1 00:00:01,080 --> 00:00:04,799 How'd you like to listen to dot net rocks with no ads? Easy? 2 00:00:05,360 --> 00:00:09,480 Become a patron for just five dollars a month. You get access to a 3 00:00:09,480 --> 00:00:14,279 private RSS feed where all the shows have no ads. Twenty dollars a month. 4 00:00:14,279 --> 00:00:18,440 We'll get you that and a special dot net Rocks patron mug. Sign 5 00:00:18,519 --> 00:00:36,000 up now at Patreon dot dot NetRocks dot com. Welcome back to dot net 6 00:00:36,119 --> 00:00:39,000 rocks. This is Carl Franklin and this is Richard Campbell. We're here in 7 00:00:39,039 --> 00:00:45,359 our respective studios. I saw you on Windows Weekly yesterday. Oh yeah, 8 00:00:45,399 --> 00:00:47,960 you watched the live stream of that. I did, So tell me what 9 00:00:48,000 --> 00:00:52,399 you're doing on Windows Weekly besides you know, stirring up hat and discontent and 10 00:00:52,399 --> 00:00:56,640 whiskey. Yet I mean, in the end, I've come to appreciate that 11 00:00:56,719 --> 00:01:00,600 Window. This all happened because Mary Joe Foley left zd net. She went 12 00:01:00,640 --> 00:01:07,799 to work for Directions on Microsoft, which is an analyst group mostly ex microsofties, 13 00:01:07,959 --> 00:01:14,000 and she's really their sort of lead writer organizer, does the blogs and 14 00:01:14,000 --> 00:01:17,400 things like that. She super loves the new roles like super Happy, but 15 00:01:17,439 --> 00:01:19,959 Windows Weekly was intmpatible with that, and so Paul and Leo were doing it 16 00:01:21,000 --> 00:01:25,239 on their own for a few weeks, and Paul was so sad. You 17 00:01:25,239 --> 00:01:27,040 guys are buddies from way. We've been buzzy for a long time. And 18 00:01:27,120 --> 00:01:30,719 so I came in and guested a couple of times, and actually it was 19 00:01:30,799 --> 00:01:33,560 Leo. He said, like, what are we going to do to make 20 00:01:33,599 --> 00:01:36,359 it permanent? That's kind of how it worked with you and me and Dot 21 00:01:36,439 --> 00:01:40,680 Rocks. Isn't it that you were Richard the toy boy? And then it's 22 00:01:40,719 --> 00:01:47,680 like, you know, how about Yeah, that's very cool. I really 23 00:01:47,760 --> 00:01:49,519 enjoyed a conversation. It was great. Yeah, you know, Paul is 24 00:01:49,560 --> 00:01:53,000 a prolific writer, like a dozen articles a week. It's insane, and 25 00:01:53,079 --> 00:01:57,879 in a lot of ways, the show is basically going over what Paul's written 26 00:01:57,879 --> 00:02:00,239 that week, which is fine. You know, I'm familiar with most of 27 00:02:00,239 --> 00:02:04,200 the subjects. I have opinions, and it's easy to make fun of Paul. 28 00:02:05,680 --> 00:02:08,039 And then they added the you know, MJ used to do this closing 29 00:02:08,080 --> 00:02:12,000 bit on her malt liquor of the week, her beer of the week. 30 00:02:12,199 --> 00:02:15,120 She has, like her craft beer. So they asked if I could put 31 00:02:15,120 --> 00:02:16,360 something there, and it's like, how about brown liquor? And of course 32 00:02:16,400 --> 00:02:22,680 he's never forgotten you and I And he refers to this day as the Bourbon 33 00:02:22,759 --> 00:02:30,560 Brothers going up there after a build and they cracked that abler abunda the cask 34 00:02:30,639 --> 00:02:35,039 string. Oh my gosh, yeah, that stuff. He was gooned by 35 00:02:35,080 --> 00:02:37,840 the time we were done. Yeah, they ran a couple of clips of 36 00:02:37,919 --> 00:02:40,960 that day right where I think Paul said something along the lines of I feel 37 00:02:40,960 --> 00:02:49,120 like I've been shot in the tongue. Anyway, So every week now I 38 00:02:49,240 --> 00:02:53,000 do a different thing on whiskey, and they'd been clipping them and picking him 39 00:02:53,000 --> 00:02:59,439 into a playlist on YouTube as well, including over eight parts I described Scottish 40 00:02:59,439 --> 00:03:02,639 whiskey creation in detail, and if you can buy those eight bits together two 41 00:03:02,680 --> 00:03:06,240 and a half hours. So basically, there's a two and a half hour 42 00:03:06,319 --> 00:03:08,599 geek out on making Scottish whiskey out there if you want it. So you 43 00:03:08,759 --> 00:03:14,080 sent me a link to the live recording. Yeah, but you guys don't 44 00:03:14,199 --> 00:03:17,120 publish a lot a live recording. You publish an edited recording, right, 45 00:03:17,199 --> 00:03:21,520 Yeah, Yeah, they publish a podcast essentially. Yeah, but yeah, 46 00:03:21,599 --> 00:03:24,120 it's still it's just that they also stream it in some respects, and like 47 00:03:24,199 --> 00:03:29,159 every podcast, they're struggling financially Yeah. So they where they used to literally 48 00:03:29,159 --> 00:03:30,639 have a twenty four hour stream with inter socials and stuff. They've cut that 49 00:03:30,719 --> 00:03:35,240 back to we just stream their recordings. Well, it's really fun, and 50 00:03:35,319 --> 00:03:38,560 I encourage all the dot ne Rocks listeners to go check out Windows Weekly, 51 00:03:38,919 --> 00:03:43,400 just google it. It's at twit dot tv and you can get more Richard 52 00:03:43,439 --> 00:03:57,719 Cammell that way, let's let's get into better no framework. Awesome, All 53 00:03:57,800 --> 00:04:00,159 right, man, what do you got? So a friend of mine said, 54 00:04:00,240 --> 00:04:04,280 have you seen this? It's a YouTube channel YouTube dot com slash at 55 00:04:04,439 --> 00:04:11,719 Jim Browning. Jim Browning is a is a faux name, a nom de 56 00:04:11,800 --> 00:04:16,800 plume. This is an anonymous guy who finds scammers. Oh yeah, and 57 00:04:17,639 --> 00:04:25,800 and confronts them and in some cases gets them convicted. Yeah, so tracking 58 00:04:25,800 --> 00:04:29,519 and identifying scammers who knock on my front door, call me or shove pop 59 00:04:29,639 --> 00:04:33,399 ups onto my computer screen. Yeah. And there's a series of amazing videos 60 00:04:33,439 --> 00:04:36,839 here where he does just that. Yeah, I know. And he's been 61 00:04:38,040 --> 00:04:43,240 like literally unraveled a huge scam group in India. Like it's amazing, it's 62 00:04:43,399 --> 00:04:46,120 remarkable the work, But there is a certain amount of just the shoden froud 63 00:04:46,160 --> 00:04:51,839 of the scammer being scam Absolutely, I love it. You know, schadenfreud 64 00:04:51,959 --> 00:04:58,800 is the perfect word here. Yeah, so anyway, I really do appreciate 65 00:04:58,839 --> 00:05:00,759 that. And I have also noticed, and you may have noticed this, 66 00:05:00,839 --> 00:05:05,399 and I'm going to bring it up on security this week later today that if 67 00:05:05,399 --> 00:05:10,720 you have any kind of Facebook page, not a personal page, but a 68 00:05:10,720 --> 00:05:15,279 Facebook page, a business page. Irene is already nodding her head. She 69 00:05:15,360 --> 00:05:18,800 knows exactly what I'm talking about. There's this scam that is going on where 70 00:05:19,040 --> 00:05:25,920 people send you, you know, instant messages, Facebook messages saying they're representing 71 00:05:26,000 --> 00:05:30,000 Facebook and your page is about to be shut down for some copyright infringements or 72 00:05:30,000 --> 00:05:33,519 whatever unless you respond to this, and blah blah blah. And at first 73 00:05:33,519 --> 00:05:38,800 it really confused me because it was very you know, official looking, and 74 00:05:38,839 --> 00:05:43,480 then I noticed I was getting the same message from like Bob and you know, 75 00:05:43,720 --> 00:05:47,079 Ahmed and all these other people, and I'm like, okay, this 76 00:05:47,160 --> 00:05:51,959 isn't real. Well, and the accounts get deleted pretty quickly too, right, 77 00:05:53,079 --> 00:05:55,439 Like, yeah they do, but I mean, talk about shadenfreud. 78 00:05:55,519 --> 00:06:00,759 I just have a strange urge to reply bite me. Yeah, you know, 79 00:06:00,360 --> 00:06:04,360 don't, don't. But don't. Don't do it. Yeah, don't 80 00:06:04,399 --> 00:06:08,439 you don't roll around with pigs. Not only will you not get anywhere, 81 00:06:08,480 --> 00:06:11,800 but the pig enjoys it. Well. Dottna Rocks has gotten just a spate 82 00:06:11,920 --> 00:06:14,959 of these messages. So if you're any of the people who sent them to 83 00:06:15,000 --> 00:06:20,240 me, bite me. And Franklin Brothers Band, another page that I have 84 00:06:20,399 --> 00:06:26,920 has gotten slews of them, Like I'm talking maybe thirty forty of them. 85 00:06:27,319 --> 00:06:30,560 Bite me, bite me, bite me. Well, and you imagine, 86 00:06:30,560 --> 00:06:32,879 when you're not a technical person, just how freaky this must be. Like 87 00:06:32,920 --> 00:06:36,720 it must be effective. But it speaks to the state of spacebook, isn't 88 00:06:36,720 --> 00:06:41,800 it. It's just a cesspool. Now. I like Spacebook better than Facebook. 89 00:06:41,839 --> 00:06:45,199 Actually that's a really much better word for it. Yeah, because it's 90 00:06:45,240 --> 00:06:48,279 out and outer space sometimes you would hope. Eh. Anyway, that's what 91 00:06:48,319 --> 00:06:50,720 I got. Richard who's talking to us today, grabbed a comment off for 92 00:06:50,759 --> 00:06:55,680 show seventeen ninety seven, which we did back in June of twenty two with 93 00:06:55,759 --> 00:06:58,800 one erneus you Tube. Maybe you've heard her, never heard of her? 94 00:06:58,879 --> 00:07:01,199 Yeah, And oddly enough we were talking about gRPC, who would have funken 95 00:07:05,920 --> 00:07:10,959 anyway. This comment comes from codepouter otherwise known as Richard Rukima, who's also 96 00:07:11,040 --> 00:07:14,439 been a guest on the show back in March twenty two and was also a 97 00:07:14,480 --> 00:07:18,680 student at Codena Castle in Tuscany. Oh, I didn't know that. That's 98 00:07:18,720 --> 00:07:21,399 awesome. Yeah, he came to Tuscany. It took me a while to 99 00:07:21,439 --> 00:07:25,199 figure out who he was because I've never seen him in person. Yeah, 100 00:07:25,240 --> 00:07:27,279 yeah, he could get him, get him out of character, right, 101 00:07:27,319 --> 00:07:29,759 What a great guy though, And he's in your neck of the witz. 102 00:07:29,800 --> 00:07:33,240 I think, yes, somewhere nearby, not too close. I have deble 103 00:07:33,319 --> 00:07:39,439 check, but it might be in said Alberta. Anyway. His comment is 104 00:07:39,759 --> 00:07:43,920 in reference to gRPC, where he says you might want to review proto buffers. 105 00:07:44,319 --> 00:07:47,920 Yeah, for tro RPC and developers, and specifically the dot net library 106 00:07:47,959 --> 00:07:51,720 for proto buffers. I never understood why I should create a proto buff file 107 00:07:51,759 --> 00:07:56,800 to describe a message in all the languages from a dot apps which uses my 108 00:07:56,879 --> 00:08:01,160 ap API for my application. So it's just one of those utilities. And 109 00:08:01,199 --> 00:08:05,879 I'll include a link on the show to dealing with idiomatic dot Net with the 110 00:08:05,120 --> 00:08:09,160 protocol buffer's library. So yeah, we went over at gRPC and Blazer in 111 00:08:09,240 --> 00:08:13,360 the masterclass thing, right, and so uh yeah, protobuff is just part 112 00:08:13,360 --> 00:08:16,920 of the thing. That's the protocol that gRPC uses. Yes, it's the 113 00:08:16,920 --> 00:08:20,040 way to go. So, Richard, thanks so much for your comment. 114 00:08:20,079 --> 00:08:22,680 I'm pretty sure you get a copy of music Code by but ping me anyway, 115 00:08:24,240 --> 00:08:26,040 and we'll hook you up. And if you'd like a copy of music 116 00:08:26,040 --> 00:08:30,120 Code buy, write a comment on the website at dot NetRocks dot com or 117 00:08:30,199 --> 00:08:31,360 on the facebooks you publish every show there, and if you comment there and 118 00:08:31,360 --> 00:08:33,480 everybody on the show, we'll send you a copy of music Go buy. 119 00:08:33,679 --> 00:08:37,519 And you could also send us a tweet or an x or whatever the hell 120 00:08:37,600 --> 00:08:41,080 they call him these days. I'm at Carl Franklin, He's at Rich Campbell, 121 00:08:41,559 --> 00:08:43,720 or you could be like all the other cool kids and go to mastonon 122 00:08:45,039 --> 00:08:48,639 I'm at Carl Franklin at tech hub dot social, and I'm Rich Campbell at 123 00:08:48,679 --> 00:08:52,279 Mastodon dot social. Send us a toot or tweet or an ex or whatever 124 00:08:52,320 --> 00:08:56,759 the bleep you call those things now. Yeah, it'll always be Twitter to 125 00:08:56,799 --> 00:08:58,279 me. I'm sorry, Elin, Yeah, I'm okay with that. You 126 00:08:58,519 --> 00:09:03,120 You're evil. Plan is not worth really that story? Yeah, no, 127 00:09:03,159 --> 00:09:07,399 okay, I'm not sorry. I'll say it like my wife, my wife 128 00:09:07,519 --> 00:09:11,559 Kelly, she has a habit of apologizing for no parent reason, you know, 129 00:09:11,120 --> 00:09:16,080 anything. And so she's realized this and now she says sorry, not 130 00:09:16,240 --> 00:09:24,440 sorry, fuck you constantly. It's the cutest thing. That's pretty good. 131 00:09:26,159 --> 00:09:30,320 She can't stop the reflex, but she can follow it up. She's right, 132 00:09:30,480 --> 00:09:33,240 exactly. It's like a napid thought. And now it's just becoming one 133 00:09:33,240 --> 00:09:37,759 of those family things. All right, let's bring on Arena back for more. 134 00:09:37,960 --> 00:09:43,240 Arena Skirt two is a Microsoft MVP for Developer Technologies, the Software architect, 135 00:09:43,240 --> 00:09:48,720 an independent consultant, and Microsoft certified trainer, always in a quest for 136 00:09:48,720 --> 00:09:52,879 the latest trends and best practices in architecture, dot net and the world around 137 00:09:52,919 --> 00:09:56,960 it. Also having more than two thousand hours of delivered trainings, classings, 138 00:09:58,000 --> 00:10:01,519 workshops and presentations, she loves to help others grow. Twice a year for 139 00:10:01,600 --> 00:10:05,320 five months, she teaches dot net and c sharp to people that either want 140 00:10:05,360 --> 00:10:11,279 to switch careers and get into software development or simply learn new things very very 141 00:10:11,279 --> 00:10:16,320 cool. She tries to lead by example and believes in the importance of growing 142 00:10:16,360 --> 00:10:20,720 tech communities. She is the co founder of dot net Yash user group that's 143 00:10:20,960 --> 00:10:30,519 I a SI Yash of course, and dot Netdays dot ro conference where she 144 00:10:30,600 --> 00:10:33,759 tries to gather people that are willing to share their knowledge with others or gather 145 00:10:33,879 --> 00:10:37,360 around in the community, and from time to time she publishes articles on her 146 00:10:37,399 --> 00:10:46,120 blog Irena dot codes awesome domain there codes. I got to look into that, 147 00:10:46,559 --> 00:10:50,759 Carl Frank, Oh, yeah, welcome back. Thank you for having 148 00:10:50,759 --> 00:10:54,080 me. You're welcome. Great heard of this dot Net Days dot ro. 149 00:10:56,600 --> 00:11:01,480 I wonder why, all right, tell everybody come on, I'm coming to 150 00:11:01,559 --> 00:11:05,039 Yash. I'm gonna come. Yeah, I'm going to Yash. I've never 151 00:11:05,080 --> 00:11:07,639 been. I've been, you know, I've been to Cluge, I've been 152 00:11:07,679 --> 00:11:11,600 around Romania, but I've never been. I've done any Yeah, I haven't 153 00:11:11,600 --> 00:11:13,399 done gone to Yash. I'm like, I got to go to Yash. 154 00:11:13,879 --> 00:11:16,639 I don't know that I'll go to Moldova while I'm there, because it's very 155 00:11:16,639 --> 00:11:20,000 close to Moldova. But it's kind of a sketchy time to go to Moldova. 156 00:11:20,120 --> 00:11:24,240 Right now, I have a yash on my wrist and you put a 157 00:11:24,240 --> 00:11:28,159 little ointment on it and it's going away. So we're talking gRPC here. 158 00:11:28,320 --> 00:11:35,559 It's been a spell more than a year last time we checked in, and 159 00:11:35,559 --> 00:11:37,399 and you've been writing new course words, stuff like there's more to no, 160 00:11:37,799 --> 00:11:43,120 right, yeah, there's more to know and the technology is well well established. 161 00:11:43,360 --> 00:11:48,600 No, no, so it's in the last year. There's been some 162 00:11:48,639 --> 00:11:54,879 innovations in gRPC, more stabilizations, if that's a word, now I get 163 00:11:54,919 --> 00:11:58,200 that. Yeah, so gRPC is here to stay if you ask me. 164 00:11:58,639 --> 00:12:03,600 And we're really talking gRPC web, right, because that's what most developers use. 165 00:12:03,039 --> 00:12:07,080 Is that true? No, No, we're talking gRPC like back end 166 00:12:07,159 --> 00:12:13,440 technology, okay, for server for the server bit server server, yeah, 167 00:12:13,799 --> 00:12:20,000 micro services, yeah, okay, so gRPC web obviously for browsers. So 168 00:12:20,240 --> 00:12:26,759 it started at Google. But they claim that gRPC is a recursive acronym called 169 00:12:26,600 --> 00:12:31,480 gRPC Remote Procedure Calls and so if you go there, gRPC io, that's 170 00:12:31,480 --> 00:12:35,080 what they say. But they built it for, as you say, micro 171 00:12:35,159 --> 00:12:41,480 services, you know, service to service kind of interaction which doesn't require HTTP. 172 00:12:43,000 --> 00:12:48,879 Where they built it to fool people that a distributed system is the monolith. 173 00:12:48,519 --> 00:12:54,879 Like the way they are behaving. The g doesn't stand for Google. 174 00:12:56,240 --> 00:13:00,639 It used to, they claim it doesn't now because they opened it up. 175 00:13:00,639 --> 00:13:03,360 I think you know, it's just kind of a wink wink, you know. 176 00:13:03,440 --> 00:13:09,600 Yeah, it's just active re you know, re acronym, yeah, 177 00:13:09,639 --> 00:13:13,080 reassigning. Yeah, that's pretty funny. So what's new? So the tooling 178 00:13:13,240 --> 00:13:20,360 is new and stable, which makes me, uh, very glad that well 179 00:13:20,639 --> 00:13:28,000 Visual Studio caught up with everything that happens in RPC, so now you don't 180 00:13:28,039 --> 00:13:33,919 have like problems with types not being recognized and stuff like that. Okay, 181 00:13:33,960 --> 00:13:39,480 awesome. Uh. And also you have like an option in Visual Studio right 182 00:13:39,519 --> 00:13:45,440 click at reference and you'll search and find the proto file and it will automatically 183 00:13:45,440 --> 00:13:50,799 install some packages for you. And that's awesome because we want straightforward things. 184 00:13:50,840 --> 00:13:54,879 We do not want to remember what packages we need to install. And there's 185 00:13:54,919 --> 00:14:00,960 a bunch of them in for the different places too. Yeah, you talked 186 00:14:00,960 --> 00:14:05,759 about types when you when you generate a client or a service, you have 187 00:14:05,799 --> 00:14:11,360 to define these types in a in a protofile, as Richard Rukumo was saying. 188 00:14:11,919 --> 00:14:16,200 And those types don't you know, if the the actual types like string 189 00:14:16,360 --> 00:14:20,440 in whatever. Some of them map, but a lot of them don't. 190 00:14:20,679 --> 00:14:26,639 And I always find it difficult if I'm doing anything complex to well, nowadays 191 00:14:26,639 --> 00:14:31,320 I ask chat GPT, but usually I would go browsing for it. Uh. 192 00:14:31,519 --> 00:14:35,399 You know what is the proto type that matches, you know, a 193 00:14:35,559 --> 00:14:39,080 double array or something like that. Well, basically they they try to help 194 00:14:39,159 --> 00:14:46,360 us with something that is called Google well Known types, which is the common 195 00:14:46,399 --> 00:14:52,039 ground between all languages out there. But if you need something specific only for 196 00:14:52,120 --> 00:14:56,159 a specific language, you'll have to mimic and code it yourself. In broad 197 00:14:56,200 --> 00:15:00,840 above. Yeah, you create your own message type that has whatever. Yeah, 198 00:15:00,960 --> 00:15:05,080 I mean the doc liarly says protocol buffers well known types. Yeah. 199 00:15:05,840 --> 00:15:07,919 Yeah, I presume as soon as you go off this list, now you've 200 00:15:07,919 --> 00:15:11,159 got to figure it out. Yeah, but this is what proto buff's all 201 00:15:11,200 --> 00:15:16,159 about, right, Yeah, it's not a language, it's a code generation 202 00:15:16,519 --> 00:15:22,840 template really yep. So basically the fact that we're writing and defining the services 203 00:15:22,840 --> 00:15:28,000 and the types in proto buffs and it makes us understand the new syntax and 204 00:15:28,039 --> 00:15:31,960 new types and all. It's overhead for us. But you can also have 205 00:15:33,039 --> 00:15:37,639 the option of starting with your own existing types and feels like code first gRPC 206 00:15:39,360 --> 00:15:45,960 due to a nice library written by Mark Gravelle, which allows you to reuse 207 00:15:46,120 --> 00:15:50,600 if you have WCS for example and you have types that are adnotating with data 208 00:15:50,600 --> 00:15:54,919 contracts and stuff like that, you can get rid of proto buffs. That 209 00:15:56,039 --> 00:15:58,840 was one of the first things that I wrote as well, and I showed 210 00:15:58,879 --> 00:16:02,279 it on Blazer Train. Back when I first learned gRPC. It was like, 211 00:16:02,320 --> 00:16:03,480 well, what if I've already got, like, say, an API 212 00:16:03,600 --> 00:16:08,000 layer, I've already got my types and I have to redo them and now 213 00:16:08,000 --> 00:16:12,360 I've got duplicate code and all that stuff. So, yeah, it's a 214 00:16:12,399 --> 00:16:21,000 smart idea, just you make a code generator generator kind of. I'm sorry 215 00:16:21,039 --> 00:16:23,360 when you say it that way. It doesn't feel like a smart idea anymore. 216 00:16:25,600 --> 00:16:29,600 I know, it's weird. It's a lot of weirdness. Yeah, 217 00:16:29,639 --> 00:16:33,320 but there's a problem. There's problems there because you have to you know, 218 00:16:33,519 --> 00:16:38,000 every time that you get a call or a bunch of variables, you have 219 00:16:38,080 --> 00:16:42,399 to convert them in the service, right, so there has to be stuff 220 00:16:42,559 --> 00:16:47,960 in there that runs at run time, and you know, you have to 221 00:16:48,120 --> 00:16:52,639 take into account all the types. It's a lot. Yeah, but basically, 222 00:16:52,759 --> 00:17:00,960 let's be honest, we all kind of use basic types with our other 223 00:17:00,039 --> 00:17:08,680 types custom ones. So as long as those are like interoperable where they can 224 00:17:08,960 --> 00:17:15,359 work with NOGS or something else. Sure, Yeah, back to why do 225 00:17:15,440 --> 00:17:18,319 we build types? Why do we do this? How does this really save 226 00:17:18,400 --> 00:17:22,279 us pain? It is just because it's easier to pass in APIs. Basically, 227 00:17:22,359 --> 00:17:26,680 once you start with gRPC, you have one source of truth, or 228 00:17:26,000 --> 00:17:30,279 you should have, like this protocol buffers file. And the point of gRPC, 229 00:17:32,000 --> 00:17:36,839 beyond the fact that it uses a binary serialization, is that now you 230 00:17:36,880 --> 00:17:44,759 can have a c sharp consumer that talks with an OGS server and the only 231 00:17:44,799 --> 00:17:48,440 thing that these two have in common is that protocol buffers file. That You're 232 00:17:48,519 --> 00:17:51,839 right, right, And that's kind of the magical part about it is. 233 00:17:52,039 --> 00:17:56,160 Yeah, this is cross platform without WS starring a whole lot of XML. 234 00:17:56,319 --> 00:18:00,119 Oh yeah, and performance is amazing. I mean I did some comparisons using 235 00:18:00,160 --> 00:18:07,200 gRPC web in Blazer to download five thousand records from an API and then from 236 00:18:07,319 --> 00:18:12,880 gRPC, and it was like the second and a half compared to like two 237 00:18:12,920 --> 00:18:18,640 hundred and fifty milliseconds. It's that different. Yeah, and that's on the 238 00:18:18,680 --> 00:18:22,079 same sheme. But still, I mean, that kind of ratio is what 239 00:18:22,119 --> 00:18:26,160 you can expect. And yet how is the adoption arena, Like I still 240 00:18:26,160 --> 00:18:32,000 don't see the huge amount of tractions For gRPC, there isn't a huge amount 241 00:18:32,039 --> 00:18:37,519 of traction. What I'm seeing in my workshops and at the conferences, I'm 242 00:18:37,559 --> 00:18:41,880 hearing people talking about transitioning from WCF to gRPC. Right, this is the 243 00:18:41,920 --> 00:18:48,119 most common scenario, and simply because they want to get to Core. Yeah, 244 00:18:48,160 --> 00:18:52,799 and WCF is not going indeed, and nobody's waiting for the open source 245 00:18:52,839 --> 00:18:59,480 version WCF exactly, and why would you Anyway, there's a whole lot of 246 00:18:59,519 --> 00:19:03,759 pain. I don't know if you guys remember, but uh yeah, configuration 247 00:19:03,960 --> 00:19:08,599 nightmare, that's what I remember. That's why you guys, guys like the 248 00:19:08,680 --> 00:19:15,119 Germans, you know, Angle Grammar and all those guys who were into WCF, 249 00:19:15,480 --> 00:19:18,799 they made the fortune on helping people configure their WCF services well. And 250 00:19:19,079 --> 00:19:23,240 it also appeals to the mind with that kind of precision, like as long 251 00:19:23,279 --> 00:19:26,839 as you do it right, it works great. The moment you do it 252 00:19:26,880 --> 00:19:32,559 wrong, it's essentially undiagnosable. Yeah, I get that. Where gRPC seems 253 00:19:32,599 --> 00:19:34,519 way more stripped down. It seems like the best part of dot net remoting, 254 00:19:34,799 --> 00:19:40,559 which I loved, like dot remoting between two dot net services back in 255 00:19:40,640 --> 00:19:47,400 the day just worked like it was nothing and it was fast form work with 256 00:19:47,480 --> 00:19:49,400 anybody else. No, So then they say, hey, you know what 257 00:19:49,400 --> 00:19:55,680 you like more soap? They were wrong, they were wrong. That was 258 00:19:55,720 --> 00:20:02,680 a lie. Soap did not clean anything what they were thinking. Yeah, 259 00:20:02,880 --> 00:20:06,920 it wasn't simple, it didn't have any objects, it didn't provide access, 260 00:20:07,079 --> 00:20:18,079 it wasn't a protocol. The name was perfect. Oh man, I could 261 00:20:18,079 --> 00:20:23,440 do this all day. So I know that you're, you know, peer 262 00:20:23,480 --> 00:20:27,599 to peer in the micro services world, so you're not really using the HDP 263 00:20:27,799 --> 00:20:36,039 web version of it. But I was following the the way that what is 264 00:20:36,160 --> 00:20:42,240 to your what's the deal with HDDB two and why didn't something work? And 265 00:20:42,319 --> 00:20:47,599 does it work? Now? I'm sorry for being vague, but I just 266 00:20:47,680 --> 00:20:53,240 remember that there was a problem with maybe Azure not supporting HDP two or and 267 00:20:53,240 --> 00:20:57,920 that's why gRPC web came around. I don't know, something like that. 268 00:20:59,359 --> 00:21:06,079 I mean, they added grbc into like dotnut world in twenty nineteen, but 269 00:21:06,440 --> 00:21:11,240 Asia wasn't supporting that until I don't know last year or something. So it 270 00:21:11,279 --> 00:21:15,559 has been Oh yeah, it has been a gap around it. Yeah, 271 00:21:15,559 --> 00:21:18,559 that's what I remember. And it's not like HGB two is new. We 272 00:21:18,559 --> 00:21:22,680 were using it it's strangely, for crying out loud, now they're talking HGDP 273 00:21:22,799 --> 00:21:27,720 three. Yeah, so it's not even new anymore. HGB two is implemented 274 00:21:27,759 --> 00:21:33,119 in twenty eighteen in the Azure App Service. Yeah, so it's been a 275 00:21:33,119 --> 00:21:37,079 few years. But still they didn't get to it soon. No, I 276 00:21:37,079 --> 00:21:42,480 don't know why. Yeah, it's we don't We don't know the cloud problems, 277 00:21:44,039 --> 00:21:45,079 and we don't want to know the cloud problems. Oh yeah, cloud, 278 00:21:45,440 --> 00:21:49,039 those are hard problems. I don't envy those guys and our browsers using 279 00:21:49,200 --> 00:21:55,279 HDTP two now, yes, they are since when for a long time, 280 00:21:55,440 --> 00:22:00,279 for a long time, because it was easy, relatively speaking. Work on 281 00:22:00,319 --> 00:22:03,400 the server was the hard part, right, and it was. Yeah, 282 00:22:03,440 --> 00:22:07,960 twenty fifteen was when it was ratified for twenty eight for browser side. Okay, 283 00:22:07,039 --> 00:22:10,599 So the fact that it took three years for Azure to get it going 284 00:22:10,680 --> 00:22:11,880 just on app service, like they still got to do a fund more and 285 00:22:11,920 --> 00:22:15,160 all those other pieces like that makes sense to me. But you know, 286 00:22:15,200 --> 00:22:18,720 we were battling that ressulcilationship with twenty thirteen, and one of the one of 287 00:22:18,720 --> 00:22:22,200 the reasons you used, strangely it was that it did terminate HTP two for 288 00:22:22,279 --> 00:22:26,960 you. So these are little bits of information that have fallen out of my 289 00:22:26,039 --> 00:22:30,880 ears, and you know, you don't want to know this stuff like that's 290 00:22:30,000 --> 00:22:34,599 just it. It's it's not a story worthy, you know. Yeah, 291 00:22:36,359 --> 00:22:40,359 this is the plumbing that makes the Internet go around. Like once upon a 292 00:22:40,440 --> 00:22:44,920 time the web was just not that robust and having two connections to a web 293 00:22:44,960 --> 00:22:52,680 server was a lot. So has the performance or scalability of gRPC improved over 294 00:22:52,720 --> 00:22:57,240 the last year. I don't know about it being improved, but I just 295 00:22:57,319 --> 00:23:03,039 recently wrote a blog about hosting in Asia and it was so simple. I 296 00:23:03,160 --> 00:23:07,640 was like, okay, no, no wonder that it took like three years 297 00:23:07,640 --> 00:23:15,039 for you to get app service and gRPC workable with each other. So it's 298 00:23:15,160 --> 00:23:19,880 very nice right now. Oh good, So again you're talking too ling as 299 00:23:19,920 --> 00:23:23,400 your support. All of those things have. It's come into its own. 300 00:23:23,799 --> 00:23:30,200 I don't know why gRPC isn't more popular because it's certainly it's certainly so usable 301 00:23:30,400 --> 00:23:34,599 in just about every scenario where you have communication. I think it's hard. 302 00:23:36,279 --> 00:23:38,519 I mean, once you have if you have a green field app where you 303 00:23:38,599 --> 00:23:42,880 have services that need to talk with other downstream services and you need some kind 304 00:23:42,920 --> 00:23:48,000 of I don't know request response scenario, then it's fine. But if you 305 00:23:48,079 --> 00:23:53,400 have a huge application that it's already well established, it's hard in terms of 306 00:23:53,519 --> 00:23:57,200 time to retrofit, so it will be a while. Yeah. No, 307 00:23:57,319 --> 00:24:00,400 no. And if you're used to living in a land and then you just 308 00:24:00,559 --> 00:24:06,200 poorted that up to app service, like you haven't really embraced a different view 309 00:24:06,200 --> 00:24:08,400 of the web. You still kind of an iis person, yeap, And 310 00:24:08,440 --> 00:24:15,759 you havn't grasped these ideas of reverse proxying and all of the other mindsets that 311 00:24:15,160 --> 00:24:19,440 other web approaches, you know, not coming from dot Netland, not coming 312 00:24:19,440 --> 00:24:23,440 from Microsoft Land, embraced years ago. So you know, I still feel 313 00:24:23,440 --> 00:24:26,440 like we're playing catch up in a lot of respect. You know, the 314 00:24:26,559 --> 00:24:30,519 day I shut off Internet Information Server was the happiest day of my life. 315 00:24:30,920 --> 00:24:36,599 Dude, dude. You know, I sent both of my old servers to 316 00:24:36,799 --> 00:24:40,279 the e cyclers to be torn, you know, taken apart. And that 317 00:24:40,759 --> 00:24:44,880 was a good day. That was a good I am now purging all racks 318 00:24:44,880 --> 00:24:48,799 out of my life. I want to see it getting shredded. I'm not 319 00:24:48,000 --> 00:24:52,799 leaving until I see it shredded. Well, I shredded the drives and then 320 00:24:52,880 --> 00:24:55,759 shut the chassis off. But the motherboards were for twenty thirteen. I didn't 321 00:24:55,759 --> 00:24:59,160 owe anybody anything. I was on my third set of fans in those things. 322 00:24:59,200 --> 00:25:03,799 But now I'm with you. You know, it's a cloud world like 323 00:25:03,160 --> 00:25:08,359 less less is more at home absolutely, Plus you know, never mind that 324 00:25:08,480 --> 00:25:15,799 the security implications of Internet information. Everything's on by default and yeah, yeah, 325 00:25:15,039 --> 00:25:19,039 it's the it's the Swiss army knife with all the blades out right. 326 00:25:21,279 --> 00:25:25,960 So what have you been doing in your day to day life with gRPC? 327 00:25:26,160 --> 00:25:33,839 That's interesting arena basically helping teams that want to transition from w CF to gRPC. 328 00:25:34,039 --> 00:25:40,319 That's the most common center that I've found. I currently try to keep 329 00:25:40,359 --> 00:25:45,359 an eye on it just because they have some bits that really interest me. 330 00:25:45,799 --> 00:25:52,000 And when those bits will be established and mature enough, I think it will 331 00:25:52,519 --> 00:25:56,720 generate a lot of traction in the industry around gRPC. And I'm talking about 332 00:25:56,759 --> 00:26:03,000 gRPC transcoding, especially Oh you said the T word, Oh yeah, all 333 00:26:03,119 --> 00:26:10,119 right, tell us transcoding. So transcoding is a package that you can install 334 00:26:10,680 --> 00:26:15,960 and allows you to also call your gRPC service from the browser as you would 335 00:26:17,000 --> 00:26:22,880 do with any other HTTP API. But and it does the translation from Jason 336 00:26:23,119 --> 00:26:30,599 that you send in the browser to your gRPC service. Now is this different 337 00:26:30,599 --> 00:26:37,200 from gRPC web kind of? I mean the thing, yeah, accomplished is 338 00:26:37,240 --> 00:26:41,559 the same thing, indeed, But it allows you from the browser to also 339 00:26:42,000 --> 00:26:49,599 specify endpoints and to define like routes can bind variables from the routes. You 340 00:26:49,640 --> 00:26:56,000 can have body that are bound to different models that you have as a parameter 341 00:26:56,119 --> 00:27:02,079 in your gRPC service. So I don't power it's yep, indeed, But 342 00:27:02,240 --> 00:27:04,759 now there's a lot of manual work that you need to do in order for 343 00:27:04,799 --> 00:27:11,559 that to work in a basic scenario. And what it happens currently is that 344 00:27:11,640 --> 00:27:18,480 if you have like an integration project or a unit testing project for your gRPC 345 00:27:18,640 --> 00:27:23,039 service, and you also want to add transcoding, bad luck. So nothing 346 00:27:23,079 --> 00:27:30,119 will work anymore, no gRPC for you. Yeah, so take your mind 347 00:27:30,119 --> 00:27:34,039 off of it, don't go that way. Yeah, things will be hard, 348 00:27:34,480 --> 00:27:38,160 and we should take a brief break. Yeah, let's do that. 349 00:27:38,279 --> 00:27:41,279 We're gonna pause here for these very important messages and won't be right back, 350 00:27:45,119 --> 00:27:48,240 and we're back. It's dot net rock some Carl Franklin, that's Richard Campbell, 351 00:27:48,400 --> 00:27:52,240 Yes, sir, and we're talking to Arena skirt you about gRPC. 352 00:27:52,680 --> 00:27:59,440 What's new in the last year so far, it's tooling and Azure support and 353 00:27:59,720 --> 00:28:03,119 training. I don't know if trans coding is new. It's not necessarily new. 354 00:28:03,519 --> 00:28:08,039 Now it's like way better, way better, right, improved, yeah, 355 00:28:08,079 --> 00:28:15,519 improved yes? And uh and the WCF problem that's uh, well, 356 00:28:15,519 --> 00:28:18,599 this really comes to that core issue, like you're doing all these migrations. 357 00:28:18,160 --> 00:28:23,960 What are the typical problems that a dot net dev coming from WCF have trying 358 00:28:23,960 --> 00:28:27,559 to understand g RPC? Like where what do you find? You're explaining over 359 00:28:27,599 --> 00:28:33,319 and over again. So I think the general the I think the concept of 360 00:28:33,440 --> 00:28:40,599 gRPC itself, the fact that it resides and lives under the HTP two protocol 361 00:28:41,039 --> 00:28:47,920 and you have these small requests that travel on one single connection. That's the 362 00:28:47,920 --> 00:28:52,920 hard beat that developers don't understand at first. And then the fact that you 363 00:28:53,000 --> 00:29:02,039 have like different set of heathers poor gRPC requests. You have that concept that 364 00:29:02,200 --> 00:29:10,279 is named trailers that comes sent as a response from the server at the end. 365 00:29:11,160 --> 00:29:15,160 So there are a lot of new concepts. Yeah. I mean you 366 00:29:15,319 --> 00:29:19,039 literally call it headers and then call it trailers. Yeah, and you work 367 00:29:19,119 --> 00:29:25,839 with one type for both of them. So a dictionary, right, same? 368 00:29:26,160 --> 00:29:29,720 Yeah, So I mean and the point being there called trailers because they 369 00:29:29,720 --> 00:29:33,200 come after the data. It's the head of the data. But the type 370 00:29:33,240 --> 00:29:37,279 that we you were you use in order to actually add key value pairs to 371 00:29:37,319 --> 00:29:45,519 those trailers. It's a metadata type that you can also use to add headers 372 00:29:45,000 --> 00:29:52,359 that goes before in your RPC requests. So yeah, it's just a naming 373 00:29:52,440 --> 00:29:56,880 problem like the seam streets forward. And yet the words are complicated. It's 374 00:29:56,920 --> 00:30:02,480 not the first industry to you confuse headers and trailers. Think of movies, 375 00:30:03,559 --> 00:30:07,200 right, right, you see a movie trailer is your introduction to the movie? 376 00:30:07,400 --> 00:30:10,680 Why is it called a trailer? It should be called a header or 377 00:30:10,720 --> 00:30:15,039 an announcement. But it turns out there's a reason for that because they used 378 00:30:15,039 --> 00:30:19,319 to show these double features and days in the theater, and the trailers for 379 00:30:19,400 --> 00:30:23,400 the next movie or whatever would come after the first one, right, And 380 00:30:23,440 --> 00:30:26,400 they just never changed the name. They just never changed it because it's a 381 00:30:26,440 --> 00:30:32,839 word and everybody knows what it is. Right. So but the point I 382 00:30:32,839 --> 00:30:36,759 think this is your point arena is like if you work in gRPC metadata, 383 00:30:36,839 --> 00:30:40,720 it's going to generate the headers and the trailers as needed. No, it's 384 00:30:40,759 --> 00:30:45,960 basically you add some extra concepts to the concept that you already know. Like 385 00:30:45,079 --> 00:30:51,880 the h protocol protocol it has headers, right, so you can add key 386 00:30:51,920 --> 00:30:57,279 value pairs in there. But underneath these you also add key value pairs that 387 00:30:57,359 --> 00:31:03,160 are specific to a gRPC request. And depending on the part where you're adding 388 00:31:03,200 --> 00:31:08,039 those with the gRPC request or with the gRPC response, they're called differently, 389 00:31:08,720 --> 00:31:17,839 right, So so the header of a response is called the trailer. Yeah, 390 00:31:18,119 --> 00:31:29,759 no problem, it's fine. I take off my parachute and okay, 391 00:31:30,279 --> 00:31:33,400 I get it, and it is all metadata. So it is good to 392 00:31:33,400 --> 00:31:36,720 think in the context of metadata. Let me tell you about the package I 393 00:31:36,759 --> 00:31:38,880 want to send, and yeah, some stuff's going to go up front, 394 00:31:38,920 --> 00:31:42,079 some stuffs going to go back. It doesn't They don't get too sweaty about 395 00:31:42,079 --> 00:31:48,200 that. And even though it's weird due to the fact that this is implemented, 396 00:31:48,920 --> 00:31:52,799 it's very helpful because you might have things that don't fit in the message 397 00:31:52,799 --> 00:31:56,759 body basically and the request itself, so you have to have a place where 398 00:31:56,799 --> 00:32:01,680 to shove those so well, and I appreciate that you don't have a separate 399 00:32:01,680 --> 00:32:05,559 set of calls for the header and the trailer and so forth. You just 400 00:32:05,640 --> 00:32:07,279 speak to metadata and you can do all the things. So in a lot 401 00:32:07,319 --> 00:32:13,480 of ways it's simple. It's just describing it well. It feels challenging, 402 00:32:13,759 --> 00:32:17,680 right, But there's also a nuance here if we're talking about different gRPC modes, 403 00:32:17,799 --> 00:32:23,559 like when we have the server streaming the server streaming modes or types what 404 00:32:23,720 --> 00:32:31,759 sends several GPS responses back to the client, But only after all those messages 405 00:32:31,799 --> 00:32:37,759 are finished are sent, you'll get those trailers, right, So this is 406 00:32:37,880 --> 00:32:42,640 the way you'll you'll know that, Okay, I'm I'm done, I received 407 00:32:42,680 --> 00:32:45,839 all the responses. Here's the trailers. Yeah. Again, it's like the 408 00:32:46,000 --> 00:32:50,240 issue here with gRPC is that it can be a lot of different things. 409 00:32:50,279 --> 00:32:54,400 It can be you know, a synchronous streamer or just a messaging and so 410 00:32:54,519 --> 00:32:58,599 the rules are different. So yeah, you gotta be careful about using trailers 411 00:32:58,599 --> 00:33:00,640 and streaming mode because you don't know when you're going to get them. It 412 00:33:00,839 --> 00:33:04,400 could be a while. Yeah. Yeah, I like to use gRPC in 413 00:33:04,440 --> 00:33:10,359 place of you know, rest API calls to controllers in a regular application, 414 00:33:10,480 --> 00:33:15,079 you know, in a Blazer app for example. That's I really love it 415 00:33:15,200 --> 00:33:16,839 and of course, it wasn't meant to do that, but it works. 416 00:33:17,079 --> 00:33:21,400 It works great fast, and yeah, I mean it's interesting you say that 417 00:33:21,440 --> 00:33:23,480 called that it wasn't meant to do that. Yeah, it wasn't built for 418 00:33:23,519 --> 00:33:27,960 that. What is gRPC meant to do? Like, it's not just about 419 00:33:27,960 --> 00:33:30,039 streaming either. Well, I mean, and we'll take that. I mean 420 00:33:30,039 --> 00:33:35,920 it was built for micro service communication, service to service. Yeah. I 421 00:33:35,960 --> 00:33:40,160 mean we can start like with a short history. How many of us heard 422 00:33:40,240 --> 00:33:45,880 the term micro services, like I'm developing micro services, the team develops micro 423 00:33:45,960 --> 00:33:52,759 services, when in fact those are not are just services that respond with Jason. 424 00:33:52,319 --> 00:33:57,920 Yeah, and they passed that data to each other with Jason. Yeah. 425 00:33:58,000 --> 00:34:02,960 So people defaulted to something that is so called REST, like to describe 426 00:34:04,119 --> 00:34:08,039 an API that responds with Jason as a data foremant, right, which made 427 00:34:08,039 --> 00:34:12,840 all the rest people angry. What I remember about REST is mostly hearing I 428 00:34:12,880 --> 00:34:15,400 was doing it wrong. I have so many beefs, and when we were 429 00:34:15,440 --> 00:34:20,119 talking to Grant Barrett about this, I have so many beefs with people who 430 00:34:20,199 --> 00:34:22,000 use the wrong word for things, and I just got to learn to let 431 00:34:22,000 --> 00:34:24,280 it go and let them play on my lawn. It's going to be okay. 432 00:34:24,320 --> 00:34:29,039 You know it's going to be okay, but I still have some Here's 433 00:34:29,079 --> 00:34:35,719 one when people describe the Browning Milliard reaction that they get on meat, they 434 00:34:35,719 --> 00:34:39,239 say it's caramelizing. No it's not. It's not caramelizing. There's no sugar 435 00:34:39,239 --> 00:34:43,719 in me. Little things like that. But people get a word for something, 436 00:34:44,519 --> 00:34:46,360 and you know, it's popular, and so then they overuse it and 437 00:34:46,360 --> 00:34:51,199 they put it on everything, and it's rest and the RESTful, which is 438 00:34:51,239 --> 00:35:00,159 a nuance. I'm happier with rest dish. I'm not going for perfection your 439 00:35:00,199 --> 00:35:05,199 voice. I'm just trying to get my wrap to work. Rested, well 440 00:35:05,280 --> 00:35:08,960 rested, it's well rested. Best to keep it simple. Yeah, we 441 00:35:08,960 --> 00:35:13,920 should just start making up things and asking people on the show. So how 442 00:35:13,960 --> 00:35:25,079 long do you let that proof? Well? I use the oiling library oiler 443 00:35:25,519 --> 00:35:34,360 and then I wrap it, oh man, and then it's only proofed when 444 00:35:34,440 --> 00:35:39,159 doubled, and then you go, you bacon, let it icd to bake 445 00:35:39,239 --> 00:35:44,760 a bit ye down the pipeline to the bank. Yeah, we're making work 446 00:35:45,159 --> 00:35:49,599 way way too an extreme that I really enjoying. This all started with the 447 00:35:49,639 --> 00:35:52,800 metadata conversation in this naming problem, and now we've promptly painted a worse one. 448 00:35:53,159 --> 00:35:57,400 You know, I'm presuming that the folks who wrote this, we're trying 449 00:35:57,440 --> 00:36:00,480 to do a good job. And it's still you know, there's some challenges 450 00:36:00,519 --> 00:36:07,000 here. Yeah, it's a technology that needs to mature, and it will 451 00:36:07,079 --> 00:36:12,440 mature enough when people will use it, because otherwise it's just a tool out 452 00:36:12,480 --> 00:36:15,679 there that nobody uses and no one will invest in developing it. Yeah, 453 00:36:16,079 --> 00:36:20,760 and can you make it simpler? Like it does feel like gRPC is mature 454 00:36:20,840 --> 00:36:23,719 enough that we are looking at some simplicity down to the tool links better. 455 00:36:23,960 --> 00:36:30,559 Yeah. Yeah, I think we're we're getting there. It is taken a 456 00:36:30,679 --> 00:36:35,239 while. Oh yeah, I mean there is a reason for it took for 457 00:36:35,280 --> 00:36:39,159 it to took that amount of time. I mean there was no support in 458 00:36:39,239 --> 00:36:45,000 Asia. I don't know about a WS, but the only moment where you 459 00:36:45,000 --> 00:36:52,039 could use RPC in a real production environment was if you had a virtual machine 460 00:36:52,639 --> 00:36:55,679 where you can't run the service. Yeah. I don't really know how many 461 00:36:55,679 --> 00:36:59,880 people are doing that anymore. Yeah, that's not a reason to run a 462 00:37:00,079 --> 00:37:02,760 VM. Reason running VM is you're trying to get you know, you're on 463 00:37:02,800 --> 00:37:06,719 your way off of it. Not that I'm building new things. I need 464 00:37:06,760 --> 00:37:09,840 a VM for it. I fairly enjoyed a moment where I had to run 465 00:37:12,280 --> 00:37:20,280 APIs as Linux services right machines start though, see what's happening. I was 466 00:37:20,360 --> 00:37:22,840 very excited to do that. That's a good way to slow people down from 467 00:37:22,880 --> 00:37:24,400 running a VM. You can run a VM, but only if it's Linux. 468 00:37:27,519 --> 00:37:32,199 I had to learn if your commands linux pseudo, make me a sandwich, 469 00:37:37,039 --> 00:37:42,360 Pseudo, make me a pseudo? What are what is server reflection and 470 00:37:42,400 --> 00:37:46,920 how does that play into all of this? So server reflection will allow you 471 00:37:47,039 --> 00:37:55,880 to discover what Jersey services you have available inside an endpoint services operations in fact 472 00:37:58,559 --> 00:38:02,519 and the tooling. I'm going to talk about tooling, especially about Postman in 473 00:38:02,559 --> 00:38:07,079 here, because now it allows you to have like an ur L of the 474 00:38:07,119 --> 00:38:12,239 service and your server reflection, and if you have reflection enable on the server 475 00:38:12,360 --> 00:38:17,480 side, it will discover what can't you with that swagger for GRPs kind of 476 00:38:17,760 --> 00:38:23,199 kind of yeah yeah, And you can also execute those because Postman will generate 477 00:38:23,599 --> 00:38:29,239 messages for you domulance, but you can use those in order to call those 478 00:38:29,320 --> 00:38:34,559 operations. Nice. Would you use discovery in production or is this really just 479 00:38:34,599 --> 00:38:37,480 a tool when you're developing to make it easier to get hooked up. I 480 00:38:37,480 --> 00:38:44,360 think it's meant to be used with an API gateway or something like that, 481 00:38:44,880 --> 00:38:47,800 right, so sort of a self service discovery. Otherwise I don't see the 482 00:38:47,840 --> 00:38:52,960 point too like, I mean, I think it kind of should know your 483 00:38:52,079 --> 00:38:55,639 end points right right, and never if you have with this whole we don't 484 00:38:55,679 --> 00:39:00,639 know where end points are. Yeah, No, because it's g RPC is 485 00:39:00,679 --> 00:39:05,320 meant for point point communication, and that assumes that you know where you will 486 00:39:05,360 --> 00:39:09,960 make requests and what should look like. Yeah, because you have the protocol 487 00:39:10,000 --> 00:39:16,039 buffers, files and types that describe basically what you have available. Now you're 488 00:39:16,079 --> 00:39:20,000 twisting my brain. It was like, am I going to generate the proto 489 00:39:20,079 --> 00:39:30,440 buffs on the fly because I've used discovery to making right now just because you 490 00:39:30,559 --> 00:39:35,480 could do it? Yep? Yeah, all right, I mean, but 491 00:39:35,519 --> 00:39:38,119 I appreciate you want something discoverable, you know, nothing else, just to 492 00:39:38,239 --> 00:39:43,599 spit out an error message, right you have a I don't know. This 493 00:39:43,599 --> 00:39:45,920 has been a thing for me lately. I kept trying to write a better 494 00:39:45,000 --> 00:39:50,400 error message because object not found pisses everybody off. Like I kind of got 495 00:39:50,440 --> 00:39:52,480 this reflexed. If we made a call and it failed, then I'd like 496 00:39:52,519 --> 00:39:57,320 to invoke the discovery and include it in the log. So it's like, 497 00:39:57,360 --> 00:40:00,760 hey, you know, I failed to make this call, but I did 498 00:40:00,760 --> 00:40:04,960 the discovery on it, and this is what it's actually expecting. Like you've 499 00:40:05,039 --> 00:40:07,920 just shortened everybody's pain up a whole bunch, right, you know, I 500 00:40:07,000 --> 00:40:09,480 have to go back and look. It's like at the moment when I made 501 00:40:09,519 --> 00:40:13,000 this call, this is what it was expecting, and this is what we 502 00:40:13,079 --> 00:40:17,079 SENTENCEDEAD. Yeah, but to be here reflections So far, for me, 503 00:40:17,400 --> 00:40:22,800 it has been useful just to play around to discover like what's available on the 504 00:40:22,840 --> 00:40:29,159 path that I know, and with Postman because it allows you to like load 505 00:40:29,199 --> 00:40:34,639 the live RL and to see what's there and to invoke it for testing purposes 506 00:40:34,679 --> 00:40:37,039 of course. Nice. Yeah, and you like those tools, so it 507 00:40:37,079 --> 00:40:39,599 just gives you confidence of what's going on. Oh yeah, I mean I 508 00:40:39,639 --> 00:40:45,639 always prefer a tool where I can click three buttons to make sure the thing 509 00:40:45,679 --> 00:40:51,079 that I want it to work works. Then I'll manually create a client try 510 00:40:51,119 --> 00:40:55,599 to debugget see what's the error if any, so it's easier. Yeah, 511 00:40:55,719 --> 00:41:00,679 I like that. I have a question for you, a particular question about 512 00:41:00,719 --> 00:41:04,639 streaming in gRPC which I've been doing for a while, but I'm not really 513 00:41:04,719 --> 00:41:07,320 sure what's going on under the hood. So in your protofile, if you 514 00:41:07,440 --> 00:41:15,800 have an RPC that returns a stream of you know, some message type that 515 00:41:15,000 --> 00:41:22,320 instead of you know, a repeated message type, then when you implement that, 516 00:41:22,400 --> 00:41:28,880 you would think that there would be an I sink enumerable in there somewhere. 517 00:41:29,679 --> 00:41:35,039 But instead what you get is the request and an I serve stream writer 518 00:41:35,320 --> 00:41:40,239 of your type, you know, and then a server call context and then 519 00:41:40,360 --> 00:41:45,639 you basically can just await you know, response streamed dot write a sink whatever 520 00:41:45,679 --> 00:41:51,000 your object is in a loop and it you know, I would I was 521 00:41:51,039 --> 00:41:55,639 expecting an I sink numerable in C sharp. Is that's what's actually happening under 522 00:41:55,679 --> 00:41:59,440 the hood, To be fair, I don't know for sure, but to 523 00:41:59,519 --> 00:42:04,719 be fair, I don't think that is what happening under the food And I 524 00:42:04,760 --> 00:42:09,559 think they allowed us to have that stream reader writer because you have the both 525 00:42:12,239 --> 00:42:15,920 just to make it as low level as they can. Yeah. So you 526 00:42:15,000 --> 00:42:20,159 have this stream, your job to do whatever you want with it. Yeah. 527 00:42:20,239 --> 00:42:22,480 I realized this is an answer I could easily find by going and looking 528 00:42:22,519 --> 00:42:27,960 at the source code and get hub. But I just wondered if you knew. 529 00:42:28,079 --> 00:42:31,400 I'm sorry, I'm going to air quote easily. Yeah, easily. 530 00:42:34,159 --> 00:42:37,800 You're right. I can. I can answer this question myself because I can 531 00:42:37,039 --> 00:42:39,559 go and look at the source code, cut and paste it into get Up 532 00:42:39,559 --> 00:42:43,679 Go. Pilots say what the hell does this mean? I can't ask chat 533 00:42:43,719 --> 00:42:45,360 jipity. I was just gonna say, or you could ask chat jipity. 534 00:42:45,559 --> 00:42:49,000 Yeah, we all went to the same thing. Give me a large language 535 00:42:49,000 --> 00:42:52,199 model is going to answer this question for me? Yeah? Not? Do 536 00:42:52,360 --> 00:42:54,199 I trust it? That's yeah, Now it wasn't correct. How would you 537 00:42:54,239 --> 00:42:59,400 know? It's like it's a great summarizer. If it wasn't a great summarizer, 538 00:42:59,599 --> 00:43:04,400 how would you you know? Yeah, sounds good, but always search 539 00:43:04,440 --> 00:43:07,199 and research of the fact that chad jipt gives you. Yeah. Yeah, 540 00:43:07,239 --> 00:43:12,079 you can say where what what's the get up page where you found this answer, 541 00:43:12,119 --> 00:43:15,880 and they'll say, I have no idea where on Google? I'm sorry, 542 00:43:15,880 --> 00:43:20,239 this was derived from a from a mirror site on four chance, so 543 00:43:20,280 --> 00:43:24,559 it must be correct. Oh my goodness. Okay, we were talking about 544 00:43:24,599 --> 00:43:30,360 gRPC. I mean because it's good. It's really good. As much as 545 00:43:30,400 --> 00:43:31,519 I think we've been hard on it today. Yeah, I know, it 546 00:43:31,559 --> 00:43:36,400 really is amazing. Yeah, but I appreciate tools like Postman, like, 547 00:43:36,480 --> 00:43:39,840 oh okay, yeah, if I was learning my way around gr PC starting 548 00:43:39,880 --> 00:43:45,000 to play with requests through Postman, so you can watch the interaction, that's 549 00:43:45,039 --> 00:43:49,280 awesome. That's awesome. Yeah, that's good stuff and get started. And 550 00:43:49,519 --> 00:43:57,239 it also makes the binary human readable, which is awesome for Postman because at 551 00:43:57,239 --> 00:44:00,400 first that was a show stopper, like how can I a thing that is 552 00:44:00,519 --> 00:44:06,599 binary? I don't know how to decipher binary? So this will makes it 553 00:44:06,880 --> 00:44:09,119 way easier. So training is a big part of what you do. Are 554 00:44:09,199 --> 00:44:13,800 there any regular classes that you teach online where we could learn this stuff? 555 00:44:14,360 --> 00:44:21,360 Not? Up until now, I'm basically working on business to business contracts. 556 00:44:21,440 --> 00:44:24,440 So there are companies if they called me, hey, we need to learn 557 00:44:24,480 --> 00:44:29,559 about this or that, I can provide them some trainings around it. So 558 00:44:29,679 --> 00:44:32,039 if somebody who's interested in how many what's the minimum number of people that you'd 559 00:44:32,079 --> 00:44:37,400 want to teach too? Minimum? I did it for five people because the 560 00:44:37,440 --> 00:44:46,559 company wanted one, right, But to be honest, I always prefer like 561 00:44:46,639 --> 00:44:52,480 having a bigger group of people. Sure, so the questions are way easier 562 00:44:52,519 --> 00:44:59,440 to answer because from thirty people or fifteen people, you'll always find some questions 563 00:44:59,440 --> 00:45:02,639 that are into and they will start debates. And I always love the interaction 564 00:45:02,719 --> 00:45:07,719 with you put your hands back, sit back. Now the classes it's called 565 00:45:07,719 --> 00:45:12,159 by itself. See what I started. Yeah, but it's everybody working on 566 00:45:12,159 --> 00:45:15,039 one project and you can get into a discussion about the project like that's that's 567 00:45:15,079 --> 00:45:21,199 good to do, yeah, without a doubt. But you also had some 568 00:45:21,480 --> 00:45:28,559 online training stuff too, some videos maybe oh yeah yeah, yeah, uh, 569 00:45:28,880 --> 00:45:35,239 forget about those. No I'm joking, No, I'm joking. Oh 570 00:45:35,280 --> 00:45:38,159 it was so hard to record record those. It's a whole new type, 571 00:45:38,239 --> 00:45:46,480 so different. Oh yeah, a different life. Yeah. So uh oh 572 00:45:46,559 --> 00:45:52,039 yeah it's hard. So uh big thumbs up for people that are doing this 573 00:45:52,159 --> 00:45:57,440 for a living or full time their content creators. And to put your cell 574 00:45:57,440 --> 00:46:04,280 phone camera and to speak like coherently. Yeah, it's it's a thing. 575 00:46:04,760 --> 00:46:08,199 Yeah, So coherently coherence comes with editing though, you know that's the good. 576 00:46:08,440 --> 00:46:15,559 Yeah, if you know how to like take breaks or have stop words. 577 00:46:15,760 --> 00:46:19,679 And I learned that the hard way. I do that. I subscribe 578 00:46:19,679 --> 00:46:22,840 to the max Headroom style of editing, where everything I just jump cut all 579 00:46:22,880 --> 00:46:27,840 the time. Yeah. As long as you can hear me and understand me, 580 00:46:27,920 --> 00:46:30,800 you can put up with a twitch or two. Yeah I think that 581 00:46:30,920 --> 00:46:38,360 and the hard hair. Yeah yeah, exactly my mom up. So if 582 00:46:38,400 --> 00:46:45,280 anyone wants to to learn more about gRPC, I have a class around four 583 00:46:45,320 --> 00:46:51,320 hours around g RPC from zero to Hero. You can find the link on 584 00:46:51,360 --> 00:46:55,039 my blog. It's there with a small discount. Excellent. Yeah cool. 585 00:46:55,079 --> 00:46:59,199 So what's next? What's on your what's in your inbox? To be here? 586 00:46:59,440 --> 00:47:04,280 In need or want to start a YouTube channel, I have a gazillion 587 00:47:04,320 --> 00:47:08,719 plans. Man. Oh yeah, meteor, because I said I did this, 588 00:47:08,760 --> 00:47:14,719 I did the course. It's hard. It might be some learning curve 589 00:47:14,719 --> 00:47:19,159 for me with YouTube tool, but it might also prove useful. Well. 590 00:47:19,199 --> 00:47:21,599 Feel free to reach out to me because I've been doing it for a long 591 00:47:21,639 --> 00:47:23,400 time and if I can help in any way, I will. For sure. 592 00:47:24,280 --> 00:47:29,480 I'm gonna book a slot in your calendar. You got And of course, 593 00:47:30,079 --> 00:47:31,840 by the time this show comes out, we'll have already been to NDYC 594 00:47:32,000 --> 00:47:35,840 London and back again. But I will you and I will see each other 595 00:47:35,880 --> 00:47:38,639 in the NDC London. Yep. Very good. Anything else that we forgot 596 00:47:38,679 --> 00:47:43,920 to talk about arena. Nothing that crosses my mind. All right. Well, 597 00:47:44,159 --> 00:47:45,920 on that note, let's call it a show. Thanks very much for 598 00:47:46,000 --> 00:47:50,119 spending this time with us. It's always great to talk to you same, 599 00:47:50,599 --> 00:47:52,800 Thank you. All right, We'll talk to you next time on dot net 600 00:47:52,920 --> 00:48:19,000 rocks. Dot net Rocks is brought to you by Franklin's Net and produced by 601 00:48:19,079 --> 00:48:24,000 Pop Studios, a full service audio, video and post production facility located physically 602 00:48:24,039 --> 00:48:30,000 in New London, Connecticut, and of course in the cloud online at pwop 603 00:48:30,239 --> 00:48:34,119 dot com. Visit our website at d O T N E t R O 604 00:48:34,199 --> 00:48:38,599 c k S dot com for RSS feeds, downloads, mobile apps, comments, 605 00:48:38,800 --> 00:48:43,920 and access to the full archives going back to show number one, recorded 606 00:48:43,960 --> 00:48:46,880 in September two thousand and two. And make sure you check out our sponsors. 607 00:48:47,039 --> 00:48:51,679 They keep us in business. Now go write some code. See you 608 00:48:51,719 --> 00:49:10,760 next time you got javans My is home, then my Texes in a line. Credit ballh