WEBVTT 1 00:00:05.360 --> 00:00:09.080 Hey, welcome back to another episode of JavaScript Jabber. 2 00:00:09.359 --> 00:00:10.880 This week, on our panel, we have Dan. 3 00:00:10.720 --> 00:00:15.359 Shapier, Hello from Tel Aviv, which is actually cold, even 4 00:00:15.400 --> 00:00:17.719 though it's probably pretty warm by your standards. 5 00:00:18.320 --> 00:00:19.120 He says cold. 6 00:00:19.120 --> 00:00:21.559 It is below freezing here, or at least was this morning. 7 00:00:22.000 --> 00:00:23.879 And we also have a j O'Neill. 8 00:00:24.800 --> 00:00:27.160 Yo ya yo coming at you live from the wonderful, 9 00:00:27.160 --> 00:00:28.000 wonderful shed. 10 00:00:30.359 --> 00:00:32.799 I'm Charles Maxwood from Top End Devs. I am jealous 11 00:00:32.799 --> 00:00:33.719 of your shed. 12 00:00:34.240 --> 00:00:36.000 We have a special guest this week, and that is 13 00:00:36.119 --> 00:00:37.840 Avishai is Shalom. 14 00:00:38.079 --> 00:00:40.000 I hope I got close on that. I'm sorry if 15 00:00:40.000 --> 00:00:48.119 I didn't. So yeah, I've thought very deeply about putting 16 00:00:48.159 --> 00:00:49.399 something like that in the back of my house. 17 00:00:50.399 --> 00:00:54.479 Anyway, Dan, you invited Avishai here, and it seems like 18 00:00:54.640 --> 00:00:57.439 you know each other pretty well. So do you want 19 00:00:57.479 --> 00:01:00.439 to introduce him and then he can fill in the 20 00:01:00.479 --> 00:01:01.600 gaps for us? 21 00:01:01.920 --> 00:01:06.480 Yeah, well I do. I've known a Vishai for a while. 22 00:01:06.920 --> 00:01:10.200 I think we work together at Wicks a long time ago. 23 00:01:10.480 --> 00:01:13.959 Correct me if I'm wrong. And I've wanted to have 24 00:01:14.000 --> 00:01:16.079 a visha on or show for a while now, and 25 00:01:16.120 --> 00:01:19.480 I'm glad that he's finally able to join us. The 26 00:01:19.599 --> 00:01:23.959 reason for this is that, similarly to Tomer Gabel, we 27 00:01:24.040 --> 00:01:26.519 had on the show a few weeks so good. Yeah, 28 00:01:26.599 --> 00:01:28.719 yes he was, We had him on the show a 29 00:01:28.719 --> 00:01:33.040 few weeks ago, so similar to him, Avishai is one 30 00:01:33.079 --> 00:01:36.920 of those people who when I communicate with often it 31 00:01:37.000 --> 00:01:40.400 forces me to rethink or even reconsider some of my 32 00:01:40.480 --> 00:01:45.599 technological assumptions and even core beliefs. Things that I hold 33 00:01:45.680 --> 00:01:49.640 to be obviously true suddenly don't seem so true anymore 34 00:01:50.239 --> 00:01:55.560 thanks to Avishai. And you know, it doesn't happen every day, 35 00:01:55.760 --> 00:01:59.239 but it does happen, and I have to say that 36 00:02:00.079 --> 00:02:04.519 to be honest, this isn't always a comfortable experience. You know, 37 00:02:04.640 --> 00:02:08.080 we as people, as humans, we dislike having our viewpoints 38 00:02:08.159 --> 00:02:12.400 challenged in such ways. Even as engineers, we tend to 39 00:02:12.439 --> 00:02:17.199 be creatures of habit. You know, what we use, what 40 00:02:17.240 --> 00:02:19.479 we like to do, we continue doing the same thing, 41 00:02:19.560 --> 00:02:22.840 and you know, we prefer to stay the course as 42 00:02:22.879 --> 00:02:27.919 it were. We prefer complacency over disruption, even when we, 43 00:02:28.240 --> 00:02:32.599 you know, proclaim otherwise. But I think it's really important 44 00:02:32.599 --> 00:02:36.000 for us to go through such experiences because that's how 45 00:02:36.039 --> 00:02:40.400 we grow. And when you have some such interactions, you know, 46 00:02:40.639 --> 00:02:43.840 you grow or the other party grows, or ideally you 47 00:02:43.960 --> 00:02:48.280 both grow as a result of such engagements. Now, I 48 00:02:48.400 --> 00:02:51.520 recently learned that a few of our guests that we 49 00:02:51.599 --> 00:02:54.919 had on the show apparently did not have a good time, 50 00:02:56.159 --> 00:03:00.360 and that's really unfortunate. They felt a bit put upon 51 00:03:00.560 --> 00:03:04.000 when we ask them questions that require them to justify 52 00:03:04.159 --> 00:03:08.800 technical decisions that they made, and when we presented alternatives 53 00:03:08.840 --> 00:03:11.919 to solutions they came up with, they felt that, like, 54 00:03:12.080 --> 00:03:16.000 you know, we were kind of attacking them. Obviously, we 55 00:03:16.039 --> 00:03:18.719 want our guests to have a great time, and I 56 00:03:18.800 --> 00:03:22.800 think that most of them definitely do, and I hope 57 00:03:22.840 --> 00:03:24.960 Avisha that you will have a good time as well, 58 00:03:25.639 --> 00:03:29.520 we will say, but more importantly than that, we want 59 00:03:29.560 --> 00:03:33.479 our audience to gain actual value from the show. We 60 00:03:33.560 --> 00:03:37.719 want our audience to learn something that actually benefits them 61 00:03:37.879 --> 00:03:43.560 in their career as engineers. And sometimes that requires, you know, 62 00:03:44.800 --> 00:03:49.719 asking harder questions, challenging our guests on those decisions that 63 00:03:49.759 --> 00:03:53.520 they've made, and if it does make them feel uncomfortable, 64 00:03:53.680 --> 00:03:58.599 well sorry, but not sorry, because it's really a part 65 00:03:58.639 --> 00:04:02.360 of what it is to, you know, on a podcast 66 00:04:02.479 --> 00:04:06.319 such as ours. I guess that there are others where 67 00:04:06.400 --> 00:04:10.360 these things don't happen. You know, fine, I love a 68 00:04:10.360 --> 00:04:14.000 lot of other technical podcasts. You know, we have different emphasies, 69 00:04:14.879 --> 00:04:17.879 but this is how we do things. So you know, 70 00:04:18.279 --> 00:04:21.480 if you want to come on our show, you need 71 00:04:21.519 --> 00:04:25.519 to be ready to answer some tougher questions. It's just 72 00:04:25.639 --> 00:04:28.600 the way that we do do things. And that's what 73 00:04:28.639 --> 00:04:29.439 I wanted to say. 74 00:04:30.399 --> 00:04:32.920 Yeah, I can give an example. You brought up Tomer 75 00:04:33.399 --> 00:04:36.120 Gabble when he was on the show, and you know, 76 00:04:36.319 --> 00:04:38.800 we had a lot of back and forth on like types, 77 00:04:38.959 --> 00:04:43.800 strong types, type systems, and you know, it was funny 78 00:04:43.800 --> 00:04:44.839 because I was thinking about it. 79 00:04:44.839 --> 00:04:46.519 And then I watched the talk because you know. 80 00:04:46.480 --> 00:04:48.800 I live a lot of my programming life in the 81 00:04:48.839 --> 00:04:52.399 Ruby world, and it was a discussion between David Heinemeier 82 00:04:52.439 --> 00:04:57.959 Hansen and Tobias or Toby Luki from from Shopify, and 83 00:04:58.279 --> 00:05:01.560 they were having that same discussion, right, and it was 84 00:05:01.639 --> 00:05:04.800 interesting just to see kind of the trade offs that 85 00:05:04.800 --> 00:05:05.639 they were talking about. 86 00:05:05.720 --> 00:05:07.319 And then I wish that I could go back on 87 00:05:07.360 --> 00:05:08.959 that episode and say, so I. 88 00:05:08.920 --> 00:05:12.519 Think you're right about this and not you know, and right, 89 00:05:12.600 --> 00:05:16.199 and I got better informed about maybe I'm not running 90 00:05:16.240 --> 00:05:18.319 into the things that bring that trade off. And I 91 00:05:18.360 --> 00:05:20.279 think the kinds of questions and back and forth that 92 00:05:20.279 --> 00:05:22.639 you're talking about, Dan are the things that allow us 93 00:05:22.680 --> 00:05:25.879 to clarify that and open it up and go, Okay, 94 00:05:26.439 --> 00:05:28.399 these are the problems you're trying to solve, not these 95 00:05:28.439 --> 00:05:30.959 other ones, or these are the you know, this is 96 00:05:30.959 --> 00:05:33.399 where the trade offs really start to click into place, 97 00:05:33.639 --> 00:05:36.519 because what I find is that it's not always black 98 00:05:36.560 --> 00:05:38.480 and white up and down the board. It may just 99 00:05:38.560 --> 00:05:41.680 be your experience leads you to make this decision where 100 00:05:42.240 --> 00:05:45.279 I'm just not working in a space where that matters, 101 00:05:45.600 --> 00:05:48.040 or maybe I just never considered that I needed to 102 00:05:48.040 --> 00:05:49.720 think about it that way, and you push me to 103 00:05:49.759 --> 00:05:50.439 do better. 104 00:05:50.240 --> 00:05:51.639 So I agree with you. 105 00:05:51.720 --> 00:05:54.480 I think I think some of that you know, back 106 00:05:54.519 --> 00:05:56.759 and forth, and hey, it looks like this solves a 107 00:05:56.839 --> 00:05:59.920 problem better than that. And you know, we're going to 108 00:06:00.079 --> 00:06:03.319 give people the tools to go, oh, yes I need that, 109 00:06:03.519 --> 00:06:06.160 I you know, or no, I'm not quite to the 110 00:06:06.160 --> 00:06:08.040 place where I need that. But then you be thinking 111 00:06:08.040 --> 00:06:10.199 about it as I get there or there, or maybe hey, 112 00:06:10.240 --> 00:06:13.959 you're talking to another group of people that think about 113 00:06:13.959 --> 00:06:15.160 things in a different way than I do. 114 00:06:15.480 --> 00:06:17.360 And all of that is constructive. 115 00:06:20.879 --> 00:06:22.319 Any thoughts of you shine. 116 00:06:23.279 --> 00:06:25.600 But I can tell you that when I'm using static types, 117 00:06:25.800 --> 00:06:29.680 like I'm enjoying myself so much, and I'm like, what 118 00:06:29.759 --> 00:06:32.160 idiot would use dynamic types? And then when I'm using 119 00:06:32.279 --> 00:06:36.439 language that it's me. It's me right way. So when 120 00:06:36.480 --> 00:06:39.519 I'm using a closure it's my favorite language, I'm enjoying 121 00:06:39.560 --> 00:06:41.480 myself so much, and I'm like, what idiot would use 122 00:06:41.519 --> 00:06:46.639 static types? So it's actually like, once you get your 123 00:06:46.680 --> 00:06:50.040 mind into a specific style of programming, it feels very 124 00:06:50.120 --> 00:06:52.399 natural to you. It's very easy to forget all the 125 00:06:52.439 --> 00:06:55.079 advantages at Allstein has so. 126 00:06:58.079 --> 00:06:58.360 Cool. 127 00:06:58.399 --> 00:07:00.879 Well let's talk about marshmallows. 128 00:07:01.680 --> 00:07:03.879 Wait a minute. First of all of Isshai, I said 129 00:07:03.920 --> 00:07:05.959 that we worked together at Wicks. Can can you tell 130 00:07:06.040 --> 00:07:07.399 us a little bit about yourself? 131 00:07:08.319 --> 00:07:12.319 Yes? I can. So I did a lot of wild stuff, 132 00:07:12.319 --> 00:07:16.839 I guess. I started my career as a physicist, not 133 00:07:16.839 --> 00:07:19.160 not really in it, and then at some point I 134 00:07:19.199 --> 00:07:22.079 discovered that I'm more interested in how fast my A 135 00:07:22.120 --> 00:07:25.959 simulations are running more than the actual results. So I 136 00:07:26.040 --> 00:07:30.720 kind of switched into into it. Initially, I went into systems. 137 00:07:31.480 --> 00:07:33.800 I did that for a while, and then I founded 138 00:07:33.839 --> 00:07:36.839 the consulting company and I worked on a lot of 139 00:07:37.000 --> 00:07:41.199 different stuff, a part of lost large scale architecture and 140 00:07:41.360 --> 00:07:47.519 performance stuff and some training of engineers, and at some 141 00:07:47.519 --> 00:07:49.720 point I stumbled into big data. It was like the 142 00:07:49.759 --> 00:07:52.959 big doup hype back in the day. So I spent 143 00:07:53.000 --> 00:07:55.680 about two years, you know, getting people off a dupe 144 00:07:55.800 --> 00:07:59.680 because they didn't have enough data, you know, the whole 145 00:08:00.079 --> 00:08:05.639 data fits in rum thing. And after that I went 146 00:08:05.839 --> 00:08:10.120 to work at Wicks. I did. I had the back 147 00:08:10.199 --> 00:08:14.079 end team at Wicks and from Wiks I went to 148 00:08:14.079 --> 00:08:17.920 work for a VC. That's quite a co change, I guess. Uh. 149 00:08:18.319 --> 00:08:21.120 So I was in in the residents at the Alish VC. 150 00:08:21.279 --> 00:08:24.680 It's one of the prominent vcs in as well. And 151 00:08:25.040 --> 00:08:27.160 after that they went to work for CLDB, which is 152 00:08:27.160 --> 00:08:31.319 a high performance Cassandra clone written C plus plus. Very 153 00:08:31.639 --> 00:08:34.039 unique and interesting thing if you want to talk about that. 154 00:08:35.279 --> 00:08:40.120 And since then I've been an independent of social Now 155 00:08:40.320 --> 00:08:44.120 nowadays I'm mostly working on performance problems and training engineers, 156 00:08:44.120 --> 00:08:47.480 which is my con passion. I think that now that 157 00:08:47.519 --> 00:08:50.440 we have the ai AH, we have a very big 158 00:08:50.480 --> 00:08:53.519 problem because everything's too easy, so you don't actually learn 159 00:08:53.600 --> 00:08:57.440 how stuff works. So the funnel of becoming a senior 160 00:08:57.519 --> 00:09:01.320 from from a junior is now gone. If you think 161 00:09:01.320 --> 00:09:04.360 about network engineers, for example, how did you train to 162 00:09:04.360 --> 00:09:06.480 be a network engineer? Back in the day, you had 163 00:09:06.480 --> 00:09:09.399 your home auto and you played with it, and you 164 00:09:09.480 --> 00:09:12.159 needed to let's say, you know, define the DHP and 165 00:09:12.200 --> 00:09:15.159 the DNS and what stuff and nat if you had 166 00:09:15.360 --> 00:09:20.360 some kind of home auto and you learned doing that. Nowadays, 167 00:09:20.399 --> 00:09:22.720 you just get Google Mesh and you install it and 168 00:09:22.840 --> 00:09:26.240 everything works. So there's no reason if you'd fiddle with 169 00:09:26.279 --> 00:09:29.200 it and actually learned networks. So there's no funnel for 170 00:09:29.559 --> 00:09:32.679 network engineers anymore. So it's a problem I'm trying to 171 00:09:32.679 --> 00:09:33.200 help solve. 172 00:09:35.240 --> 00:09:37.679 You know. I'm thinking about all, you know, all the 173 00:09:37.720 --> 00:09:42.240 American TV shows where you have the father lifting up 174 00:09:42.279 --> 00:09:46.759 the hood of the car and showing their son or 175 00:09:46.840 --> 00:09:50.600 daughter how it actually works, how it runs. And I'm 176 00:09:50.600 --> 00:09:54.679 thinking about all the new electrical cars that's pretty much gone. 177 00:09:56.240 --> 00:09:59.559 Yeah, I had experienced and motorcycles. I had one of 178 00:09:59.600 --> 00:10:04.240 the last motorcycles that still had colpools. And then the 179 00:10:04.320 --> 00:10:09.559 fulling models already have the fuel injections. Fuel injection, yeah, yeah, exactly. 180 00:10:09.600 --> 00:10:13.000 And you can't you can't tune it, you can't fix it. 181 00:10:13.080 --> 00:10:14.399 So the day. 182 00:10:16.000 --> 00:10:20.480 Yeah, I've never heard anyone argue against fuel injection. 183 00:10:21.919 --> 00:10:27.679 I've never but I mean the trade off. Yeah, it's 184 00:10:27.720 --> 00:10:31.159 it's more fuel efficient, it has no performance, it's no reliable. 185 00:10:31.840 --> 00:10:33.879 But the downside is that you don't get to learn 186 00:10:34.120 --> 00:10:36.799 how motorcycles work anymore. You can. 187 00:10:38.399 --> 00:10:39.399 Controlled by the computer. 188 00:10:39.840 --> 00:10:42.559 Yeah, exactly, there's no no. 189 00:10:42.159 --> 00:10:43.840 Mechanical management of it. 190 00:10:45.399 --> 00:10:49.320 That's that's interesting because I so, I just got a motorcycle, 191 00:10:49.399 --> 00:10:51.519 and I'm actually glad that I got one that you know, 192 00:10:51.519 --> 00:10:53.519 I got an old one, so it was cheap, but 193 00:10:53.759 --> 00:10:56.320 it was just new enough that it was fuel injected, 194 00:10:57.120 --> 00:11:00.279 and which motorcycles got fuel injected much later than car did. 195 00:11:00.320 --> 00:11:01.919 I mean a lot of four wheelers I think are 196 00:11:01.919 --> 00:11:05.480 still carbureted. But yeah, I kind of I kind of 197 00:11:05.519 --> 00:11:07.200 felt like I dodged the bullet. I still feel like 198 00:11:07.240 --> 00:11:09.759 I'm learning about it, but I didn't. I didn't have 199 00:11:09.840 --> 00:11:15.519 that sense of understanding the carburetor is is key to 200 00:11:16.080 --> 00:11:22.679 understanding the maintenance of the motorcycle. Set's I'm yeah, I'm really, 201 00:11:22.720 --> 00:11:24.960 I'm really interested that you said that. I'm I have 202 00:11:25.039 --> 00:11:27.039 to sit on that for a minute. I wonder what 203 00:11:27.080 --> 00:11:28.799 you know that I don't know that makes you feel 204 00:11:28.799 --> 00:11:30.000 that way obviously something. 205 00:11:30.639 --> 00:11:33.519 So I'll give you an example. You live in a 206 00:11:33.519 --> 00:11:36.080 colder climate, right, so you need to heat up your 207 00:11:36.120 --> 00:11:39.120 motorcycle for a little while before you start gliding it. 208 00:11:39.600 --> 00:11:42.000 And the reason for that is because of the cowboy 209 00:11:42.039 --> 00:11:47.000 tool and because you actually need heat for the fuel 210 00:11:47.039 --> 00:11:50.480 to spray in, you know, to the mixture. So did 211 00:11:50.480 --> 00:11:52.279 you get you actually get it into the engine and 212 00:11:52.320 --> 00:11:56.159 it blows up? Was with you know, the injection system? 213 00:11:56.240 --> 00:11:58.559 You don't need to do that. The computer text out 214 00:11:58.600 --> 00:12:03.759 of it. So whole like even this bit of thinking 215 00:12:03.799 --> 00:12:05.960 for pausing for a second thinking, Okay, I need to 216 00:12:06.000 --> 00:12:08.720 warm up my car because you know call the climate 217 00:12:09.000 --> 00:12:13.440 fuel mixture that stuff that's gone. You're not even aware 218 00:12:13.440 --> 00:12:14.120 of it anymore. 219 00:12:15.480 --> 00:12:19.159 I so my mine is fuel injected. It's just new 220 00:12:19.240 --> 00:12:22.159 enough to be fuel injected. But it does it does 221 00:12:22.320 --> 00:12:26.519 definitely behave differently like I can tell. Yeah, I do 222 00:12:26.600 --> 00:12:28.200 get a sense of it. But yeah, the four wheeler, 223 00:12:28.200 --> 00:12:30.519 I actually have to actually have to tune the carburetor 224 00:12:30.519 --> 00:12:32.679 of the four wheeler so it'll start in the winter 225 00:12:33.279 --> 00:12:34.720 and not flood in the summer. 226 00:12:37.159 --> 00:12:40.159 I'm still thinking about the fact that you define yourself 227 00:12:40.200 --> 00:12:44.080 as a researcher, and I'm curious what that means and 228 00:12:44.120 --> 00:12:45.759 how that translates to real life. 229 00:12:47.039 --> 00:12:49.720 So a lot of the stuff they do right now 230 00:12:50.039 --> 00:12:52.480 is first of all, a lot of reading, but also 231 00:12:52.480 --> 00:12:57.200 a lot of observing and experimenting. So some of it 232 00:12:57.039 --> 00:13:00.200 is like on how systems work, and some of it 233 00:13:00.240 --> 00:13:04.080 is into models. And I'm saying models specifically because it 234 00:13:04.120 --> 00:13:09.559 relates to how people understand systems. You don't actually understand 235 00:13:09.600 --> 00:13:13.240 a system because you don't actually get to direct experience 236 00:13:13.279 --> 00:13:16.360 the system, and you cannot perceive a system. What you 237 00:13:16.399 --> 00:13:18.480 perceive is a model of the system that you build 238 00:13:18.480 --> 00:13:21.120 inside of your head. And there are some common models 239 00:13:21.200 --> 00:13:23.960 that we all know and they play a very important 240 00:13:23.960 --> 00:13:26.799 part in how we pass the world around us. So 241 00:13:26.960 --> 00:13:30.840 understanding those models, or inventing new models and making them 242 00:13:31.120 --> 00:13:35.200 approachable to people is a very important part of teaching 243 00:13:35.240 --> 00:13:39.440 and also debugging and architecture and all that stuff that 244 00:13:39.480 --> 00:13:44.559 we do. So the research part, some of it is 245 00:13:45.600 --> 00:13:49.799 just going through the field and seeing what is comprehensible 246 00:13:49.840 --> 00:13:52.399 to people and how people pass the world that they 247 00:13:52.440 --> 00:13:54.720 see and how they look at the world that they see. 248 00:13:54.720 --> 00:13:58.679 Because you've viewed the system through dashboards and graphs and 249 00:13:58.720 --> 00:14:02.080 metrics and of course all the tools that you have, 250 00:14:03.519 --> 00:14:08.039 and some of it is learning psychology and learning ecology 251 00:14:08.279 --> 00:14:11.360 and you know, how people think about stuff, how the 252 00:14:11.399 --> 00:14:15.240 brain works, all the cognitive biases that go into play. 253 00:14:15.519 --> 00:14:18.000 And some of it is actually researching the tech stack 254 00:14:18.080 --> 00:14:23.799 because that also informs and kind of conditions the way 255 00:14:23.799 --> 00:14:25.399 that we experience things. 256 00:14:27.840 --> 00:14:30.440 And you're able to make a living doing that. Cool. 257 00:14:32.399 --> 00:14:37.039 Not yet, So I'm doing consulting mostly the performance stuff. 258 00:14:37.120 --> 00:14:40.200 I am giving some workshops, which is starting to get 259 00:14:40.639 --> 00:14:42.639 more and more like a bigger and bigger part of 260 00:14:43.840 --> 00:14:45.639 the work that I do. So at some point, I 261 00:14:45.679 --> 00:14:47.519 guess it will become the main thing that I do. 262 00:14:48.000 --> 00:14:52.440 Hopefully I can make a living just doing that, but 263 00:14:52.480 --> 00:14:53.039 we'll see. 264 00:14:54.279 --> 00:14:57.360 I just wanted to also add that, you know, when 265 00:14:57.399 --> 00:15:00.639 you were speaking about the importance of models, it kind 266 00:15:00.639 --> 00:15:04.240 of also directly translates to the importance of naming things 267 00:15:04.360 --> 00:15:05.759 and naming things properly. 268 00:15:08.120 --> 00:15:10.360 Yeah, because it's just it's not just the name, it's 269 00:15:10.399 --> 00:15:13.320 the name and everything, you know, all the associations that 270 00:15:13.360 --> 00:15:14.120 it tells with it. 271 00:15:15.720 --> 00:15:21.679 I totally agree. So I actually brought you here because 272 00:15:21.759 --> 00:15:24.720 of an I don't know if you know, going back 273 00:15:24.759 --> 00:15:28.399 to my kind of little speech at the beginning because 274 00:15:28.440 --> 00:15:33.639 of this argument we had online about binary protocols versus 275 00:15:33.679 --> 00:15:39.200 sexual protocols. But while preparing I also read or reread 276 00:15:39.240 --> 00:15:41.960 some of the articles that you posted on your excellent 277 00:15:42.039 --> 00:15:47.399 website website, uh and so, and there were two the 278 00:15:47.480 --> 00:15:50.559 latest two posts. They're kind of caught my eye. One 279 00:15:50.600 --> 00:15:54.960 of them is titled Don't Paint the Roses and the 280 00:15:55.000 --> 00:15:59.000 other is titled the Marshmallow Effect. And I definitely want 281 00:15:59.039 --> 00:16:01.279 to talk about these as well. So I'll leave it 282 00:16:01.360 --> 00:16:03.879 up to you which which one of those three you 283 00:16:03.919 --> 00:16:04.639 want to start with. 284 00:16:05.840 --> 00:16:08.320 I'm cool with those all of them, but let's talk 285 00:16:08.320 --> 00:16:11.360 about text and binary protocols for a second. 286 00:16:11.480 --> 00:16:12.440 Okay for sure. 287 00:16:13.000 --> 00:16:16.879 It's also like it ties very nicely into the other topics. 288 00:16:17.320 --> 00:16:22.320 So actually, one of the points that I want make 289 00:16:22.440 --> 00:16:24.519 is that there is no such thing as a text protocol. 290 00:16:25.879 --> 00:16:27.840 You kind of stole the one of what I wanted 291 00:16:27.840 --> 00:16:30.600 to say. I basically wanted to say that if you 292 00:16:30.679 --> 00:16:34.960 go down low enough in the stack, everything is binary. 293 00:16:35.000 --> 00:16:36.879 It's at the end of the day, it's bits over 294 00:16:36.919 --> 00:16:42.399 the wire, so obviously it's it's binary. And you know, 295 00:16:42.679 --> 00:16:47.279 you encrypt things, you compress things, so so obviously that 296 00:16:47.320 --> 00:16:52.000 turns them into binary data. But I was more thinking 297 00:16:52.039 --> 00:16:55.200 about the applicative layer. So I was making the argument 298 00:16:55.279 --> 00:16:59.159 that at the applicative layer, in a lot of cases, 299 00:16:59.320 --> 00:17:01.759 not always, but in a lot of cases, there are 300 00:17:01.840 --> 00:17:07.039 benefits to text based protocols, for example, protocols that send 301 00:17:07.960 --> 00:17:13.759 JSON or XML or god forbid or or HTTP over 302 00:17:13.799 --> 00:17:18.279 the wire. And from my understanding, you were making the 303 00:17:18.440 --> 00:17:23.559 argument that in most cases, a binary protocol is what 304 00:17:23.599 --> 00:17:24.119 you prefer. 305 00:17:25.279 --> 00:17:28.640 So that's one that's part of it. But I think 306 00:17:28.680 --> 00:17:32.599 that the problem with this discussion is that it's about, 307 00:17:33.200 --> 00:17:35.880 you know, the civilization and the protocol itself, and it 308 00:17:35.960 --> 00:17:40.200 ignoes the larger context of how we write software, you know, 309 00:17:40.279 --> 00:17:42.079 when we have a sort of protocol. And I'll give 310 00:17:42.119 --> 00:17:45.839 you an example of what I mean. So Jason, for example, 311 00:17:46.200 --> 00:17:52.240 doesn't actually specify what the size of an integer is, okay, 312 00:17:52.359 --> 00:17:55.680 or what the actual precision of the float is. And 313 00:17:56.279 --> 00:17:58.519 and they kind of fixed it in the following out 314 00:17:58.519 --> 00:18:01.599 of seas, but not all implements, of course, adhere to it. 315 00:18:02.880 --> 00:18:08.079 But the thing about this is that if you convert 316 00:18:08.240 --> 00:18:12.039 to a binyl protocol corpor binyl protocol, it takes much 317 00:18:12.119 --> 00:18:16.279 less and much more like bytes, much more verbals in 318 00:18:16.319 --> 00:18:19.160 how it in codes things. So a specific number, for example, 319 00:18:19.160 --> 00:18:22.640 if you basically you have one byte per digit, so 320 00:18:22.799 --> 00:18:26.559 very large number actually takes a lot of space, which 321 00:18:26.599 --> 00:18:28.960 means it has a lot of facts. But it means 322 00:18:29.039 --> 00:18:31.880 that when you're writing software you would not use the 323 00:18:31.880 --> 00:18:38.000 same precision because of this. So the protocol actually affects 324 00:18:38.039 --> 00:18:41.079 how you write the code. And we have a lot 325 00:18:41.119 --> 00:18:44.799 of code that is written under the you know, assuming 326 00:18:44.839 --> 00:18:47.480 that you're going to use certain protocol, and because of it, 327 00:18:47.480 --> 00:18:50.960 it's written differently. You're not actually using the same position, 328 00:18:51.079 --> 00:18:53.039 you're not actually using the same data types, you're not 329 00:18:53.079 --> 00:18:56.200 actually actually writing the same software that you would otherwise. 330 00:18:56.920 --> 00:19:00.119 So even if the efficiency of binary protocols would be 331 00:19:00.160 --> 00:19:03.680 the same, and it's not, you would so profell to 332 00:19:03.720 --> 00:19:06.640 have them because of because it allows you to write 333 00:19:06.640 --> 00:19:10.319 better software. And that's an angle that is not often 334 00:19:10.359 --> 00:19:14.960 discussed when you're talking about protocols. And just to clarify 335 00:19:15.000 --> 00:19:19.160 this example more, think about what happens if you want 336 00:19:19.240 --> 00:19:23.160 to do a denial of service attack against you know, 337 00:19:23.200 --> 00:19:27.440 some server that accept Jason XML, you actually have to 338 00:19:27.480 --> 00:19:31.160 lead the entire Jason before you understand that the message 339 00:19:31.279 --> 00:19:34.240 is done. So you have to have very large buffles 340 00:19:35.799 --> 00:19:37.920 just in case, you know, someone actually decides to send 341 00:19:37.960 --> 00:19:42.039 you know, like eight megabytes of Jason. And if I 342 00:19:42.160 --> 00:19:44.720 send a very large and very deep Jason to you, 343 00:19:45.319 --> 00:19:48.599 I can very easily do like denial of service attack 344 00:19:48.759 --> 00:19:51.920 against parser. Well as in YO protocol, most of them 345 00:19:52.200 --> 00:19:55.440 have very strict schemas which allow them to detect no 346 00:19:55.559 --> 00:19:59.759 phone messages very very easily and very early. They don't 347 00:19:59.759 --> 00:20:02.759 actually need to have the entire buffal. Most of them 348 00:20:02.759 --> 00:20:07.599 can also stream, so it makes the website of those 349 00:20:07.759 --> 00:20:12.559 much more efficient, much more, much cheaper, but also immune 350 00:20:12.599 --> 00:20:16.000 to the now service tax. And this guy, you can 351 00:20:16.079 --> 00:20:19.240 actually say this effect in all the text based protocols. 352 00:20:20.599 --> 00:20:25.039 So even if the protocols, you know, they have the 353 00:20:25.039 --> 00:20:28.279 same payload size and they you know, network was the 354 00:20:28.319 --> 00:20:30.839 same and everything else was the same, just because we 355 00:20:30.839 --> 00:20:33.720 can write better software, I would prefer to have banier protocols. 356 00:20:34.720 --> 00:20:37.559 So here's the thing. First of all, I can get 357 00:20:37.599 --> 00:20:42.920 into the technicalities of you know, some of the downsides 358 00:20:42.960 --> 00:20:47.160 that you mentioned, but before that, I want to explain, 359 00:20:47.319 --> 00:20:50.079 like what caused me to say that in many cases 360 00:20:50.119 --> 00:20:54.160 I prefer textual protocols. But first, before that, even I 361 00:20:54.200 --> 00:20:57.319 want to say that obviously there are scenarios where textual 362 00:20:57.519 --> 00:21:01.440 protocols are totally not appropriate yet. So for example, nobody 363 00:21:01.440 --> 00:21:05.920 would send a video stream as the textual protocol, or 364 00:21:06.359 --> 00:21:09.119 even images over the wire. It's probably a bad idea 365 00:21:09.200 --> 00:21:14.119 to send them as textual protocol. And I'll let you 366 00:21:14.920 --> 00:21:17.359 respond in a second age. I just wanted to say 367 00:21:17.759 --> 00:21:22.279 that the two reasons that I wanted to say as 368 00:21:22.400 --> 00:21:27.640 positives for textual protocol before we get into the possibilities 369 00:21:27.680 --> 00:21:30.559 of you know, some of the negatives that you mentioned 370 00:21:31.119 --> 00:21:35.160 is one that it's easily humanly readable, which is significant 371 00:21:35.200 --> 00:21:40.240 benefit for me when I have to work with protocols. 372 00:21:41.480 --> 00:21:45.319 Second is the fact that in many cases I don't 373 00:21:45.359 --> 00:21:49.039 really care. I care that something is a number rather 374 00:21:49.400 --> 00:21:53.000 than exactly how many bits it takes to represent it 375 00:21:53.079 --> 00:21:58.720 in the exact format. And finally is the fact that 376 00:21:58.799 --> 00:22:03.279 you can make text show protocols and binary protocols together, 377 00:22:03.839 --> 00:22:06.839 which can get you the benefit of both worlds. 378 00:22:06.880 --> 00:22:12.119 And now to you aging, So there's there's somewhat of 379 00:22:12.119 --> 00:22:15.359 a snuck premise in what you've said, which is that 380 00:22:15.920 --> 00:22:20.759 me or or a h ah, sorry, there's there's somewhat 381 00:22:20.759 --> 00:22:24.880 of a snuck premise in that you're saying if people 382 00:22:25.000 --> 00:22:30.400 are using text protocols, they're less efficient and they're more 383 00:22:30.519 --> 00:22:38.440 subject to these issues because because of the text protocol itself. 384 00:22:38.640 --> 00:22:42.440 When I'm going to argue that, because of my own experience, 385 00:22:42.480 --> 00:22:45.200 and I imagine you've seen this too, if you gave 386 00:22:45.279 --> 00:22:48.240 someone a binary protocol and it were to become popular, 387 00:22:49.400 --> 00:22:52.160 the binary protocol would have all of the same problems 388 00:22:52.200 --> 00:22:56.039 because the issue with the dialogue, to the denial of service, 389 00:22:56.119 --> 00:23:01.039 not bounds checking, not having a bite limits, to rate limits, 390 00:23:01.519 --> 00:23:06.559 those are engineering issues. They're not protocol issues. They're not 391 00:23:06.759 --> 00:23:11.480 the fault of the data format. Likewise, what goes over 392 00:23:11.519 --> 00:23:15.839 the network is a bit different depending on the size 393 00:23:15.839 --> 00:23:19.640 of message. But we're talking about communication, not binary files. 394 00:23:19.680 --> 00:23:23.200 We're not talking about videos. We're talking about communication. Normally, 395 00:23:23.279 --> 00:23:26.240 those messages are going to fit into a TCP packet, 396 00:23:26.599 --> 00:23:28.960 and if they're not going to fit into a TCP packet, 397 00:23:29.279 --> 00:23:34.720 taking away ten to twenty percent of the brackets embraces 398 00:23:34.960 --> 00:23:40.240 and replacing that with little, little size headers is not 399 00:23:41.119 --> 00:23:45.920 actually going to be what makes the difference in you know, 400 00:23:46.039 --> 00:23:48.519 in the long run of when we're exchanging messages back 401 00:23:48.519 --> 00:23:51.160