survive things that aren't in specification thinks 1052 00:49:29,560 --> 00:49:32,640 it hasn't been designed for. And I call this off spec 1053 00:49:32,640 --> 00:49:34,600 architecture, which is a big part of our jobs. 1054 00:49:34,600 --> 00:49:37,960 But we have to design a system that survives off spec and 1055 00:49:38,000 --> 00:49:41,680 criticality is appropriate exists in biology and it's a 1056 00:49:41,680 --> 00:49:43,760 function of the number of components and the number of 1057 00:49:43,760 --> 00:49:46,520 links between the components. So if you have a biological 1058 00:49:46,520 --> 00:49:49,760 system, which is very simple, very small number of components 1059 00:49:49,760 --> 00:49:52,440 and links, what we would call a monolith, that system is 1060 00:49:52,440 --> 00:49:55,400 generally not critical because it's considered weak. 1061 00:49:55,400 --> 00:49:58,480 It's only got a few components. So if you are one of them, the 1062 00:49:58,600 --> 00:50:01,800 creature dies in biology or the system crash, or the monolith 1063 00:50:01,800 --> 00:50:05,880 crashes on the other side, you have your microservices, right? 1064 00:50:05,880 --> 00:50:07,920 So you've got a very high number of components, a very high 1065 00:50:07,920 --> 00:50:10,480 number of links. And if you have a biological 1066 00:50:10,480 --> 00:50:13,720 system like that, if you had one, it would be very hard to 1067 00:50:13,720 --> 00:50:17,680 kill it in one go because it has so many pieces and so many 1068 00:50:17,680 --> 00:50:20,440 links. It'll find new ways to survive. 1069 00:50:20,640 --> 00:50:21,960 And, and our brains are like this. 1070 00:50:21,960 --> 00:50:25,080 They have so many components and so many connections that you can 1071 00:50:25,080 --> 00:50:27,520 get a, you know, you can get a bang in the head and a bit of 1072 00:50:27,520 --> 00:50:29,720 your brain stops working and another bit of your brain will 1073 00:50:29,720 --> 00:50:31,280 pick up the slack and cover it for you. 1074 00:50:31,280 --> 00:50:33,360 And that's sort of part of the goal of criticality. 1075 00:50:33,800 --> 00:50:35,720 But the problem is that you can get to a point where you have 1076 00:50:35,720 --> 00:50:38,840 too many components and too many links when the system then 1077 00:50:38,840 --> 00:50:42,480 collapses under the weight of its own management. 1078 00:50:42,520 --> 00:50:45,920 OK, so in biological systems that turns up in one way as 1079 00:50:45,920 --> 00:50:47,720 cancer. There's too many connections, 1080 00:50:47,720 --> 00:50:49,360 too many things, too much can go wrong. 1081 00:50:49,480 --> 00:50:52,360 In another way, as in seeing energy needs. 1082 00:50:52,360 --> 00:50:54,680 It's like you have all these components where it's very hard 1083 00:50:54,680 --> 00:50:56,760 to kill you. You're going to have to eat all 1084 00:50:56,760 --> 00:50:59,400 day every day to keep the energy in that system. 1085 00:50:59,480 --> 00:51:01,200 And it makes the system very, very vulnerable. 1086 00:51:01,720 --> 00:51:04,480 Micro service people have already realized this. 1087 00:51:04,680 --> 00:51:07,880 It's the big problem with micro services is that it is hard to 1088 00:51:07,880 --> 00:51:09,560 kill the entire system at one go. 1089 00:51:09,880 --> 00:51:12,960 But when something goes wrong, it takes like 3 weeks to figure 1090 00:51:12,960 --> 00:51:15,960 out what happened and you've got to sit reading logs in the 1091 00:51:16,000 --> 00:51:18,280 middle of the night and what happened, why did it go wrong? 1092 00:51:18,280 --> 00:51:20,640 We don't know. And the system will collapse 1093 00:51:20,640 --> 00:51:22,360 under the weight of managing itself. 1094 00:51:22,400 --> 00:51:24,760 For us, that's the cost of operations going up. 1095 00:51:25,040 --> 00:51:27,920 So criticality is this magic lying in between the two where 1096 00:51:27,920 --> 00:51:30,640 you have the right balance of components and links and 1097 00:51:30,640 --> 00:51:32,320 residuality. These ideas that we're talking 1098 00:51:32,320 --> 00:51:34,120 about, all this focus on philosophy. 1099 00:51:34,120 --> 00:51:37,160 The point of all of this is to be able to drive our systems to 1100 00:51:37,160 --> 00:51:40,440 that point of criticality where the system has the right balance 1101 00:51:40,440 --> 00:51:44,560 of components and linkage to survive off spec, but not so 1102 00:51:45,040 --> 00:51:47,680 many that it's over engineered that it collapses under the 1103 00:51:47,680 --> 00:51:50,920 weight of its own management. This is a very important thing 1104 00:51:50,920 --> 00:51:54,080 for us because generally the way we do this is we just guess at 1105 00:51:54,080 --> 00:51:55,880 it. And this gives us a sort of a 1106 00:51:55,880 --> 00:51:59,600 very pragmatic and very structured way of getting to 1107 00:51:59,600 --> 00:52:02,320 have we actually achieved criticality in the system or 1108 00:52:02,320 --> 00:52:04,600 have we gone too far? Have we all overshot? 1109 00:52:04,840 --> 00:52:07,440 And so this protects us from 2 very dangerous things. 1110 00:52:07,640 --> 00:52:10,560 The first is over engineering, which we love, and the second is 1111 00:52:10,560 --> 00:52:13,160 under engineering, which we we have a tendency to start 1112 00:52:13,160 --> 00:52:16,560 building systems without really exploring the conceptual models 1113 00:52:17,160 --> 00:52:19,640 that we work with. Thanks for explaining that. 1114 00:52:19,920 --> 00:52:22,360 So what about takeaways for non-technical people? 1115 00:52:22,360 --> 00:52:24,520 Maybe those business stakeholders or people who 1116 00:52:24,520 --> 00:52:26,800 always complain that the software engineering team or the 1117 00:52:26,800 --> 00:52:30,040 architects cannot come up with some solution that is stable and 1118 00:52:30,040 --> 00:52:33,120 can survive over time. Yeah, so this is one of the 1119 00:52:33,120 --> 00:52:36,560 strange side effects of my work. I started talking about these 1120 00:52:36,560 --> 00:52:38,840 things and I thought, well, architects should come and 1121 00:52:38,840 --> 00:52:42,440 listen to me, and then I would go to companies and developers 1122 00:52:42,440 --> 00:52:44,200 were coming as well. And they were really interested 1123 00:52:44,200 --> 00:52:45,960 in what I was saying, though. Well, that's good. 1124 00:52:45,960 --> 00:52:48,120 That's interesting. And as time went on, I started 1125 00:52:48,120 --> 00:52:50,840 to notice agile coaches turning up. 1126 00:52:51,000 --> 00:52:53,360 And I was like, oh, do you know, do you know who I am? 1127 00:52:53,360 --> 00:52:54,440 Do you know what I'm going to talk about? 1128 00:52:54,440 --> 00:52:57,840 This is really not interesting. And then more and more managers 1129 00:52:57,840 --> 00:53:00,280 were turning up, but business people who had nothing to do 1130 00:53:00,280 --> 00:53:03,080 with software started to turn up and listen to my talks. 1131 00:53:03,680 --> 00:53:06,560 I've noticed that these ideas are really popular with them 1132 00:53:06,560 --> 00:53:08,760 because once they start to understand this idea about the 1133 00:53:08,760 --> 00:53:11,680 default philosophy of software architecture, they start to 1134 00:53:11,680 --> 00:53:14,720 understand that uncertainty is at the core of this and that 1135 00:53:14,720 --> 00:53:18,720 requirements and kneeling down all of these old waterfall 1136 00:53:18,720 --> 00:53:22,280 engineering ideas and being more strict isn't the right answer. 1137 00:53:22,280 --> 00:53:25,120 That there needs to be more looseness, essentially. 1138 00:53:25,320 --> 00:53:27,520 And it's a different kind of agility that you start to get 1139 00:53:27,520 --> 00:53:32,600 when you architect like this. It starts to get really easy to 1140 00:53:32,600 --> 00:53:35,280 communicate with architects and with business people. 1141 00:53:35,280 --> 00:53:38,400 And I've actually coached senior executives and this is how you 1142 00:53:38,400 --> 00:53:40,560 manage your architects. This is why they're so weird. 1143 00:53:40,640 --> 00:53:43,120 This is how they think. This is how you talk to them to 1144 00:53:43,120 --> 00:53:45,800 get the most out of them. This is why they get frustrated 1145 00:53:45,800 --> 00:53:49,160 with you a lot of times. And, and a lot of it is about 1146 00:53:49,160 --> 00:53:52,360 getting them to understand this organizational complexity and 1147 00:53:52,360 --> 00:53:55,480 the structure and order of the machines we work with are two so 1148 00:53:55,480 --> 00:53:58,440 wildly, completely different things that they cause all kinds 1149 00:53:58,440 --> 00:54:02,760 of contradictions and problems. And then this leads to a 1150 00:54:02,760 --> 00:54:05,600 rebuilding of bridges between architecture and development, 1151 00:54:05,600 --> 00:54:08,640 between architecture and the executive function, between 1152 00:54:08,640 --> 00:54:11,440 stakeholders and business people, and getting them to 1153 00:54:11,440 --> 00:54:14,640 understand that it's not as simple as it looks. 1154 00:54:14,720 --> 00:54:17,640 These ideas actually trigger a lot of thinking when business 1155 00:54:17,640 --> 00:54:20,920 people realize that what we do as architects is closer to what 1156 00:54:20,920 --> 00:54:25,600 they do in this uncertainty, but that we are kind of limited by 1157 00:54:25,600 --> 00:54:29,440 our educations, by this logical way of looking at the world and 1158 00:54:29,440 --> 00:54:31,600 getting them to understand. That opens up a lot of doors. 1159 00:54:32,480 --> 00:54:34,160 Thank you for introducing this concept. 1160 00:54:34,160 --> 00:54:36,480 I think I can see some encouraging signs, right? 1161 00:54:36,480 --> 00:54:39,880 Where maybe the architect, software engineering team and 1162 00:54:39,880 --> 00:54:42,560 the business stakeholders, they can empathize with each other, 1163 00:54:42,560 --> 00:54:44,200 right? So business stakeholders can 1164 00:54:44,200 --> 00:54:47,880 understand why software sometimes it's not so easy to 1165 00:54:47,920 --> 00:54:50,960 actually evolve and change. And also for software engineers 1166 00:54:50,960 --> 00:54:53,960 and architects to know that why business is complex and keeps 1167 00:54:53,960 --> 00:54:55,720 changing, right? We always complain if, let's 1168 00:54:55,720 --> 00:54:58,760 say, you know, your requirement is not up to what you describe a 1169 00:54:58,960 --> 00:55:00,320 few months ago, for example, right? 1170 00:55:00,360 --> 00:55:01,880 And why we need to change things. 1171 00:55:02,080 --> 00:55:05,120 So I think hopefully there's one day that we can align and kind 1172 00:55:05,120 --> 00:55:07,640 of like empathize each other and we can come up with a much 1173 00:55:07,640 --> 00:55:09,320 better software and architecture altogether. 1174 00:55:09,760 --> 00:55:11,800 So Barry, it's been a pleasant conversation. 1175 00:55:11,920 --> 00:55:14,600 Before I let you go, I have one last question that normally I 1176 00:55:14,600 --> 00:55:17,200 ask for all my guests. I call this the three technical 1177 00:55:17,200 --> 00:55:19,360 leadership wisdom. So think of it like advice that 1178 00:55:19,360 --> 00:55:21,520 you want to give to us. Maybe if you can share your 1179 00:55:21,520 --> 00:55:23,440 version of three technical leadership wisdom. 1180 00:55:24,200 --> 00:55:29,880 OK, so the first one I've stolen from ATV character called Logan 1181 00:55:29,880 --> 00:55:31,400 Roy. I don't know if you've seen the 1182 00:55:31,400 --> 00:55:34,840 TV show Succession. Logan Roy is a billionaire, 1183 00:55:34,880 --> 00:55:39,280 media magnets, a very angry man, but very successful man, very 1184 00:55:39,280 --> 00:55:42,080 shrewd. And he's trying to teach his 1185 00:55:42,080 --> 00:55:44,760 children because he's very old. He's trying to teach his 1186 00:55:44,760 --> 00:55:46,760 children how to take over his business. 1187 00:55:46,960 --> 00:55:49,520 It doesn't go very well. It takes Four Seasons, and they 1188 00:55:49,520 --> 00:55:52,240 all hate each other. And he says in one scene that 1189 00:55:52,240 --> 00:55:55,200 really stuck in my head, that had me screaming, yes, at the 1190 00:55:55,200 --> 00:55:56,720 TV, he's talking to his daughter. 1191 00:55:56,720 --> 00:55:58,880 He's trying to teach her how to live in this world where 1192 00:55:58,880 --> 00:56:02,480 everything is chaos. And he said this is the rule. 1193 00:56:02,520 --> 00:56:08,560 You know, Everything Everywhere is moving, always and forever. 1194 00:56:09,040 --> 00:56:12,680 And that is the first piece of wisdom that every architect 1195 00:56:12,680 --> 00:56:15,640 needs to grasp immediately. And that's the difference 1196 00:56:15,640 --> 00:56:17,280 between architecture and programming. 1197 00:56:17,680 --> 00:56:21,480 Everything, everywhere is moving, always and forever. 1198 00:56:21,680 --> 00:56:24,320 If you learn that, then, and if you understand that as an 1199 00:56:24,320 --> 00:56:26,960 architect, everything gets easier. 1200 00:56:27,280 --> 00:56:30,600 And as you see in the book, the tragedy is that we waste half 1201 00:56:30,600 --> 00:56:33,200 our careers trying to do what every other architect has ever 1202 00:56:33,200 --> 00:56:36,480 done and hold it all still. If you try to hold it still, it 1203 00:56:36,480 --> 00:56:40,080 will eventually wash you away. The second thing in order to be 1204 00:56:40,080 --> 00:56:44,000 able to succeed with that is to advise everyone who wants to be 1205 00:56:44,000 --> 00:56:48,120 an architect to learn how to and to practice thinking for 1206 00:56:48,120 --> 00:56:50,320 yourself. Thinking for yourself is 1207 00:56:50,320 --> 00:56:53,840 probably the most important skill you can have as an 1208 00:56:53,840 --> 00:56:56,600 architect. You cannot just think like 1209 00:56:56,600 --> 00:56:59,200 everyone else because you'll end up looking at problems that you 1210 00:56:59,200 --> 00:57:02,120 have to solve that are unique to your context, to your customer, 1211 00:57:02,120 --> 00:57:04,360 to your environment, and you'll try to solve it with someone 1212 00:57:04,360 --> 00:57:06,920 else's thinking. Lifting and shifting someone 1213 00:57:06,920 --> 00:57:10,440 else's ideas is not thinking. It's not architecture. 1214 00:57:10,800 --> 00:57:13,320 And we get in this industry because there's so much 1215 00:57:13,320 --> 00:57:15,920 insecurity. We get really trapped easily in 1216 00:57:15,920 --> 00:57:18,440 dogma. We want to believe in something, 1217 00:57:18,440 --> 00:57:20,880 we need to believe in something, and we believe in these things a 1218 00:57:20,880 --> 00:57:23,160 little bit too much, a little bit too often. 1219 00:57:23,760 --> 00:57:27,640 There's an excellent book called Lateral Thinking by Edward de 1220 00:57:27,640 --> 00:57:32,000 Bono that teaches you how to really think outside of dogma, 1221 00:57:32,000 --> 00:57:35,200 outside of narrow confines, that I recommend to all architects. 1222 00:57:35,600 --> 00:57:39,160 And the third thing, and this is probably the most important, is 1223 00:57:39,160 --> 00:57:43,040 that it has to be fun. You have to think this is fun. 1224 00:57:43,360 --> 00:57:48,280 If you come into architecture as a developer and you're upset 1225 00:57:48,360 --> 00:57:50,680 because everything keeps moving and everything keeps changing, 1226 00:57:50,680 --> 00:57:53,200 it won't stay stale and you can't develop your perfect 1227 00:57:53,200 --> 00:57:56,280 customer object and you're blaming everyone else and you 1228 00:57:56,280 --> 00:57:58,680 make your stakeholders feel like we know the reason that you 1229 00:57:58,680 --> 00:58:00,600 can't do this is because you're bad at this. 1230 00:58:00,960 --> 00:58:03,280 If it's not fun, then don't do it. 1231 00:58:03,560 --> 00:58:06,120 It's architecture is really, really difficult. 1232 00:58:06,480 --> 00:58:09,520 And there's a certain kind of person who's possibly a little 1233 00:58:09,520 --> 00:58:12,480 bit twisted like me, who thinks that I'd love it whenever 1234 00:58:12,480 --> 00:58:14,840 everything's moving and changing and it's crazy and it won't 1235 00:58:14,840 --> 00:58:17,680 stand still and nothing works and, and it's frustrating. 1236 00:58:17,680 --> 00:58:21,120 I actually think that's fun. And I've noticed that the people 1237 00:58:21,120 --> 00:58:23,400 who've been really, really successful as architects in 1238 00:58:23,400 --> 00:58:26,680 their careers, they also think it's fun and they get bored. 1239 00:58:26,680 --> 00:58:29,360 If everything's easy and predictable and it's a crud up, 1240 00:58:29,360 --> 00:58:32,480 you're like, why am I here? You know, we want it to be 1241 00:58:32,480 --> 00:58:34,840 messy. We want it to constantly move. 1242 00:58:34,840 --> 00:58:38,560 And it's really important that you find architecture fun. 1243 00:58:38,760 --> 00:58:42,840 And if it's not fun, then being an engineer is a perfectly 1244 00:58:42,840 --> 00:58:46,680 acceptable career, and a rewarding career is good work. 1245 00:58:46,840 --> 00:58:49,640 But trying to make architecture like engineering is where all 1246 00:58:49,640 --> 00:58:53,960 the suffering begins. So having fun is probably the 1247 00:58:53,960 --> 00:58:56,720 most important thing that any of us can do in our careers, I 1248 00:58:56,800 --> 00:58:59,240 think. Wow, lovely wisdom, right? 1249 00:58:59,240 --> 00:59:01,560 So definitely, everything is moving, everything is chaos, 1250 00:59:01,560 --> 00:59:03,640 right? So think laterally, right? 1251 00:59:03,640 --> 00:59:05,760 So think outside the box, think for yourself. 1252 00:59:06,000 --> 00:59:08,000 And the last thing is, you know, make it fun, right? 1253 00:59:08,000 --> 00:59:10,720 So you don't want to suffer because of all the changes that 1254 00:59:10,720 --> 00:59:12,720 is happening without you controlling it. 1255 00:59:13,200 --> 00:59:15,960 So Barry, for people who love this conversation, they want to 1256 00:59:15,960 --> 00:59:18,400 reach out to you, connect with you or ask you more questions. 1257 00:59:18,400 --> 00:59:20,400 Is there a place where they can find you online? 1258 00:59:21,040 --> 00:59:24,520 Yes, I'm I'm very active on LinkedIn, which I think a lot of 1259 00:59:24,520 --> 00:59:28,080 people might have seen already. I do all of my thinking outlawed 1260 00:59:28,160 --> 00:59:32,000 for better or worse, mostly on LinkedIn these days because it's 1261 00:59:32,000 --> 00:59:35,080 become the most stable social media platform for doing these 1262 00:59:35,080 --> 00:59:38,520 kind of things. I'm also on blue sky, but I'm 1263 00:59:38,520 --> 00:59:41,880 more often LinkedIn allows more sort of long form conversations 1264 00:59:41,880 --> 00:59:43,520 and articles and things like that. 1265 00:59:44,000 --> 00:59:46,720 I'm sure you'll share the links. There are books and there are 1266 00:59:46,720 --> 00:59:50,680 videos that go into detail on the kind of technical details on 1267 00:59:50,720 --> 00:59:52,320 residues. Now those things work. 1268 00:59:52,680 --> 00:59:57,040 I'm on stage constantly all over the place, so there'll be an 1269 00:59:57,040 --> 01:00:00,440 opportunity to come and see me, talk and talk to me if anyone 1270 01:00:00,440 --> 01:00:04,400 wants to do that across most of the globe across the next year 1271 01:00:04,400 --> 01:00:06,520 or so. Yeah, very nice. 1272 01:00:06,800 --> 01:00:09,320 So I think if people are interested in what we just 1273 01:00:09,320 --> 01:00:12,760 discussed, I highly recommend various books and also talks 1274 01:00:12,760 --> 01:00:15,280 right in the conferences. If you can check out here's 1275 01:00:15,280 --> 01:00:17,920 presentation live, definitely highly recommend that. 1276 01:00:18,240 --> 01:00:19,840 So Barry, thank you so much for your time. 1277 01:00:20,120 --> 01:00:22,880 I'm sure we all learned something new about architecture 1278 01:00:22,880 --> 01:00:24,160 today. Thank you.