1 00:00:01,520 --> 00:00:03,520 Join us as we gather around the hedge, 2 00:00:03,520 --> 00:00:06,500 where we dig into technology, business, and culture 3 00:00:06,559 --> 00:00:08,980 with the finest minds in computer networking. 4 00:00:20,734 --> 00:00:23,454 Well, hello, Tom. How are you? Hi, Russ. 5 00:00:23,454 --> 00:00:24,914 I'm doing great. How are you doing? 6 00:00:25,695 --> 00:00:27,614 Fine. We saw that you were trying to 7 00:00:27,614 --> 00:00:28,754 read all the books. 8 00:00:29,550 --> 00:00:31,869 Yeah. I decided to podcast. Try to cram 9 00:00:31,869 --> 00:00:33,149 in this morning and read everything on my 10 00:00:33,149 --> 00:00:34,590 shelf so you couldn't harass me, but I 11 00:00:34,590 --> 00:00:35,729 failed. So 12 00:00:38,189 --> 00:00:40,189 Could you just find something else? I would 13 00:00:40,189 --> 00:00:41,649 find something else. No worries. 14 00:00:43,710 --> 00:00:44,689 So I was saying 15 00:00:45,685 --> 00:00:47,204 on no. It wasn't a this it wasn't 16 00:00:47,204 --> 00:00:48,804 a podcast. I was saying that I went 17 00:00:48,804 --> 00:00:49,865 to my aunt's house 18 00:00:51,204 --> 00:00:53,524 and over last weekend, whatever, a couple of 19 00:00:53,524 --> 00:00:54,984 weeks ago, and 20 00:00:55,364 --> 00:00:56,965 she was like, I'm getting rid of all 21 00:00:56,965 --> 00:00:59,704 my my husband's books because he passed away. 22 00:01:00,700 --> 00:01:01,520 And so 23 00:01:02,940 --> 00:01:04,700 I went back in the in his shop, 24 00:01:04,700 --> 00:01:06,859 and he did electronics. So I bought I 25 00:01:06,859 --> 00:01:08,859 picked up, didn't buy, picked up, didn't bring 26 00:01:08,859 --> 00:01:11,439 home with me a couple of electronics books 27 00:01:11,819 --> 00:01:14,045 because they're, like, from my childhood and stuff 28 00:01:14,045 --> 00:01:15,724 as well. I mean, these are old, old, 29 00:01:15,724 --> 00:01:16,944 old books on electronics. 30 00:01:17,805 --> 00:01:20,604 Old books if they're from Russ' childhood. Yes. 31 00:01:20,604 --> 00:01:21,984 From my childhood. Yes. 32 00:01:22,525 --> 00:01:24,604 And I realized that actually one of them 33 00:01:24,604 --> 00:01:25,104 is 34 00:01:25,805 --> 00:01:27,984 a book that I had on my shelf 35 00:01:28,579 --> 00:01:31,159 when I was a kid that somehow migrated 36 00:01:32,579 --> 00:01:33,640 to his shelf 37 00:01:33,939 --> 00:01:36,359 and has now migrated back to my shelf. 38 00:01:36,979 --> 00:01:39,299 And another one was had migrated from my 39 00:01:39,299 --> 00:01:41,159 stepfather's shelf to his shelf. 40 00:01:42,064 --> 00:01:44,545 So it's it's just kind of like, wow. 41 00:01:44,545 --> 00:01:45,685 That's that's just 42 00:01:46,064 --> 00:01:46,564 odd. 43 00:01:49,105 --> 00:01:49,605 So 44 00:01:50,385 --> 00:01:51,765 Yvonne is with us. 45 00:01:52,465 --> 00:01:54,245 Yvonne Hello? She shed. 46 00:01:54,625 --> 00:01:56,325 I am. I'm always here. 47 00:01:56,729 --> 00:01:59,129 Does the she shed have a circular door 48 00:01:59,129 --> 00:02:00,030 like a hobbit? 49 00:02:00,969 --> 00:02:03,290 I do not have a hobbit door. No. 50 00:02:03,290 --> 00:02:06,329 That's just really sad. I know. I just 51 00:02:06,329 --> 00:02:08,250 have a regular door, but I do have 52 00:02:08,250 --> 00:02:09,389 a frog lamp. 53 00:02:10,724 --> 00:02:12,985 I do have a plant that's off screen. 54 00:02:13,525 --> 00:02:15,544 You know, I've got my books. 55 00:02:16,084 --> 00:02:17,621 Yeah. But you need a 56 00:02:18,516 --> 00:02:20,324 dolls. You need a and my He Man 57 00:02:20,324 --> 00:02:20,824 toys, 58 00:02:21,284 --> 00:02:22,905 all the relics of my childhood. 59 00:02:23,205 --> 00:02:23,705 Yeah. 60 00:02:25,740 --> 00:02:27,980 Well, so you don't like Skeletor. Is that 61 00:02:27,980 --> 00:02:30,620 it? I don't have a Skeletor. No. Actually, 62 00:02:30,620 --> 00:02:31,520 I just have, 63 00:02:32,379 --> 00:02:32,879 Teela. 64 00:02:33,420 --> 00:02:35,099 If you remember if you watched, you know, 65 00:02:35,099 --> 00:02:37,040 she was the the female character 66 00:02:37,735 --> 00:02:40,235 in He Man because all of those stories, 67 00:02:40,614 --> 00:02:41,914 they at least had one girl. 68 00:02:42,215 --> 00:02:44,074 And so for He Man, it was Selah, 69 00:02:44,134 --> 00:02:45,834 so she's on my shelf back there. 70 00:02:46,854 --> 00:02:48,314 Okay. Well, good. 71 00:02:50,550 --> 00:02:52,469 So let's start today. This is just a 72 00:02:52,469 --> 00:02:53,449 roundtable show 73 00:02:53,750 --> 00:02:56,250 with this blog post from Thinking Sideways, 74 00:02:56,870 --> 00:02:59,370 talking about the software documentation trifecta. 75 00:02:59,669 --> 00:03:02,229 And, you know, I know we all love 76 00:03:02,229 --> 00:03:03,209 to hate documentation. 77 00:03:07,294 --> 00:03:09,455 Some people like it. Some people like it. 78 00:03:09,615 --> 00:03:11,455 Well Depends on if you're using it or 79 00:03:11,455 --> 00:03:13,875 creating it. Exactly. Yes. 80 00:03:14,335 --> 00:03:15,235 Yes. Exactly. 81 00:03:15,615 --> 00:03:17,375 I was gonna say, Tom, if you actually 82 00:03:17,375 --> 00:03:18,675 enjoy creating documentation, 83 00:03:19,569 --> 00:03:20,930 I just don't know if we can even 84 00:03:20,930 --> 00:03:22,930 be friends. Oh, am I not allowed to 85 00:03:22,930 --> 00:03:23,990 admit that publicly? 86 00:03:27,090 --> 00:03:28,389 Oh my goodness. 87 00:03:29,330 --> 00:03:31,250 This should be the tagline on this episode. 88 00:03:31,250 --> 00:03:33,030 It gets real with Tom and Russ. 89 00:03:33,784 --> 00:03:35,705 Russ learns something about Tom that makes him 90 00:03:35,705 --> 00:03:37,245 question everything about it. 91 00:03:39,384 --> 00:03:40,364 That could be. 92 00:03:40,905 --> 00:03:41,405 So 93 00:03:42,104 --> 00:03:44,104 I thought it was interesting, this article talking 94 00:03:44,104 --> 00:03:46,824 about first of all, this this the blog 95 00:03:46,824 --> 00:03:49,490 post starts talking about it's from thinking sideways 96 00:03:49,490 --> 00:03:51,189 for anybody who wants to look it up. 97 00:03:51,650 --> 00:03:53,909 Talks about the three kinds of documentation, 98 00:03:55,169 --> 00:03:55,669 documentation 99 00:03:55,969 --> 00:03:58,150 for developers, customers, and management. 100 00:03:59,485 --> 00:04:01,084 And when we think about document you know, 101 00:04:01,084 --> 00:04:03,084 we can convert that to network engineering. This, 102 00:04:03,084 --> 00:04:04,544 of course, is a coding blog 103 00:04:04,925 --> 00:04:06,705 that we're we're looking at here. 104 00:04:07,245 --> 00:04:08,144 But documentation 105 00:04:08,525 --> 00:04:09,025 for 106 00:04:10,604 --> 00:04:11,745 what would we say? 107 00:04:12,125 --> 00:04:12,625 Operations? 108 00:04:14,310 --> 00:04:16,790 Yeah. Let's say operations is the equivalent of 109 00:04:16,790 --> 00:04:18,870 developers in this context. Yeah. Well, actually, I 110 00:04:18,870 --> 00:04:20,870 would split developers in two. It would be 111 00:04:20,870 --> 00:04:21,370 operations 112 00:04:21,990 --> 00:04:22,490 and 113 00:04:23,350 --> 00:04:24,089 the why, 114 00:04:25,589 --> 00:04:26,649 future designers, 115 00:04:28,545 --> 00:04:30,564 or design documents, maybe. 116 00:04:31,345 --> 00:04:33,904 And then management, what would that look like? 117 00:04:33,904 --> 00:04:35,904 Like, that's a little that's a little I 118 00:04:35,904 --> 00:04:37,845 think that would be more high level, 119 00:04:38,545 --> 00:04:39,045 documentation 120 00:04:39,584 --> 00:04:42,705 diagrams. I know that, I was once asked 121 00:04:42,705 --> 00:04:43,605 by a leader 122 00:04:44,050 --> 00:04:44,870 to put together 123 00:04:45,250 --> 00:04:46,069 a a diagram, 124 00:04:46,689 --> 00:04:47,189 and, 125 00:04:47,730 --> 00:04:49,410 you know, the the response I got back 126 00:04:49,410 --> 00:04:51,650 is, like, take all these IP addresses off. 127 00:04:51,650 --> 00:04:53,410 Take all you know, like, it's it's got 128 00:04:53,410 --> 00:04:55,590 too much information on it. Like, I want 129 00:04:58,305 --> 00:05:00,944 a a a notional diagram or something that's 130 00:05:00,944 --> 00:05:03,585 more business focused that gives us a very 131 00:05:03,585 --> 00:05:05,824 high level view of what's going on, but 132 00:05:05,824 --> 00:05:08,305 I don't want any of the specifics. Nobody 133 00:05:08,305 --> 00:05:09,444 cares about those. 134 00:05:09,930 --> 00:05:12,569 It's like and and so there's a there 135 00:05:12,569 --> 00:05:15,069 is and and sometimes that can be extreme, 136 00:05:15,769 --> 00:05:17,790 but there is a level of documentation 137 00:05:18,889 --> 00:05:20,029 of your environment 138 00:05:21,050 --> 00:05:21,550 that 139 00:05:21,850 --> 00:05:23,310 helps provide a conceptual 140 00:05:23,689 --> 00:05:24,189 overview 141 00:05:24,854 --> 00:05:27,414 without all of the implementation details. So I 142 00:05:27,414 --> 00:05:28,154 think that's, 143 00:05:29,654 --> 00:05:32,615 that's one level that might map to, 144 00:05:33,254 --> 00:05:36,074 to management. And and the other is still 145 00:05:36,134 --> 00:05:39,095 the same that's here, like issue trackers, project 146 00:05:39,095 --> 00:05:39,595 plans, 147 00:05:40,800 --> 00:05:42,419 schedules, things like that, 148 00:05:43,279 --> 00:05:44,259 status documents. 149 00:05:44,639 --> 00:05:46,659 All of all of that is still important 150 00:05:46,720 --> 00:05:49,060 for management of a, 151 00:05:50,000 --> 00:05:50,579 you know, 152 00:05:51,039 --> 00:05:52,719 a a a operational networking team or an 153 00:05:52,719 --> 00:05:54,675 engineering team or any of those things. Yeah. 154 00:05:54,675 --> 00:05:57,235 I thought when like, what are the clear 155 00:05:57,235 --> 00:05:59,395 features? And when we think about this, the 156 00:05:59,395 --> 00:06:00,935 problem for management is 157 00:06:01,395 --> 00:06:03,735 management thinks of features as products 158 00:06:04,995 --> 00:06:06,295 that are shipped to customers. 159 00:06:07,800 --> 00:06:09,319 But that's not really what we're talking about. 160 00:06:09,319 --> 00:06:10,379 We talk about networks. 161 00:06:10,839 --> 00:06:12,460 We talk about features like 162 00:06:12,920 --> 00:06:15,819 a virtualization overlay that enables products 163 00:06:16,839 --> 00:06:17,500 to me 164 00:06:19,160 --> 00:06:21,020 rather than the products themselves. 165 00:06:21,480 --> 00:06:22,939 I face this all the time. 166 00:06:23,455 --> 00:06:25,074 What products are there? Well, 167 00:06:25,855 --> 00:06:27,935 you know, they wanna think about things like 168 00:06:27,935 --> 00:06:29,715 what applications are being supported. 169 00:06:30,254 --> 00:06:31,555 Well, that's not really 170 00:06:31,935 --> 00:06:33,555 a feature of the network. 171 00:06:34,574 --> 00:06:37,479 The feature of the network is whatever enables 172 00:06:37,479 --> 00:06:39,259 that application to be supported, 173 00:06:39,639 --> 00:06:40,699 not the application. 174 00:06:41,719 --> 00:06:42,219 Virtualization 175 00:06:42,759 --> 00:06:43,259 or 176 00:06:44,360 --> 00:06:47,000 quality of service or traffic steering or traffic 177 00:06:47,000 --> 00:06:47,500 engineering. 178 00:06:49,464 --> 00:06:51,144 And and what I think you'd find if 179 00:06:51,144 --> 00:06:52,904 you start looking at that way is that 180 00:06:52,904 --> 00:06:54,824 almost all networks have the same set of 181 00:06:54,824 --> 00:06:55,324 things 182 00:06:55,944 --> 00:06:56,845 that they support 183 00:06:57,545 --> 00:06:59,785 in one way or the other in maybe 184 00:06:59,785 --> 00:07:01,165 with different levels of emphasis. 185 00:07:01,830 --> 00:07:03,750 I think what you would find so the 186 00:07:03,750 --> 00:07:06,310 level of interest that different audiences have in 187 00:07:06,310 --> 00:07:08,149 the in the knowledge that's available about a 188 00:07:08,149 --> 00:07:10,250 given network versus code base, 189 00:07:11,029 --> 00:07:13,910 you know, a manager doesn't really care what 190 00:07:13,910 --> 00:07:15,610 variables are named in code 191 00:07:15,944 --> 00:07:17,865 or or what how the data structures are 192 00:07:17,865 --> 00:07:19,625 laid out. That that doesn't really matter. Those 193 00:07:19,625 --> 00:07:21,224 are those are below the level of a 194 00:07:21,224 --> 00:07:22,284 feature to them. 195 00:07:22,584 --> 00:07:24,504 And and I think the analog in the 196 00:07:24,504 --> 00:07:27,305 networking world would be like like, I don't 197 00:07:27,305 --> 00:07:28,985 I don't really care about an l two 198 00:07:28,985 --> 00:07:29,485 overlay. 199 00:07:30,100 --> 00:07:32,580 What what probably what what a business leader 200 00:07:32,580 --> 00:07:33,319 cares about 201 00:07:33,779 --> 00:07:36,100 is what revenue stream does this enable. And 202 00:07:36,100 --> 00:07:37,540 that would to me would be the the 203 00:07:37,540 --> 00:07:39,220 equivalent of like a feature. Like if you're 204 00:07:39,220 --> 00:07:40,500 looking at a code base and you're purely 205 00:07:40,500 --> 00:07:41,480 just software development. 206 00:07:41,935 --> 00:07:43,694 Like, the feature is the thing that makes 207 00:07:43,694 --> 00:07:45,535 money for the business. That's what the that's 208 00:07:45,535 --> 00:07:46,915 why customers buy the thing. 209 00:07:47,375 --> 00:07:49,774 You know, and and with networking, it's kinda 210 00:07:49,854 --> 00:07:52,254 it's one layer hidden below. Customers don't buy 211 00:07:52,254 --> 00:07:53,790 overlays. Customers buy, 212 00:07:54,509 --> 00:07:57,389 caching services or customers buy, you know, the 213 00:07:57,389 --> 00:08:00,029 content delivery network. They don't they really don't 214 00:08:00,110 --> 00:08:01,790 the inners of it. Like, when I go 215 00:08:01,790 --> 00:08:03,550 to buy a car, I don't buy a 216 00:08:03,550 --> 00:08:05,230 spare tire. I don't I don't buy an 217 00:08:05,230 --> 00:08:07,509 engine. And so I think at least and 218 00:08:07,509 --> 00:08:08,589 the one of the things I think is 219 00:08:08,589 --> 00:08:10,134 interesting about this article, you look at the 220 00:08:10,134 --> 00:08:12,294 volume of words, the one that is the 221 00:08:12,294 --> 00:08:15,034 biggest by far is the management focused one. 222 00:08:15,094 --> 00:08:16,774 The rest of them are it's just it's 223 00:08:16,774 --> 00:08:18,935 it's small. I think that's really interesting. Yeah. 224 00:08:18,935 --> 00:08:20,294 And I and I agree with you about 225 00:08:20,294 --> 00:08:22,134 the car. But on the other hand, if 226 00:08:22,134 --> 00:08:22,875 I'm documenting 227 00:08:24,214 --> 00:08:26,649 what the car does or or I'm documenting 228 00:08:26,649 --> 00:08:27,790 the car to the management, 229 00:08:28,410 --> 00:08:29,709 I'm not gonna say, 230 00:08:32,090 --> 00:08:33,549 the car gets groceries. 231 00:08:34,649 --> 00:08:37,309 I'm gonna say the car enables the application 232 00:08:38,434 --> 00:08:39,654 of getting groceries. 233 00:08:40,195 --> 00:08:41,575 How does it do that? 234 00:08:41,955 --> 00:08:43,095 It has an engine. 235 00:08:43,955 --> 00:08:45,014 It has wheels. 236 00:08:45,634 --> 00:08:48,595 Right? It has things that make it able 237 00:08:48,595 --> 00:08:51,360 to go do that. Or I mean and, 238 00:08:51,360 --> 00:08:52,959 actually, that's not even true either. I would 239 00:08:52,959 --> 00:08:55,120 actually say, how does the car enable me 240 00:08:55,120 --> 00:08:57,059 to go get groceries? It is a trunk. 241 00:08:58,639 --> 00:09:00,579 It is a place to store groceries. 242 00:09:00,959 --> 00:09:02,899 It is a place to put people. 243 00:09:03,360 --> 00:09:05,379 It has a places to do things. 244 00:09:06,284 --> 00:09:07,804 And so I think we think of the 245 00:09:07,804 --> 00:09:09,664 network as either this detailed, 246 00:09:10,445 --> 00:09:12,865 like I said before, wheels and engine. 247 00:09:13,325 --> 00:09:15,804 That's not really what management cares about. But 248 00:09:15,804 --> 00:09:18,465 on the other hand, management doesn't necessarily 249 00:09:18,980 --> 00:09:20,600 we shouldn't push to management. 250 00:09:21,700 --> 00:09:22,100 It's 251 00:09:22,580 --> 00:09:24,340 it it it's for you know, it gets 252 00:09:24,340 --> 00:09:24,840 groceries. 253 00:09:25,300 --> 00:09:27,399 Like, not just what we apply it to. 254 00:09:27,540 --> 00:09:30,019 What features of the network allow it to 255 00:09:30,019 --> 00:09:32,420 be enabled to enable you to go get 256 00:09:32,420 --> 00:09:32,920 groceries? 257 00:09:34,019 --> 00:09:36,475 And and why are yeah. I I agree 258 00:09:36,475 --> 00:09:38,075 because because the thing is they're gonna be 259 00:09:38,075 --> 00:09:39,835 they're gonna be decisions they have to make 260 00:09:39,835 --> 00:09:41,855 in those components. They're gonna be cost differentials. 261 00:09:41,995 --> 00:09:43,215 Right? There's gonna be, 262 00:09:44,154 --> 00:09:45,674 a certain engine is needed to carry a 263 00:09:45,674 --> 00:09:47,799 certain size of a trunk. And so, yeah, 264 00:09:47,799 --> 00:09:49,559 the trunk is is what you care about 265 00:09:49,559 --> 00:09:51,799 from your balance sheet. But there there is 266 00:09:51,799 --> 00:09:54,440 some cost to this particular trunk, and here's 267 00:09:54,440 --> 00:09:56,759 why. That's that's the thing that I like, 268 00:09:56,759 --> 00:10:00,514 good documentation skips over the nerd stuff insofar 269 00:10:00,514 --> 00:10:02,434 as it doesn't serve the purpose of getting 270 00:10:02,434 --> 00:10:04,375 money for the thing that drives the revenue 271 00:10:04,674 --> 00:10:06,215 is is the way it's it. 272 00:10:07,075 --> 00:10:09,335 Well, I think the the inherent, 273 00:10:10,035 --> 00:10:10,615 you know, 274 00:10:10,995 --> 00:10:13,335 the the entire focus of this audience 275 00:10:13,715 --> 00:10:15,495 or this document is 276 00:10:16,009 --> 00:10:18,009 who's your audience. Right? Who are you writing 277 00:10:18,009 --> 00:10:20,250 for, and what's the right context? I think 278 00:10:20,250 --> 00:10:22,889 the part of the reason the management section 279 00:10:22,889 --> 00:10:24,730 is so long is that a lot of 280 00:10:24,730 --> 00:10:26,330 the things that are covered in here are 281 00:10:26,330 --> 00:10:28,250 not things that we necessarily think of just 282 00:10:28,250 --> 00:10:28,909 as documentation. 283 00:10:29,210 --> 00:10:29,870 They're not 284 00:10:30,464 --> 00:10:33,284 static pieces of information that describe 285 00:10:33,664 --> 00:10:35,764 either what the code is doing 286 00:10:36,384 --> 00:10:38,164 or what it does or why it exists. 287 00:10:38,625 --> 00:10:41,184 There it's it's status of a project. Right? 288 00:10:41,184 --> 00:10:43,824 Like, you know, are are there are there 289 00:10:43,824 --> 00:10:46,160 blockers that are are stopping the progress from 290 00:10:46,160 --> 00:10:48,399 going forward? Is there a bottleneck? What's our 291 00:10:48,399 --> 00:10:50,980 capacity as a team? Are we over overloaded 292 00:10:51,840 --> 00:10:54,179 or underloaded? And and all of these, 293 00:10:55,920 --> 00:10:57,620 really, this management documentation 294 00:10:58,174 --> 00:11:00,834 section is a communication of the state 295 00:11:01,375 --> 00:11:04,014 of a project, is really what's communicate is 296 00:11:04,014 --> 00:11:07,454 is being outlined here. You know? And and 297 00:11:07,454 --> 00:11:07,954 that, 298 00:11:08,815 --> 00:11:11,534 you know, as as a now manager, like, 299 00:11:11,534 --> 00:11:14,259 those are things I'm really concerned about. I'm 300 00:11:14,259 --> 00:11:15,399 not so much concerned 301 00:11:16,179 --> 00:11:18,740 about all of the the specific details that 302 00:11:18,740 --> 00:11:21,059 my folks talk to customers about. I'm certain 303 00:11:21,059 --> 00:11:22,339 that they know all those things, that they 304 00:11:22,339 --> 00:11:25,079 understand them. What I'm concerned about is 305 00:11:25,404 --> 00:11:27,725 if I get five more of these requests 306 00:11:27,725 --> 00:11:28,625 in tomorrow, 307 00:11:29,085 --> 00:11:31,644 am I gonna have the right distribution of 308 00:11:31,644 --> 00:11:32,144 people 309 00:11:32,764 --> 00:11:34,705 to respond to those requests? 310 00:11:35,164 --> 00:11:38,705 And that's a very different information set 311 00:11:39,120 --> 00:11:39,620 than 312 00:11:40,080 --> 00:11:41,679 what my folks need to go do their 313 00:11:41,679 --> 00:11:42,340 day job. 314 00:11:43,120 --> 00:11:46,080 Right. Yeah. Yep. Yeah. I think and I 315 00:11:46,080 --> 00:11:47,519 I think some of the questions they ask 316 00:11:47,519 --> 00:11:49,460 you are like, is anything at risk? 317 00:11:50,160 --> 00:11:52,240 Okay. Well, I mean, I deal all the 318 00:11:52,240 --> 00:11:54,585 time, like Tom was saying on the management 319 00:11:54,585 --> 00:11:55,245 side that 320 00:11:55,865 --> 00:11:57,945 I deal all the time where you wanna 321 00:11:57,945 --> 00:12:00,524 add this technology to the network. Why? 322 00:12:02,264 --> 00:12:04,345 Well, let me tell you about the business 323 00:12:04,345 --> 00:12:06,045 cases you're making to me 324 00:12:06,720 --> 00:12:08,179 and how I can enable 325 00:12:09,279 --> 00:12:10,339 those business cases. 326 00:12:11,279 --> 00:12:13,600 You give me five business cases. I want 327 00:12:13,600 --> 00:12:16,240 to enable those business cases in the simplest 328 00:12:16,240 --> 00:12:17,059 way possible. 329 00:12:17,519 --> 00:12:20,179 That's why this technology solves all of those. 330 00:12:20,985 --> 00:12:22,664 But, again, you can't even really put it 331 00:12:22,664 --> 00:12:23,725 in terms of technology. 332 00:12:24,105 --> 00:12:24,605 Right? 333 00:12:25,144 --> 00:12:26,365 It's all about capability, 334 00:12:27,464 --> 00:12:29,404 and that's and that's the thing. 335 00:12:31,225 --> 00:12:32,445 Yeah. And and 336 00:12:33,769 --> 00:12:35,370 the other things that are in here on 337 00:12:35,370 --> 00:12:37,690 the management side are really you know, teams 338 00:12:37,690 --> 00:12:39,690 are under loaded or overloaded. Right? Like, you're 339 00:12:39,690 --> 00:12:42,029 talking about Yvonne. Right? That's really 340 00:12:42,889 --> 00:12:45,470 a big deal. Are we on track? 341 00:12:45,850 --> 00:12:47,850 What kinds of changes have been made to 342 00:12:47,850 --> 00:12:48,590 these features, 343 00:12:48,954 --> 00:12:49,774 and why? 344 00:12:51,355 --> 00:12:53,615 Like, what what has evolved about 345 00:12:54,475 --> 00:12:54,975 the 346 00:12:55,674 --> 00:12:57,514 status of the network? What new things have 347 00:12:57,514 --> 00:12:59,115 we done, and why have those things been 348 00:12:59,115 --> 00:12:59,615 done? 349 00:12:59,995 --> 00:13:01,514 So I think I think those are all 350 00:13:01,514 --> 00:13:02,014 very 351 00:13:03,820 --> 00:13:06,139 good questions we can convert into the network 352 00:13:06,139 --> 00:13:08,000 world. And people always say, 353 00:13:08,379 --> 00:13:10,620 how can I relate to management? How can 354 00:13:10,620 --> 00:13:12,320 I get them to invest in the network? 355 00:13:13,340 --> 00:13:15,120 Well, there's your answer right there. 356 00:13:15,820 --> 00:13:17,039 This is your answer. 357 00:13:18,904 --> 00:13:20,904 Thinking about them as an audience and figuring 358 00:13:20,904 --> 00:13:22,845 out what that what that looks like. 359 00:13:24,024 --> 00:13:26,664 I really liked the, the last section customer 360 00:13:26,664 --> 00:13:27,164 documentation. 361 00:13:27,465 --> 00:13:29,784 Customer documentation is highly beneficial as it helps 362 00:13:29,784 --> 00:13:31,945 prevent customer issues and makes the software easier 363 00:13:31,945 --> 00:13:34,250 to implement, which in turn makes it more 364 00:13:34,250 --> 00:13:34,750 valuable. 365 00:13:36,169 --> 00:13:38,730 I I really like that documentation makes the 366 00:13:38,730 --> 00:13:41,610 product more valuable. What? Yes. Totally. I totally 367 00:13:41,610 --> 00:13:44,429 agree with that. Yeah. And who's the customer 368 00:13:44,570 --> 00:13:45,389 for a network? 369 00:13:45,835 --> 00:13:47,835 That's always a question we forget. We we 370 00:13:47,835 --> 00:13:49,455 don't tend to think about a lot. 371 00:13:50,075 --> 00:13:51,615 Who is the network's customer? 372 00:13:52,075 --> 00:13:54,174 I would say it's application owners. 373 00:13:56,235 --> 00:13:57,934 I wouldn't say it's the business. 374 00:13:59,139 --> 00:14:01,940 The biz the the applications customer is the 375 00:14:01,940 --> 00:14:02,440 business. 376 00:14:03,700 --> 00:14:05,639 The business's customer is the customer. 377 00:14:06,500 --> 00:14:09,240 The network's customer is the application owner. 378 00:14:09,700 --> 00:14:11,139 That's the way I would think of it. 379 00:14:11,139 --> 00:14:12,615 I don't know. Yvonne's 380 00:14:13,394 --> 00:14:15,735 thinking about it. I think I think, 381 00:14:17,555 --> 00:14:20,035 certainly. I also think it depends on how 382 00:14:20,035 --> 00:14:22,215 you break up your organization. Like, 383 00:14:22,595 --> 00:14:24,835 is is there an infrastructure team that's also 384 00:14:24,915 --> 00:14:26,409 like, or is it is it the people 385 00:14:26,409 --> 00:14:28,669 that are deploying, building, maintaining 386 00:14:30,169 --> 00:14:31,470 the servers and infrastructure 387 00:14:31,850 --> 00:14:33,149 in your organization? 388 00:14:33,610 --> 00:14:35,629 Are those customers of your, 389 00:14:36,409 --> 00:14:38,350 network? They very well may be. 390 00:14:38,945 --> 00:14:39,845 Are there, 391 00:14:40,304 --> 00:14:42,705 is there any automation that's built inside of 392 00:14:42,705 --> 00:14:45,904 your network that other teams interact with? Is 393 00:14:45,904 --> 00:14:46,404 there 394 00:14:46,945 --> 00:14:48,384 I'll tell you who else could be a 395 00:14:48,384 --> 00:14:48,884 customer 396 00:14:49,504 --> 00:14:50,004 is, 397 00:14:50,625 --> 00:14:53,149 finance or billing. Like, if you if you're 398 00:14:53,149 --> 00:14:53,950 running a, 399 00:14:57,629 --> 00:14:59,329 if you sell network services, 400 00:15:00,350 --> 00:15:03,389 there are absolutely logs and data that come 401 00:15:03,389 --> 00:15:05,089 out of your network infrastructure 402 00:15:05,815 --> 00:15:09,175 that feed into workflows for your, you know, 403 00:15:09,175 --> 00:15:10,154 billing, accounting, 404 00:15:10,535 --> 00:15:12,535 operations teams that they need. So there can 405 00:15:12,535 --> 00:15:14,875 be lots of customers depending on your organization, 406 00:15:14,935 --> 00:15:17,735 what what you do. Yep. Billing is actually 407 00:15:17,735 --> 00:15:19,815 a big one too. I always think of 408 00:15:19,815 --> 00:15:21,360 application owners, but 409 00:15:22,159 --> 00:15:24,419 I do interface a lot 410 00:15:24,959 --> 00:15:26,179 with billing people 411 00:15:26,639 --> 00:15:27,860 in my current job. 412 00:15:29,600 --> 00:15:32,480 It's not something I'm really accustomed to. But 413 00:15:32,720 --> 00:15:34,079 Well, in some and I think I think 414 00:15:34,079 --> 00:15:36,644 app the concept of application owner can stretch 415 00:15:36,725 --> 00:15:37,625 even a little bit 416 00:15:38,085 --> 00:15:39,764 further. So if you're 417 00:15:40,245 --> 00:15:42,485 say you're running, like a little eyeball network, 418 00:15:42,485 --> 00:15:44,184 a regional network, regional ISP, 419 00:15:44,804 --> 00:15:46,644 you still have application owners. It's just that 420 00:15:46,644 --> 00:15:48,184 there's one of them per subscriber. 421 00:15:48,899 --> 00:15:50,039 And the application 422 00:15:50,579 --> 00:15:53,559 is whatever application they're using from their home. 423 00:15:54,339 --> 00:15:56,740 So and and in that case, it it 424 00:15:56,740 --> 00:15:59,139 gets really interesting because you you could call 425 00:15:59,139 --> 00:16:01,379 the customer the application owner because they're using 426 00:16:01,379 --> 00:16:02,839 your network to move their bits. 427 00:16:03,245 --> 00:16:05,725 But then but then also, I mean, they 428 00:16:05,725 --> 00:16:06,784 they are a customer 429 00:16:07,245 --> 00:16:08,464 of another application, 430 00:16:09,325 --> 00:16:11,804 most of them. And so I think that's 431 00:16:12,764 --> 00:16:14,125 I I think it's another way to look 432 00:16:14,125 --> 00:16:16,304 at it. Like, even if you're totally commodity 433 00:16:16,445 --> 00:16:17,985 just moving bits for a living, 434 00:16:18,430 --> 00:16:19,170 you still 435 00:16:19,550 --> 00:16:22,690 you still have, application owners and you have 436 00:16:23,149 --> 00:16:24,670 as you get to more as you get 437 00:16:24,670 --> 00:16:26,670 to a subscriber centric like that like an 438 00:16:26,670 --> 00:16:27,889 ISP type of business, 439 00:16:28,430 --> 00:16:30,269 you get less and less control over what 440 00:16:30,269 --> 00:16:32,455 those application owners are doing with your network, 441 00:16:32,455 --> 00:16:34,554 which I think is also an interesting property 442 00:16:35,014 --> 00:16:37,095 of documenting a network. The other thing I 443 00:16:37,095 --> 00:16:39,115 would always point out is that documentation 444 00:16:39,815 --> 00:16:41,735 in the network world he doesn't talk about 445 00:16:41,735 --> 00:16:42,634 it in this article, 446 00:16:43,254 --> 00:16:45,995 but documentation has layers. There are layers 447 00:16:46,799 --> 00:16:47,539 of documentation. 448 00:16:50,799 --> 00:16:53,039 I the number of times I've walked into 449 00:16:53,039 --> 00:16:54,980 a situation when I was on global escalation 450 00:16:55,360 --> 00:16:57,679 and said, give me your network documentation, they 451 00:16:57,679 --> 00:16:59,919 walked me into this conference room, and there's 452 00:16:59,919 --> 00:17:00,419 these 453 00:17:00,799 --> 00:17:01,539 flip charts 454 00:17:02,774 --> 00:17:03,995 of physical connectivity. 455 00:17:06,454 --> 00:17:07,355 And you're like, 456 00:17:07,734 --> 00:17:09,494 yeah. I didn't need to know your circuit 457 00:17:09,494 --> 00:17:09,994 grooming. 458 00:17:11,255 --> 00:17:11,755 Seriously. 459 00:17:12,214 --> 00:17:12,714 Like, 460 00:17:13,654 --> 00:17:15,595 BGP and OSPF and EIGRP 461 00:17:15,974 --> 00:17:18,750 and IS IS, they don't care about circuit 462 00:17:18,750 --> 00:17:19,250 grooming. 463 00:17:19,630 --> 00:17:21,970 Okay? That's not where the problem is. 464 00:17:22,429 --> 00:17:25,069 And so that's a very common problem as 465 00:17:25,069 --> 00:17:27,470 well, I find, is that people don't know 466 00:17:27,470 --> 00:17:29,329 how to write documentation 467 00:17:29,630 --> 00:17:30,369 in layers 468 00:17:30,990 --> 00:17:33,474 per se. Well, the there's and and and 469 00:17:33,474 --> 00:17:35,315 we can talk through those quickly. Right? There's 470 00:17:35,315 --> 00:17:37,575 the physical layer that is, you know, 471 00:17:37,954 --> 00:17:38,115 the 472 00:17:38,755 --> 00:17:40,694 this physical wire is connected 473 00:17:41,075 --> 00:17:44,194 to this physical port on this physical device 474 00:17:44,194 --> 00:17:47,109 that lives in this physical rack, in this 475 00:17:47,109 --> 00:17:49,190 physical room, in this physical data center. Like, 476 00:17:49,190 --> 00:17:50,170 there's all of that. 477 00:17:50,710 --> 00:17:53,450 And then you've got your more logical documentation 478 00:17:54,470 --> 00:17:54,970 that's 479 00:17:56,390 --> 00:17:58,230 all of these things are connected to one 480 00:17:58,230 --> 00:18:00,390 another in in this sort of arrangement. It 481 00:18:00,390 --> 00:18:01,884 may not represent what it looks like in 482 00:18:01,884 --> 00:18:04,684 the physical world, but it helps you understand 483 00:18:04,684 --> 00:18:06,144 logically what's going on. 484 00:18:06,765 --> 00:18:09,325 And and then there's overlays that you might 485 00:18:09,325 --> 00:18:11,244 have over top of that, which is a 486 00:18:11,244 --> 00:18:14,125 completely different logical diagram. And then at each 487 00:18:14,125 --> 00:18:15,825 of those layers, there are different 488 00:18:16,460 --> 00:18:18,320 protocols that could be in play. 489 00:18:19,099 --> 00:18:22,380 Separate from that are things like IPAM. What's 490 00:18:22,380 --> 00:18:25,099 what's what's your IP address scheme? Where is 491 00:18:25,099 --> 00:18:27,099 all of that? How are you tracking all 492 00:18:27,099 --> 00:18:30,480 of that information and documenting it? There's configuration. 493 00:18:31,815 --> 00:18:33,034 There's automation, 494 00:18:33,654 --> 00:18:35,494 which is which is another place where there 495 00:18:35,494 --> 00:18:36,474 could be documentation. 496 00:18:37,174 --> 00:18:39,274 Should be absolutely be documentation. 497 00:18:39,734 --> 00:18:41,255 I think I'll get a hear you from 498 00:18:41,255 --> 00:18:42,154 Tom on that. 499 00:18:42,774 --> 00:18:45,994 You absolutely need to document your your, automation. 500 00:18:47,309 --> 00:18:49,789 And and then there's like, what do we 501 00:18:49,789 --> 00:18:52,109 actually accomplish with this thing? Right? And and 502 00:18:52,109 --> 00:18:53,869 there are all those different layers that have 503 00:18:53,869 --> 00:18:55,089 all these different audiences, 504 00:18:55,869 --> 00:18:58,529 and you you have to have some concept 505 00:18:58,669 --> 00:19:01,034 of what the documentation is, who the user 506 00:19:01,034 --> 00:19:02,154 is, and what it's gonna be used for 507 00:19:02,154 --> 00:19:03,694 before you even know how to write it. 508 00:19:03,754 --> 00:19:05,514 But there's a lot. There's a lot a 509 00:19:05,514 --> 00:19:08,575 lot. Yeah. Yeah. I think so too. So 510 00:19:09,034 --> 00:19:11,115 alright. So any other hits on that one 511 00:19:11,115 --> 00:19:13,214 before we move on to our next little? 512 00:19:14,759 --> 00:19:16,220 I'll just say I 513 00:19:16,600 --> 00:19:19,320 my experience with documentation mostly it's just like, 514 00:19:19,640 --> 00:19:21,400 what most people would call boring work, but 515 00:19:21,400 --> 00:19:22,380 I have had times 516 00:19:22,840 --> 00:19:25,160 when it the stuff that I that I 517 00:19:25,160 --> 00:19:26,299 documented was 518 00:19:26,855 --> 00:19:28,295 was it had to be up to date 519 00:19:28,295 --> 00:19:30,215 because I was proposing changes, and so I 520 00:19:30,215 --> 00:19:30,795 did it. 521 00:19:31,174 --> 00:19:33,115 And that had had some surprising, 522 00:19:34,215 --> 00:19:35,275 political conversations. 523 00:19:35,575 --> 00:19:37,654 And, like, I I've had bosses take my 524 00:19:37,654 --> 00:19:39,275 documentation to other people 525 00:19:39,589 --> 00:19:41,130 and say, this is what we're doing. 526 00:19:41,430 --> 00:19:41,910 And, 527 00:19:42,309 --> 00:19:43,589 we talked about this on the phone, and 528 00:19:43,589 --> 00:19:44,710 I know you're not used to us actually 529 00:19:44,710 --> 00:19:46,150 taking any notes, but we wrote it all 530 00:19:46,150 --> 00:19:48,150 down. Here's the documentation, and so we're moving 531 00:19:48,150 --> 00:19:49,990 forward with this. And they're like, wait. No. 532 00:19:49,990 --> 00:19:51,670 No. No. No. And it's like, well, sorry. 533 00:19:51,670 --> 00:19:54,410 It's written. This is it's here. It's recorded. 534 00:19:54,634 --> 00:19:56,475 And I was I he told me he 535 00:19:56,475 --> 00:19:58,075 did this this when this happened, he told 536 00:19:58,075 --> 00:19:59,195 me he did this after the fact and 537 00:19:59,195 --> 00:20:01,835 I was like, woah. That's pretty bold. But 538 00:20:01,835 --> 00:20:04,154 if it's written, if it's there, it it, 539 00:20:04,315 --> 00:20:06,154 you know, it trumps what's in somebody's mind 540 00:20:06,154 --> 00:20:07,755 and what they believe and what they think 541 00:20:07,755 --> 00:20:09,419 and Yeah. So, I mean, we can't just 542 00:20:09,579 --> 00:20:10,480 A %. 543 00:20:10,700 --> 00:20:12,460 Yep. There was a book I read once 544 00:20:12,460 --> 00:20:13,839 about a market crash, 545 00:20:14,299 --> 00:20:16,779 like a big financial market crash, and it 546 00:20:16,779 --> 00:20:19,259 was brought on by some hackers who had 547 00:20:19,259 --> 00:20:21,359 broken into the market and caused havoc. 548 00:20:22,460 --> 00:20:22,960 And 549 00:20:23,455 --> 00:20:25,715 the solution they came back to was, 550 00:20:26,654 --> 00:20:28,894 well, if it's not written down, it didn't 551 00:20:28,894 --> 00:20:29,394 happen. 552 00:20:30,575 --> 00:20:31,634 They just literally 553 00:20:31,934 --> 00:20:33,234 rewound the market 554 00:20:34,015 --> 00:20:35,955 to five minutes before the hack 555 00:20:36,740 --> 00:20:37,559 and said, 556 00:20:37,940 --> 00:20:38,440 go. 557 00:20:38,900 --> 00:20:40,900 I thought that was that's a pretty creative 558 00:20:40,900 --> 00:20:42,820 way of of looking at it. If it's 559 00:20:42,820 --> 00:20:44,440 not written down, it didn't happen. 560 00:20:45,380 --> 00:20:45,880 So 561 00:20:46,420 --> 00:20:46,920 alright. 562 00:20:47,220 --> 00:20:48,740 So the next thing we're gonna talk about 563 00:20:48,740 --> 00:20:51,559 is legacy versus EOL versus tech debt. 564 00:20:53,184 --> 00:20:56,244 I don't know. I have very specific definitions 565 00:20:56,304 --> 00:20:58,004 for these and that y'all probably 566 00:20:58,544 --> 00:20:59,684 have different definitions. 567 00:21:00,224 --> 00:21:02,384 Well, let's hear yours first, Russ. Oh, no. 568 00:21:02,384 --> 00:21:03,444 That's no fun. 569 00:21:06,490 --> 00:21:08,490 Well, legacy just means it's been here a 570 00:21:08,490 --> 00:21:09,150 long time. 571 00:21:09,849 --> 00:21:11,869 Yeah. Legacy just means But often, 572 00:21:12,250 --> 00:21:14,990 legacy means this is a thing that's worked. 573 00:21:15,769 --> 00:21:18,970 Yeah. Right. Often, people see legacy as bad. 574 00:21:18,970 --> 00:21:20,190 It's not bad 575 00:21:21,265 --> 00:21:21,765 necessarily. 576 00:21:22,704 --> 00:21:24,865 Legacy is just what's been here for a 577 00:21:24,865 --> 00:21:27,605 while. Right? And sometimes it is bad. 578 00:21:28,065 --> 00:21:30,065 Oh, we've always done it that way. That's 579 00:21:30,065 --> 00:21:31,204 a statement of legacy. 580 00:21:31,505 --> 00:21:33,365 Oh, no. Don't do that. That's nonsense. 581 00:21:33,744 --> 00:21:36,529 Right? Or that old AGS plus has been 582 00:21:36,529 --> 00:21:38,230 sitting in the corner for ten years. 583 00:21:38,529 --> 00:21:40,130 Why don't don't fix it because it ain't 584 00:21:40,130 --> 00:21:40,630 broken. 585 00:21:41,410 --> 00:21:43,250 No. No. Those are not those are not 586 00:21:43,250 --> 00:21:46,609 good either. Right? So legacy has positive and 587 00:21:46,609 --> 00:21:47,109 negative 588 00:21:47,835 --> 00:21:48,335 connotations, 589 00:21:48,875 --> 00:21:50,894 and we take it as negative, 590 00:21:51,275 --> 00:21:52,974 like, 90% of the time. 591 00:21:53,515 --> 00:21:56,075 In our minds, legacy means I need to 592 00:21:56,075 --> 00:21:56,974 replace this. 593 00:21:58,154 --> 00:22:00,849 No. That's that's that's not true. Right? 594 00:22:01,309 --> 00:22:03,789 It just means something that's been there. I 595 00:22:03,789 --> 00:22:06,029 can't tell you how many customers I've talked 596 00:22:06,029 --> 00:22:08,289 to who say to me almost apologetically 597 00:22:09,070 --> 00:22:10,609 that they still run a mainframe. 598 00:22:11,309 --> 00:22:12,450 But the thing is, 599 00:22:12,944 --> 00:22:15,044 it's all of them. Yeah. Right? 600 00:22:15,424 --> 00:22:18,164 Any company of any sufficient size 601 00:22:18,464 --> 00:22:20,325 that's been around for a few decades 602 00:22:20,944 --> 00:22:21,765 has one. 603 00:22:22,544 --> 00:22:23,044 And 604 00:22:23,345 --> 00:22:26,565 unless they have a very clear business case 605 00:22:27,240 --> 00:22:27,980 and a 606 00:22:29,319 --> 00:22:31,019 very driven project plan, 607 00:22:31,799 --> 00:22:34,140 any attempt to replace that thing 608 00:22:34,680 --> 00:22:37,480 is a fool's errand unless folks really know 609 00:22:37,480 --> 00:22:39,480 what they're doing and they have a very 610 00:22:39,480 --> 00:22:42,279 committed plan. If it's just like somebody's wild 611 00:22:42,279 --> 00:22:44,575 hair, like, oh, this year, we're gonna replace 612 00:22:44,575 --> 00:22:45,234 the mainframe. 613 00:22:46,095 --> 00:22:47,875 The those just don't go well. 614 00:22:48,894 --> 00:22:51,154 And it's because those systems 615 00:22:51,934 --> 00:22:53,615 do what they were designed to do, do 616 00:22:53,615 --> 00:22:55,474 it well, and have been sitting there running 617 00:22:55,535 --> 00:22:56,595 for decades. 618 00:22:57,214 --> 00:22:57,714 Yep. 619 00:22:58,279 --> 00:22:58,779 Yep. 620 00:22:59,240 --> 00:23:01,579 I mean, COBOL still exists, unfortunately. 621 00:23:02,039 --> 00:23:04,140 It's a horrible language. But anyway, 622 00:23:04,599 --> 00:23:05,659 there I mean, 623 00:23:06,279 --> 00:23:08,679 I worked for BASF for some period of 624 00:23:08,679 --> 00:23:10,200 time, way back in my 625 00:23:11,000 --> 00:23:13,375 you know, Yvonne figured out how how old 626 00:23:13,375 --> 00:23:15,775 I am, and she's like, no. It can't 627 00:23:15,775 --> 00:23:17,315 be, but it is. Anyway 628 00:23:18,414 --> 00:23:18,914 and 629 00:23:20,974 --> 00:23:22,815 and one of the things that BASF tried 630 00:23:22,815 --> 00:23:25,375 to do way back in the days just 631 00:23:25,375 --> 00:23:26,970 after I got out of the air force 632 00:23:27,369 --> 00:23:29,390 was to replace their 633 00:23:30,250 --> 00:23:30,750 mainframe. 634 00:23:31,369 --> 00:23:33,309 I believe it was an s three ninety. 635 00:23:33,529 --> 00:23:34,970 I could be wrong, but I believe it 636 00:23:34,970 --> 00:23:36,190 was an s three ninety. 637 00:23:36,730 --> 00:23:38,009 By the way, in the Air Force, we 638 00:23:38,009 --> 00:23:39,309 ran a Burroughs mainframe 639 00:23:40,234 --> 00:23:42,494 with TN thirty two seventy terminal emulators. 640 00:23:43,355 --> 00:23:46,335 I mean, way even before IBM s March. 641 00:23:46,955 --> 00:23:48,555 And I worked in another company where we 642 00:23:48,555 --> 00:23:50,875 had a bunch of IBM minis instead of 643 00:23:50,875 --> 00:23:51,775 having mainframes. 644 00:23:52,075 --> 00:23:54,315 And we had all this fancy middleware that 645 00:23:54,315 --> 00:23:54,815 combined 646 00:23:55,720 --> 00:23:58,279 queries from the different minis to get to, 647 00:23:58,279 --> 00:24:00,299 like give me this is, like, really crazy. 648 00:24:00,679 --> 00:24:02,839 But, anyway, they had an s three ninety, 649 00:24:02,839 --> 00:24:04,460 and they were gonna replace it. 650 00:24:05,079 --> 00:24:05,899 Okay. Fine. 651 00:24:07,079 --> 00:24:07,819 It took 652 00:24:08,599 --> 00:24:09,419 two years 653 00:24:10,414 --> 00:24:12,194 and an entire department 654 00:24:12,494 --> 00:24:13,234 of people, 655 00:24:13,934 --> 00:24:16,355 and they were bringing people out of retirement 656 00:24:17,134 --> 00:24:20,194 because no one knew how certain things worked 657 00:24:21,615 --> 00:24:23,339 to replace that s three ninety. 658 00:24:24,299 --> 00:24:26,220 Now in the long run, was it worth 659 00:24:26,220 --> 00:24:26,720 replacing? 660 00:24:27,740 --> 00:24:28,640 I don't know. 661 00:24:29,019 --> 00:24:30,400 Sometimes you just think, 662 00:24:30,940 --> 00:24:32,700 let it just sit there and do its 663 00:24:32,700 --> 00:24:35,019 thing. And as you build new stuff, you 664 00:24:35,019 --> 00:24:36,480 don't build it on that. 665 00:24:37,174 --> 00:24:39,434 You build it on on new new platforms. 666 00:24:40,134 --> 00:24:42,634 And it might take twenty years for everything 667 00:24:42,694 --> 00:24:44,554 to rotate out of being useful. 668 00:24:45,414 --> 00:24:46,075 Oh, well, 669 00:24:47,015 --> 00:24:47,914 like, really? 670 00:24:48,375 --> 00:24:49,595 Is it hurting anything? 671 00:24:50,369 --> 00:24:52,289 Yeah. But but but the tax you pay 672 00:24:52,289 --> 00:24:53,970 for that I hear you. But the tax 673 00:24:53,970 --> 00:24:55,349 you pay for that is interop 674 00:24:55,730 --> 00:24:57,650 because your new systems are going to come 675 00:24:57,650 --> 00:24:59,089 online and they are going to have to 676 00:24:59,089 --> 00:24:59,589 interoperate. 677 00:25:00,529 --> 00:25:02,049 To me, if you can put something on 678 00:25:02,049 --> 00:25:04,525 an island and never and not have strong 679 00:25:04,525 --> 00:25:07,245 dependencies and infrastructure, then, yeah, just let it 680 00:25:07,245 --> 00:25:08,924 let it go. And, you know, the finance 681 00:25:08,924 --> 00:25:10,445 people will love you. It'll you know, it's 682 00:25:10,445 --> 00:25:10,684 it's, 683 00:25:11,485 --> 00:25:12,865 as it, depreciates, 684 00:25:13,565 --> 00:25:15,485 over a much longer period of time, you're 685 00:25:15,485 --> 00:25:17,730 gonna it's it's good for the business. But 686 00:25:17,890 --> 00:25:19,809 but that's not for free. You you have 687 00:25:19,809 --> 00:25:20,470 an interop, 688 00:25:20,929 --> 00:25:22,950 surface on almost every system. 689 00:25:23,809 --> 00:25:26,049 And that's where I think I think where 690 00:25:26,049 --> 00:25:27,970 people are like, well, we don't let's just 691 00:25:27,970 --> 00:25:30,049 keep the old thing. Sometimes I think we 692 00:25:30,049 --> 00:25:32,174 don't take into account the interop tax and 693 00:25:32,174 --> 00:25:34,115 because it doesn't become apparent until later. 694 00:25:35,375 --> 00:25:37,295 That and just finding people to work on 695 00:25:37,295 --> 00:25:39,615 it. Right? That's the other thing is, like, 696 00:25:39,615 --> 00:25:41,535 at some point you reach this critical mass 697 00:25:41,535 --> 00:25:42,035 where, 698 00:25:42,495 --> 00:25:43,154 you know, 699 00:25:43,535 --> 00:25:44,434 is is it 700 00:25:44,750 --> 00:25:47,230 is it supportable, maintainable, and all that, which 701 00:25:47,230 --> 00:25:49,490 will lead us into our our next topic. 702 00:25:49,549 --> 00:25:51,630 But legacy in and of itself is not 703 00:25:51,630 --> 00:25:53,309 bad if you have the support that you 704 00:25:53,309 --> 00:25:54,369 need and 705 00:25:54,750 --> 00:25:57,710 the capabilities that you need. I mean, but 706 00:25:57,710 --> 00:25:59,410 in and of itself, it's not bad. 707 00:26:01,205 --> 00:26:03,285 Yeah. So EOL is the next thing you're 708 00:26:03,285 --> 00:26:05,924 talking about. Right, Yvonne? End of life? It 709 00:26:05,924 --> 00:26:06,424 is. 710 00:26:07,684 --> 00:26:09,224 Yeah. What does that mean? 711 00:26:10,325 --> 00:26:12,724 To me, it means you can't get support 712 00:26:12,724 --> 00:26:13,625 for it anymore. 713 00:26:14,440 --> 00:26:16,940 That's really what it means. It doesn't mean 714 00:26:17,080 --> 00:26:17,820 the manufacturer 715 00:26:18,279 --> 00:26:19,660 has declared it EOL. 716 00:26:21,160 --> 00:26:23,559 It means you can't get support for it 717 00:26:23,559 --> 00:26:24,059 anymore. 718 00:26:25,320 --> 00:26:25,820 Right? 719 00:26:26,565 --> 00:26:27,704 It's a mechanism 720 00:26:29,365 --> 00:26:31,524 that the vendor is using to drive a 721 00:26:31,524 --> 00:26:33,924 purchasing cycle. Yes. At the end of the 722 00:26:33,924 --> 00:26:34,424 day, 723 00:26:34,964 --> 00:26:37,444 it is a way for the vendor to 724 00:26:37,444 --> 00:26:38,345 force you 725 00:26:39,319 --> 00:26:41,019 to buy the next thing. 726 00:26:41,559 --> 00:26:43,640 It is it is its own kind of 727 00:26:43,640 --> 00:26:44,779 planned obsolescence. 728 00:26:46,039 --> 00:26:47,900 Now sometimes there are good reasons 729 00:26:48,599 --> 00:26:50,839 for technology to go end of life. Either 730 00:26:50,839 --> 00:26:53,174 it's it's difficult it gets to all this 731 00:26:53,174 --> 00:26:55,575 difficult to patch. There's newer technology that's gonna 732 00:26:55,575 --> 00:26:57,974 function a lot better. There's all kinds of 733 00:26:57,974 --> 00:26:59,494 reasons for that. But at the end of 734 00:26:59,494 --> 00:27:01,515 the day, when something's end of life, 735 00:27:02,535 --> 00:27:04,695 that really just means the person you bought 736 00:27:04,695 --> 00:27:07,509 it bought it from is no longer is 737 00:27:07,509 --> 00:27:09,349 saying I'm off the hook for dealing with 738 00:27:09,349 --> 00:27:10,009 this thing. 739 00:27:10,389 --> 00:27:12,329 Use it. Use it at your own risk, 740 00:27:12,710 --> 00:27:15,049 but don't come to me asking for patches 741 00:27:15,109 --> 00:27:17,929 or asking for help. You're on your own. 742 00:27:18,855 --> 00:27:21,835 And the reason that works is because most 743 00:27:22,055 --> 00:27:22,555 organizations 744 00:27:23,414 --> 00:27:25,195 aren't willing to assume that risk. 745 00:27:25,575 --> 00:27:27,434 Yep. That's exactly right. Yeah. 746 00:27:27,815 --> 00:27:29,414 As a matter of fact, again, an old 747 00:27:29,414 --> 00:27:30,474 story from Cisco, 748 00:27:30,855 --> 00:27:32,075 we had a 749 00:27:32,400 --> 00:27:33,619 customer come to us, 750 00:27:34,080 --> 00:27:36,080 and they had purchased a particular brand of 751 00:27:36,080 --> 00:27:37,140 routers, Bay routers. 752 00:27:38,240 --> 00:27:38,740 Wow. 753 00:27:39,200 --> 00:27:40,099 Bay routers. 754 00:27:40,559 --> 00:27:43,519 Does anybody remember Bay routers? The user interface 755 00:27:43,519 --> 00:27:45,859 on Bay routers was the SNMP tree. 756 00:27:46,845 --> 00:27:48,704 You literally had to go in and 757 00:27:49,325 --> 00:27:50,545 query the SNMP 758 00:27:52,125 --> 00:27:54,605 nodes in the tree and change them through 759 00:27:54,605 --> 00:27:55,345 a CLI 760 00:27:55,804 --> 00:27:58,704 to change the configuration. There was no configuration 761 00:27:58,924 --> 00:28:00,865 language or anything. Anyway, whatever. 762 00:28:01,829 --> 00:28:03,509 And they were having bugs with it, and 763 00:28:03,509 --> 00:28:04,650 Bay was out of business. 764 00:28:05,909 --> 00:28:08,490 And they were like, we can't justify 765 00:28:09,109 --> 00:28:10,569 replacing these routers 766 00:28:11,509 --> 00:28:13,049 for another five years, 767 00:28:13,429 --> 00:28:15,494 but we need somebody to support them. 768 00:28:16,295 --> 00:28:18,295 If we got you the source code, could 769 00:28:18,295 --> 00:28:19,035 you support 770 00:28:19,494 --> 00:28:21,275 these bay routers for us? 771 00:28:22,375 --> 00:28:23,194 We were like, 772 00:28:24,134 --> 00:28:26,375 no. And the sales rep said for a 773 00:28:26,375 --> 00:28:27,595 fee. Yeah. 774 00:28:30,289 --> 00:28:31,669 You just set your price 775 00:28:32,130 --> 00:28:35,029 higher than the cost of replacing them. Right. 776 00:28:37,730 --> 00:28:39,409 You know? I don't know what to say. 777 00:28:39,409 --> 00:28:41,349 It's those types of things happen. 778 00:28:42,049 --> 00:28:43,669 But, yeah, end of life 779 00:28:44,049 --> 00:28:44,710 to me 780 00:28:45,295 --> 00:28:47,454 can from like you said, Yvonne, from a 781 00:28:47,454 --> 00:28:49,934 from a vendor side, it means I'm not 782 00:28:49,934 --> 00:28:51,454 gonna support this. I want you to buy 783 00:28:51,454 --> 00:28:52,355 the new generation. 784 00:28:53,375 --> 00:28:56,095 And there are valid reasons for coming to 785 00:28:56,095 --> 00:28:57,555 that conclusion as a vendor. 786 00:28:59,420 --> 00:29:00,559 On the other side, 787 00:29:01,259 --> 00:29:02,880 end of life can mean, 788 00:29:03,900 --> 00:29:06,400 for you as a network operator, 789 00:29:07,420 --> 00:29:10,059 that equipment or that software is end of 790 00:29:10,059 --> 00:29:10,559 life. 791 00:29:11,259 --> 00:29:13,039 It may be supportable. 792 00:29:13,945 --> 00:29:15,705 There may be a vendor that will support 793 00:29:15,705 --> 00:29:18,765 it, but we're not deploying more of it. 794 00:29:18,984 --> 00:29:21,384 You can flip the tables on it and 795 00:29:21,384 --> 00:29:21,884 say, 796 00:29:22,265 --> 00:29:24,265 I don't care what the vendor's doing. We're 797 00:29:24,265 --> 00:29:26,205 end of lifing that in our network. 798 00:29:27,359 --> 00:29:29,359 And that's more of what people think of 799 00:29:29,359 --> 00:29:31,680 legacy as. I I think there's an 800 00:29:32,640 --> 00:29:35,440 there's also, a software aspect to this. A 801 00:29:35,440 --> 00:29:36,480 lot of people, when they think of end 802 00:29:36,480 --> 00:29:38,799 of blights, they think of hardware failing, which 803 00:29:38,799 --> 00:29:40,880 is an important part of the equation. And 804 00:29:40,880 --> 00:29:43,204 if you can't get that that chip anymore, 805 00:29:44,065 --> 00:29:46,224 to repair it and you don't have, any 806 00:29:46,224 --> 00:29:48,144 units on the on the shelf to replace, 807 00:29:48,144 --> 00:29:50,304 then obviously, I mean, that's that's I think 808 00:29:50,304 --> 00:29:51,744 a lot of people think of hardware as 809 00:29:51,744 --> 00:29:53,025 as that, but there's also, I think, a 810 00:29:53,025 --> 00:29:55,200 really important software conversation about this. 811 00:29:56,079 --> 00:29:57,859 This as a software supply chain 812 00:29:58,480 --> 00:29:59,619 speeds up and accelerates, 813 00:30:00,720 --> 00:30:02,819 you know, just keeping up with things and 814 00:30:03,200 --> 00:30:04,720 how much software is out there and how 815 00:30:04,720 --> 00:30:06,819 many dependencies every project has. 816 00:30:07,440 --> 00:30:09,200 It can get it can get really crazy 817 00:30:09,200 --> 00:30:11,035 and it can get I feel like it's 818 00:30:11,035 --> 00:30:12,315 a lot faster now than it used to 819 00:30:12,315 --> 00:30:14,654 be that you go to upgrade a dependency 820 00:30:14,795 --> 00:30:16,795 and then, you know, four others aren't ready 821 00:30:16,795 --> 00:30:19,035 yet. And so it's the whole thing is 822 00:30:19,035 --> 00:30:20,474 sort of stuck. And then by the time 823 00:30:20,474 --> 00:30:22,599 you get all those other four other all 824 00:30:22,599 --> 00:30:24,519 those other four already, now you have to 825 00:30:24,519 --> 00:30:27,259 leave the old ones behind because there's incompatibilities. 826 00:30:27,639 --> 00:30:28,700 Like, there's a lot of, 827 00:30:29,240 --> 00:30:31,240 software supply chain sort of things, I think, 828 00:30:31,240 --> 00:30:32,779 that are becoming more complex, 829 00:30:33,319 --> 00:30:34,700 sort of requiring us, 830 00:30:35,005 --> 00:30:37,484 whether it's more secure or not, requiring us 831 00:30:37,484 --> 00:30:39,565 to stay closer to the front edge, the 832 00:30:39,565 --> 00:30:41,984 the leading edge of of development these days. 833 00:30:42,684 --> 00:30:45,184 When we talk about security, there's also, 834 00:30:45,804 --> 00:30:48,204 you know, encryption and the ciphers that were 835 00:30:48,204 --> 00:30:50,250 deployed on that box and whether the hardware 836 00:30:50,549 --> 00:30:53,670 can support the new methods of encryption and 837 00:30:53,670 --> 00:30:55,769 all and all of those things which become 838 00:30:55,990 --> 00:30:57,210 incredibly important, 839 00:30:58,710 --> 00:31:01,755 in in our totally connected world. You know, 840 00:31:01,835 --> 00:31:03,355 you you've got to be able to patch. 841 00:31:03,355 --> 00:31:05,694 You've gotta be able to secure your environment, 842 00:31:05,755 --> 00:31:08,634 and those things also become critically important. And 843 00:31:08,634 --> 00:31:09,134 that's 844 00:31:09,434 --> 00:31:12,095 more a function of software than hardware, 845 00:31:12,555 --> 00:31:13,055 although, 846 00:31:13,595 --> 00:31:15,355 you know, if if your hardware has not 847 00:31:15,355 --> 00:31:17,194 got the the power to support it, then 848 00:31:17,194 --> 00:31:18,440 that's a related 849 00:31:19,059 --> 00:31:20,819 issue. So the third one, which is tech 850 00:31:20,819 --> 00:31:23,399 debt. And to me, this is the most 851 00:31:23,460 --> 00:31:23,960 misused 852 00:31:25,220 --> 00:31:25,720 phrase 853 00:31:26,740 --> 00:31:28,359 in all of the tech world. 854 00:31:29,380 --> 00:31:31,380 The way I think people use tech debt 855 00:31:31,380 --> 00:31:31,880 is 856 00:31:32,265 --> 00:31:34,525 some combination of legacy and EOL. 857 00:31:36,984 --> 00:31:38,605 Tech debt is old stuff. 858 00:31:39,224 --> 00:31:41,724 You know, it's basically equivalent to legacy, 859 00:31:42,105 --> 00:31:43,884 but it's all the bad legacy. 860 00:31:45,950 --> 00:31:47,549 And that, to me, is not what tech 861 00:31:47,549 --> 00:31:49,150 debt is. Of course, I'm I'm looking at 862 00:31:49,150 --> 00:31:50,930 it from a purely coding perspective. 863 00:31:51,630 --> 00:31:53,710 My definition of tech debt comes out of 864 00:31:53,710 --> 00:31:55,809 programming. It doesn't come out of networking. 865 00:31:56,430 --> 00:31:58,349 But when people say tech debt, they mostly 866 00:31:58,349 --> 00:32:00,029 mean they seem to mostly mean when I 867 00:32:00,029 --> 00:32:02,005 hear them use it. They say, 868 00:32:02,384 --> 00:32:03,125 well, that 869 00:32:03,505 --> 00:32:05,585 that router is five years old. That's tech 870 00:32:05,585 --> 00:32:06,085 debt. 871 00:32:06,545 --> 00:32:08,705 That needs to be replaced because it's tech 872 00:32:08,705 --> 00:32:09,205 debt. 873 00:32:10,705 --> 00:32:13,125 What tech debt really means to me 874 00:32:13,529 --> 00:32:16,650 is the differential between my understanding of how 875 00:32:16,650 --> 00:32:17,710 the system works 876 00:32:18,090 --> 00:32:19,549 and how it actually works. 877 00:32:21,210 --> 00:32:22,269 Tech debt is 878 00:32:22,809 --> 00:32:24,029 my level of 879 00:32:25,130 --> 00:32:28,025 the network works a certain way, but because 880 00:32:28,025 --> 00:32:30,285 things have changed in the network over time, 881 00:32:30,585 --> 00:32:33,164 no one really understands how it works anymore. 882 00:32:33,865 --> 00:32:36,505 That is tech debt. Because you go into 883 00:32:36,505 --> 00:32:37,804 a situation troubleshooting 884 00:32:38,585 --> 00:32:39,724 or making changes, 885 00:32:40,105 --> 00:32:42,430 and you think it works one way, and 886 00:32:42,430 --> 00:32:44,670 it actually works a different way. So you 887 00:32:44,670 --> 00:32:46,529 push the red button on the right, 888 00:32:46,990 --> 00:32:48,670 and it turns out that it makes the 889 00:32:48,670 --> 00:32:50,670 car go to the right instead of the 890 00:32:50,670 --> 00:32:51,170 left 891 00:32:51,549 --> 00:32:53,390 or whatever it is. You turn the wheel 892 00:32:53,390 --> 00:32:55,410 to the right, and unknown to you, 893 00:32:56,355 --> 00:32:58,934 somebody reversed the gears in the steering box 894 00:32:59,555 --> 00:33:01,255 for some reason down the road. 895 00:33:02,035 --> 00:33:04,515 That to me is tech debt. It's not 896 00:33:04,515 --> 00:33:06,214 old stuff. It's not legacy. 897 00:33:07,075 --> 00:33:09,795 Something that's totally brand new can be chock 898 00:33:09,795 --> 00:33:11,015 full of tech debt. 899 00:33:11,710 --> 00:33:13,410 I think there's also you, 900 00:33:13,870 --> 00:33:15,809 you're you're implying sort of an unintentional 901 00:33:16,110 --> 00:33:17,170 state imposed, 902 00:33:18,110 --> 00:33:20,590 by virtue of not understanding the system. I 903 00:33:20,590 --> 00:33:22,990 I think there's also, I think there's also 904 00:33:22,990 --> 00:33:23,650 an intentional 905 00:33:24,184 --> 00:33:25,065 flavor to this, 906 00:33:25,704 --> 00:33:28,125 when you're, specifically when you're building software. 907 00:33:28,904 --> 00:33:30,345 You know, if you have if you have 908 00:33:30,345 --> 00:33:32,984 the experience to anticipate certain things but decide 909 00:33:32,984 --> 00:33:33,724 not to, 910 00:33:34,265 --> 00:33:35,085 if you 911 00:33:35,464 --> 00:33:37,304 you you you a lot a lot of 912 00:33:37,537 --> 00:33:38,809 like, when I think of tech debt, I 913 00:33:38,809 --> 00:33:41,210 think of the more intentional type, like like, 914 00:33:41,210 --> 00:33:43,049 physically a credit card. When you take out 915 00:33:43,049 --> 00:33:44,410 a credit card and swipe it in the 916 00:33:44,410 --> 00:33:46,650 terminal terminal, you're saying, I wanna buy now 917 00:33:46,650 --> 00:33:48,970 and pay later. And I think the, like, 918 00:33:48,970 --> 00:33:50,410 the one flavor of this I think what 919 00:33:50,410 --> 00:33:51,450 you said is really important to us, but 920 00:33:51,450 --> 00:33:53,664 I think another flavor of it is, making 921 00:33:53,664 --> 00:33:54,724 the decision today 922 00:33:55,105 --> 00:33:57,025 to do something at the expense of effort 923 00:33:57,025 --> 00:33:58,484 that will be required tomorrow. 924 00:33:59,105 --> 00:34:02,144 And it's not tech technical debt is not 925 00:34:02,144 --> 00:34:03,365 a moral judgment. 926 00:34:03,825 --> 00:34:05,744 It is, it's no more moral judgment than 927 00:34:05,744 --> 00:34:07,105 the judgment to use a credit card and 928 00:34:07,105 --> 00:34:08,139 pay it off at the end the month, 929 00:34:08,139 --> 00:34:10,140 but I think it gets really emotionally sort 930 00:34:10,140 --> 00:34:12,300 of, loaded with us when we think when 931 00:34:12,300 --> 00:34:14,139 we think about it. Well and I was 932 00:34:14,139 --> 00:34:16,619 thinking too as Russ was talking that when 933 00:34:16,619 --> 00:34:19,099 we're talking about tech debt, that's the what 934 00:34:19,099 --> 00:34:21,900 trade offs have we made. Right? We we 935 00:34:21,900 --> 00:34:22,880 made a trade 936 00:34:23,434 --> 00:34:26,014 to either do something quickly or imperfectly 937 00:34:26,474 --> 00:34:30,175 or without the proper engineering rigor or whatever 938 00:34:30,234 --> 00:34:30,894 it is. 939 00:34:31,594 --> 00:34:32,094 And 940 00:34:33,275 --> 00:34:33,775 now 941 00:34:34,074 --> 00:34:36,815 the interest that we pay on that 942 00:34:37,269 --> 00:34:38,090 tech debt 943 00:34:38,550 --> 00:34:39,610 is complexity, 944 00:34:40,470 --> 00:34:41,369 is slowness, 945 00:34:41,750 --> 00:34:42,809 is downtime, 946 00:34:43,750 --> 00:34:44,250 is, 947 00:34:44,710 --> 00:34:46,090 you know, all of these 948 00:34:47,510 --> 00:34:48,010 results 949 00:34:48,390 --> 00:34:50,490 of a less elegant solution. 950 00:34:50,925 --> 00:34:52,945 And we and we we made that trade. 951 00:34:53,485 --> 00:34:55,885 What happens though, and I think the reason 952 00:34:55,885 --> 00:34:57,824 we talk about tech debt so much, 953 00:34:58,204 --> 00:34:59,585 is because often 954 00:35:00,364 --> 00:35:01,744 we make those trades 955 00:35:02,204 --> 00:35:03,985 without realizing it. 956 00:35:04,480 --> 00:35:06,660 You know, we we signed the paperwork 957 00:35:07,519 --> 00:35:09,199 to buy a $20,000 958 00:35:09,199 --> 00:35:10,800 car, but we didn't do the math and 959 00:35:10,800 --> 00:35:13,599 realized that we're actually gonna be paying $40,000 960 00:35:13,599 --> 00:35:15,860 for that car over the next five years. 961 00:35:16,239 --> 00:35:18,900 Right? Nobody actually did the math and understood 962 00:35:18,960 --> 00:35:20,994 the trade off. And then at some some 963 00:35:20,994 --> 00:35:22,914 point, you're like, I only have this, you 964 00:35:22,914 --> 00:35:24,914 know, I only have $25,000 965 00:35:24,914 --> 00:35:27,315 worth of stuff. Why am I spending 40,000? 966 00:35:27,315 --> 00:35:29,155 And and eventually, you just don't have the 967 00:35:29,155 --> 00:35:30,695 resources to do anything 968 00:35:31,394 --> 00:35:34,090 because you you're paying the debt on all 969 00:35:34,090 --> 00:35:35,929 these trade offs that you've made. And that's 970 00:35:35,929 --> 00:35:37,929 why we talk about it so much. But 971 00:35:37,929 --> 00:35:39,309 it is a much more nebulous 972 00:35:40,809 --> 00:35:41,309 category. 973 00:35:41,769 --> 00:35:43,949 So in software world, we say 974 00:35:44,329 --> 00:35:46,010 that to get rid of tech debt, we 975 00:35:46,010 --> 00:35:46,510 refactor, 976 00:35:47,885 --> 00:35:49,744 which means you tear things apart. 977 00:35:50,204 --> 00:35:51,885 You try to go back to the basis 978 00:35:51,885 --> 00:35:53,505 of what was trying to be accomplished, 979 00:35:54,445 --> 00:35:55,984 and you rebuild the software 980 00:35:56,605 --> 00:35:58,545 to fit what you were trying to accomplish. 981 00:35:59,180 --> 00:36:01,420 And oftentimes, you discover that there are hidden 982 00:36:01,420 --> 00:36:04,079 requirements you didn't know about in the process. 983 00:36:04,460 --> 00:36:05,980 If somebody's made this change to do this 984 00:36:05,980 --> 00:36:08,239 or that, and nobody wrote it down, 985 00:36:08,619 --> 00:36:10,860 bah, you know, six years from now or 986 00:36:10,860 --> 00:36:12,700 whatever, they'll figure it out. And they do. 987 00:36:12,700 --> 00:36:14,239 They figure it out. Trust me. 988 00:36:14,655 --> 00:36:16,755 In the network world, I think about situations 989 00:36:16,894 --> 00:36:17,394 like, 990 00:36:18,015 --> 00:36:19,715 okay. I'm going to use 991 00:36:20,175 --> 00:36:23,155 VXLAN to solve this problem because it's 992 00:36:23,454 --> 00:36:26,255 cool, and my chipset supports it. I'm going 993 00:36:26,255 --> 00:36:26,914 to use 994 00:36:27,215 --> 00:36:29,074 MPLS to solve that problem. 995 00:36:29,869 --> 00:36:31,869 And I'm gonna use GRE to solve this 996 00:36:31,869 --> 00:36:32,369 problem. 997 00:36:32,829 --> 00:36:35,809 And I'm gonna use SRV six or MPLS 998 00:36:35,949 --> 00:36:37,809 SR to solve that problem. 999 00:36:38,349 --> 00:36:40,110 And before you know it, you have five 1000 00:36:40,110 --> 00:36:42,210 different tunnel mechanisms in your network. 1001 00:36:42,905 --> 00:36:44,505 How many tunnels does it take to make 1002 00:36:44,505 --> 00:36:45,085 a network? 1003 00:36:45,464 --> 00:36:47,484 All of them. All of them. Exactly. 1004 00:36:48,105 --> 00:36:48,605 Exactly. 1005 00:36:49,704 --> 00:36:52,364 And refactoring would be saying, alright. 1006 00:36:53,065 --> 00:36:53,885 Stop it. 1007 00:36:54,585 --> 00:36:56,445 I don't care which one you pick. 1008 00:36:57,170 --> 00:36:57,670 Pick 1009 00:36:58,130 --> 00:36:59,510 one and make it work, 1010 00:37:00,690 --> 00:37:03,090 and stop piling tunnel mechanisms on top of 1011 00:37:03,090 --> 00:37:03,829 this network. 1012 00:37:04,130 --> 00:37:04,630 Right? 1013 00:37:05,809 --> 00:37:06,789 That's refactoring. 1014 00:37:07,170 --> 00:37:07,829 It's like, 1015 00:37:08,530 --> 00:37:10,635 I have all this stuff I'm using. I 1016 00:37:10,635 --> 00:37:13,215 have five different ways to allocate IP addresses. 1017 00:37:13,675 --> 00:37:14,175 Really? 1018 00:37:14,954 --> 00:37:16,335 Stop that. Right? 1019 00:37:17,195 --> 00:37:19,114 There's gotta be one way to do this 1020 00:37:19,114 --> 00:37:20,954 that makes sense. I was gonna say one 1021 00:37:20,954 --> 00:37:22,474 of the things I think that I see 1022 00:37:22,474 --> 00:37:24,579 now with age and experience that I did 1023 00:37:24,579 --> 00:37:25,800 not see in my youth 1024 00:37:26,500 --> 00:37:28,280 is that we should never underestimate 1025 00:37:28,739 --> 00:37:30,679 the value of a working system. 1026 00:37:32,019 --> 00:37:32,679 My example 1027 00:37:33,139 --> 00:37:34,760 is I I'm a loyal 1028 00:37:35,300 --> 00:37:35,800 Sonos 1029 00:37:36,434 --> 00:37:39,875 user, Sonos speakers. And they, in the past 1030 00:37:39,875 --> 00:37:41,715 year, if you're not familiar with all the 1031 00:37:41,715 --> 00:37:43,574 drama, they rewrote their application. 1032 00:37:44,114 --> 00:37:46,434 But when they released their app, it was 1033 00:37:46,434 --> 00:37:46,934 not, 1034 00:37:48,195 --> 00:37:50,535 feature rich. It didn't have all the features. 1035 00:37:51,059 --> 00:37:54,019 They upgraded some software on their speakers, so 1036 00:37:54,019 --> 00:37:56,680 they couldn't go back to the old app. 1037 00:37:57,140 --> 00:37:58,840 And you and and 1038 00:37:59,780 --> 00:38:01,699 those of us who are users lost at 1039 00:38:01,699 --> 00:38:03,320 least 50% of the functionality 1040 00:38:03,860 --> 00:38:06,019 of our app, and that's how you controlled 1041 00:38:06,019 --> 00:38:06,680 your speakers. 1042 00:38:07,594 --> 00:38:10,074 And they under so they they went through 1043 00:38:10,074 --> 00:38:13,034 this process to say, okay. We're gonna start 1044 00:38:13,034 --> 00:38:15,114 over. We're gonna rebuild our app. It's gonna 1045 00:38:15,114 --> 00:38:16,815 allow us to do all these new things, 1046 00:38:17,194 --> 00:38:18,255 but they underestimated 1047 00:38:19,719 --> 00:38:21,960 all of the level of effort required to 1048 00:38:21,960 --> 00:38:23,739 get back to feature parity. 1049 00:38:24,199 --> 00:38:25,340 That's the risk 1050 00:38:25,880 --> 00:38:27,180 in starting over, 1051 00:38:27,880 --> 00:38:30,059 is that if you're not really clear 1052 00:38:31,079 --> 00:38:33,019 on all the features, the requirements 1053 00:38:35,105 --> 00:38:35,168 that you were talking about, Russ, and the 1054 00:38:35,168 --> 00:38:36,784 capabilities that you need, it's gonna take you 1055 00:38:36,784 --> 00:38:38,085 a lot more resources 1056 00:38:38,864 --> 00:38:40,164 to get back to parity 1057 00:38:40,625 --> 00:38:42,864 than what you originally thought, and that can 1058 00:38:42,864 --> 00:38:44,085 be a huge risk. 1059 00:38:44,465 --> 00:38:46,484 So you just need to be really sure 1060 00:38:46,500 --> 00:38:47,400 what you're undertaking 1061 00:38:47,860 --> 00:38:50,199 when you do that because that's you know, 1062 00:38:50,420 --> 00:38:52,760 it it becomes a you can get underwater 1063 00:38:52,900 --> 00:38:55,139 really, really quickly and end up in a 1064 00:38:55,139 --> 00:38:56,039 bad spot. 1065 00:38:56,579 --> 00:38:58,340 I think that pretty well defines tech debt. 1066 00:38:58,340 --> 00:38:59,699 I don't know, Tom. Did you have more 1067 00:38:59,699 --> 00:39:00,599 to add to that? 1068 00:39:00,905 --> 00:39:03,704 No. No. I think that's great. Cool. Awesome. 1069 00:39:03,704 --> 00:39:05,385 Well, I'm kind of out of questions and 1070 00:39:05,385 --> 00:39:05,885 topics. 1071 00:39:06,505 --> 00:39:07,784 I don't have anything else to add, so 1072 00:39:07,784 --> 00:39:09,565 I guess we'll just go around. And 1073 00:39:10,264 --> 00:39:12,664 I'll start with Yvonne because Yvonne has a 1074 00:39:12,664 --> 00:39:15,099 longer spiel. You can find me on LinkedIn 1075 00:39:15,099 --> 00:39:17,579 at Yvonne Sharp. I'm on x at Sharp 1076 00:39:17,579 --> 00:39:19,739 Network. I'm on Blue Sky at e sharp 1077 00:39:19,739 --> 00:39:20,400 dot net, 1078 00:39:20,940 --> 00:39:21,920 and writing, 1079 00:39:22,300 --> 00:39:24,300 periodically on Substack. So you can search for 1080 00:39:24,300 --> 00:39:25,840 me there too, Yvonne Sharp. 1081 00:39:26,219 --> 00:39:26,719 Awesome. 1082 00:39:27,099 --> 00:39:27,599 And 1083 00:39:28,380 --> 00:39:29,924 I don't know if you're gonna ask Tom. 1084 00:39:30,224 --> 00:39:31,125 Oh, come on. 1085 00:39:31,905 --> 00:39:33,684 Go ahead. Do it. LinkedIn. 1086 00:39:34,984 --> 00:39:35,484 That's 1087 00:39:36,784 --> 00:39:37,284 it. 1088 00:39:38,224 --> 00:39:40,385 That's it. Alright. And I'm Russ White. You 1089 00:39:40,385 --> 00:39:41,880 can always find me here at the hedge. 1090 00:39:42,039 --> 00:39:43,559 On rule11.tech, 1091 00:39:43,559 --> 00:39:46,119 on x as routing geek. Not that I 1092 00:39:46,119 --> 00:39:48,219 ever I don't log in there very often, 1093 00:39:48,279 --> 00:39:49,719 but I am I am there every now 1094 00:39:49,719 --> 00:39:50,299 and again. 1095 00:39:51,079 --> 00:39:51,900 On LinkedIn 1096 00:39:52,440 --> 00:39:52,940 and, 1097 00:39:54,199 --> 00:39:55,719 I don't know, packet pushers every now. I 1098 00:39:55,719 --> 00:39:57,465 don't know. I'm you'll find me around. I'm 1099 00:39:57,465 --> 00:39:58,684 pretty easy to find. 1100 00:39:59,065 --> 00:40:00,985 We know that we live in an attention 1101 00:40:00,985 --> 00:40:03,385 driven economy, and we thank you for listening 1102 00:40:03,385 --> 00:40:04,985 all the way to the bitter end of 1103 00:40:04,985 --> 00:40:07,224 this episode of The Hedge. And thanks for 1104 00:40:07,224 --> 00:40:08,925 joining us, and we will catch you next. 1105 00:40:17,980 --> 00:40:18,480 Next