Speaker 1 00:00:01 Join us as we gather around the hedge, where we dig into technology, business, and culture with the finest minds in computer networking. Speaker 2 00:00:20 Well, hello Tom. It's good to see you today. Speaker 3 00:00:23 You too, Russ. Speaker 2 00:00:25 I read someplace that Austin is one of the fastest growing cities in the United States right now. Speaker 3 00:00:30 I would believe it three years ago, the traffic was not nearly as bad as it is now. Speaker 2 00:00:35 Wow. Yeah, I was, I mean, Knoxville's growing, but it ain't growing like Austin is right now. I think it's bad and good things about that. Right? Yep. I don't know. Yep. And then today we have the She shed, I think there is a, she in the She Shed. I am, Speaker 4 00:00:52 I'm here. I'm here today. I got to expose a whole new group of people last week to the She Shed. Oh. Um, because I was, I was traveling and doing some cross team meetings, and so yeah, it was fun. Speaker 2 00:01:05 Nice. Oh my goodness. So, so virtually, I hope you weren't doing tours behind your house. Speaker 4 00:01:10 Yes. No, I, I, um, so in, during the creation of the She Shed, there was a large running document that I created. Oh, okay. And I was able to share that. So yeah. Speaker 2 00:01:20 Cool. I was actually kind of near, not near Austin, I was outside of Dallas last weekend. That's where I caught this cold. Tom. Speaker 4 00:01:30 I was only in the Dallas airport last Speaker 2 00:01:32 Week. Yeah. I might have overlapped with you there a minute for, for Yvonne. I don't know, because the flights were so bad last week. Speaker 3 00:01:41 Yep. I, I was in my home office last week. It was great. Good Speaker 2 00:01:45 For Speaker 4 00:01:45 You. Mm. Speaker 2 00:01:47 I was, I was at the sales conference, not that anyone knows what that is, but that's where I was. Denton, Denton, Texas. So, yeah. And my flight, oh my goodness. The flights were a disaster last week. I have no idea what happened, but like, I tried to fly home on Saturday night and I just couldn't get home. I thought I was gonna have to get a hotel room in Dallas or something and like stay for a week, . I could not get any, I could not get back to Knoxville. It was so ridiculous. I found a flight back to Chattanooga. My wife said, I'll just drive to Chattanooga and pick you up. It's only an hour and a half. That's how bad it was. It was like, oh my goodness, this is terrible. So, I don't know. It's crazy. So here we are in another round table, even though we're really kind of vertically aligned or side aligned on Zoom and, you know, maybe I need to find a round table someplace. You know, remember the old Cisco video conferencing systems where they had the half the round table stuck against the wall, and then the other participants were on the video monitors around the other half of the round table. So it looked like you were looking over a conference room round table today. It's, that's like I do, I mean, we need that. This feels like over-engineering , but Speaker 3 00:03:06 , Speaker 2 00:03:09 That's what engineers do. They over-engineer things. This is true. Yeah. It's, you know, the pessimist says the glass is half empty. The optimist says the glass is half full and the engineer says the glass is over-engineered . Oh no. It's just too large for the amount of water you're attempting to hold. So, alright. Today we start with an article for Larry Peterson at the systems approach talking about open source value proposition. So talk us through this, Tom, what's going on here? Speaker 3 00:03:43 So he, uh, is, is basically talking about, um, the value of open source, which has been written at, written about at length. Um, but he had, there are a couple of things in here that I, that just really impressed me, made me think of things in a way I hadn't before. Um, one of the things that I thought was really cool was he talked about how in the SDN days they started to write about how to build SDN systems. And it became pretty clear that in order to really explain the whole thing, they had to explain the whole stack, which meant all the stuff that goes into a network operating system. And that is, up until that point really had been the domain of, of networking vendors. They were the only ones who really cared about it. Why would anybody else care about writing an os? Speaker 3 00:04:22 And so, um, as they started to write about SDN concepts, uh, obviously they came into the, the OS that's running one of the elements forwarding packets. And so, um, I thought that was really interesting that up until, you know, up until that point, um, mostly, so a lot of what they were doing, they, they have a, they, they wrote a book about SDN and they were talking about reproducing functionality that already exists, which totally rings true to me in my current experience. And, and, but the interesting stuff was that now that that has has taken place, you're starting to see, um, computer science courses. They're talking about how to build a network operating system, which you didn't have before. It was if you wanted to know that you'd go work for a vendor writing a commercial nos and it would, you know, largely be a proprietary thing. Speaker 3 00:05:08 And so having, you know, them doing this writing kind of opened it up. The other thing that I thought was really interesting in this article was about, um, they had, they've written another book about 5G and private 5g and in this book they go through a hands-on exercise of bringing up your own private 5G with open source software, which I didn't even know was possible. I didn't know that Open ran got that far that you could actually install software and it works without having to write a bunch of stuff. Uh, which was super cool. And that, you know, this link that led to another link in another article that kind of got me on that whole thing. And they explained, um, you know, doing that and I don't know. And, and the last little point I got out of it was that, um, if you, if it's open source, you get to see how the machine works. And that's important. It's important not just to live in PowerPoint land, but to actually look at how stuff works. So that was the I thought it was a cool article. Speaker 2 00:05:55 It is. Cool. I mean, and the thing is, is that he talks about TCP a little bit here is that we don't really think about it, but a lot of what we know in networking is and happened because of open source and Right. You know, it's not just open standards. Like we think about the TF writing a standard, well they don't just write standards. There has to be interoperable implementations. And the standard has to be good enough that someone can write an implementation from the standard. That level of documentation for protocols didn't exist. And so it was impossible to break into the market unless you were a big player like IBM could write your own system 360 code and you know, all your own communication protocols and Noval NetWare, or in fact, a lot of people think Microsoft kind of put Noval NetWare and and stuff like that. Speaker 2 00:06:45 And Linux put the Microsoft, uh, Noval NetWare and the Microsoft operating systems and stuff out of business. It's really not true. What really I think happened was they standardized all the protocols and then like there was no point in the proprietary stuff anymore. And so, yeah, it just kinda like, it, it's just, and I don't know Yvonne, if you have something to throw in there, but like it's, for me it's even colleges, right? Colleges don't want to teach proprietary stuff, right? So you get all of the colleges and you get people to teach open source stuff in colleges and all of a sudden it's gonna win the marketplace. Cuz that's all anybody knows is what's in the colleges. Speaker 4 00:07:28 You know, I I, my, uh, my oldest is a software developer now and we've been having some conversations and he's working on technology stacks, um, with which I'm not terribly familiar, but we'll, we'll talk through that and I'll say, okay, you're using this thing, tell me what it is. I'm like, oh, okay. You're talking about an abstraction for X. Or Oh, okay, you're using this tool and that really fits in this part of the stack. And I think one of the things that these kind of efforts give us is, is categories and help us understand, um, yes to Tom's point how things work, but, but what the individual pieces are and why they matter. Um, it's, it's interesting that, um, that there was not a lot of conversation around creating, um, you know, network operating systems UN until this kind of effort was put forth. Speaker 4 00:08:20 And there's, and it's, and I find it equally interesting cuz there's all kinds of domains of knowledge out there, right? That are in pockets that are accessible only inside a very difficult to reach proprietary places that when you, uh, put them out in the marketplace, it, it creates a platform that you can build on, you know, more, more broadly that people will adopt those ideas and use them, um, for other, other pursuits maybe in a creative way for which they were not necessarily intended. And we've seen that networking protocols forever, right, Russ? Like you build it Yeah. And then you have no idea what people are actually gonna use it for. Speaker 2 00:08:57 Yep. The 5G ran example is a great example and I, you know, I actually of course have been following it so I know how far along they've gotten. And it is incredible how far the cell phone providers have gotten with open source, open, openly available pushing the 5G ran story. And I'm not convinced of the 5g, oh, it's gonna have less latency. Oh, it's gonna be so much more bandwidth. Um, I read an article the other day about how they're trying to pursue junk bandwidth now they're trying to find ways to, to use statistical multiplexing and stuff to get into bandwidth ranges, into frequency ranges that can't be used because they're just, they're just, they don't carry data very well. Um, you know, like for instance, you get to really high frequencies that the wave wavelength is so short and you lose power so fast. Speaker 2 00:09:49 It's a typical power law. You lose three DB every wavelength basically that you go. So it's, you know, you're having your signal basically every time you, you move one wavelength. So when you get to wavelengths that are nanometers long, your power goes away immediately. Like there's no power after like, you know, two feet. And so how do you transmit, this is where NF NFC and stuff like that come in is they use these really high frequencies so they don't have to go very far. But anyway, I guess my point was is that, you know, you have all this stuff going on and these guys have been pushing these networks, but I think the biggest thing that comes out of 5G is the open ran, in my opinion. I think that's the biggest thing is that it could make it where it's possible for a small local provider to build a 5G deployment for their city and do fixed wireless access to the internet and not have to buy any proprietary equipment to do it, do the whole thing software. I think it's, I think it's fantastic. Speaker 4 00:10:55 Well, and I'm, I, I like Martine Cato had had a project, and I heard about it a few years ago, I don't know if it's still ongoing, but where, um, they were supporting tribal lands, right? And every school system in those tribal areas had connectivity and they were using open source tools and um, you know, commodity hardware to build, you know, connectivity for these communities that otherwise wouldn't have it. Right? And so there have been some efforts like that to do it. Um, and I think there's a, there's a lot of, you know, value for humanity Speaker 2 00:11:26 Yeah. Speaker 4 00:11:26 To have that kind of technology available. Speaker 2 00:11:28 Yeah. It's very, it's very interesting. And I think it's, um, also, it's also been helpful even against wifi because now that wifi has real competition, the people who manage the wifi space and everything else are suddenly simplifying all the, it's now wifi six and wifi seven. Like it used to be wifi 8 0 2 11 A P C G, and he had all this stuff and nobody knew how to run it. And you had to be an expert on wifi. And now it all largely just works, right? And that's a result of, well, if we don't do this, then we're going to end up with 5G taking over all of our wifi applications. Speaker 3 00:12:11 That, that's the thing I think is maybe a little counterintuitive about open source, the idea of you build something out of open source and just works all. I think most of us think those are two different things that you can't have one without the other. We just think of open source like that, like just works and d i y or like they're diametrically opposed. But I think I, I think the trend that we see is that yes, those things are diametrically opposed because you have to deal with all the reality. But, but I think the, that open source drives just works in other parts of the industry. Yes. When something commoditizes something else, then the thing that was not the commodity now has to improve in order to become profitable. And, and I think that, that this sort of thing will have the same effect on networking. Speaker 3 00:12:54 I think the stuff that can be commoditized will, and eventually stuff will become just as reliable as like the Apache web server. I know I use that in all my examples, but like, but then the stuff that's surrounding it is going to get better. And if everybody just said, oh, well no, open source is too hard. We can't do it. We're not going to do it. We'll let someone else do it, that's fine. You'll end up with premium price products that don't work very well. And that's kind of how, that's how it will work. But, uh, open source I feel like has this influence in industry, um, through the incentives that it provides. And maybe that's the same thing that's, that's going to happen to networking, you know, as a whole. Speaker 2 00:13:28 Yeah. There's a danger of course, because DNS is open source, find is open source and it just works. And so because it works, it's a working system that we can pile new functionality into Speaker 3 00:13:41 , Speaker 2 00:13:42 Right? Right. And oh, I'd rather use something that already works than something that I have to develop. Oh, look, this already has market share, it has mind share, 5 million people deploy it, why not just pile into that? And that drives centralization to some degree. And it also drives complexity because now you look at dns, like I'm involved in a project right now where I'm working with DNS and I don't always get involved in dns. DNS is kind of like, not exactly like I'm a routing guy, but still I get involved in dns, like when I was at ign. And now that I'm at outcome, I do a lot of DNS stuff and I am amazed at how complex S is and how much it really doesn't follow the standards. Every implementation has a little side thing that they added to make it better. And you're like, stop making it better. Speaker 2 00:14:37 Okay. That is just, you're breaking things and it makes certain things about DNS nearly impossible to figure out. And so, I mean, we just keep tossing stuff and now we have D records or whatever they're called where we talked about this on the hedge in the past and some, some episode a long time ago about how complex these records that are wild cards and how you can no longer analyze a zone file to make sure it's correct. And it's, and it's not atomic any longer. Like BGP is, it's not atomic. BGP never converges because BGP converges differently every time it converges. And so you get enough BGP speakers in the world and everybody's converged different, converging differently every time. And it just never converges it's done. And so, you know, you tolerate it. So there is a negative side to this whole open source if, if it just works thing, right? So it's cool, but, but it's a little dangerous sometimes Speaker 4 00:15:36 . But I think you also said something earlier, Russ, about people looking at, at a new technology emerging, especially open source technology and thinking like it'll never be efficient enough for X, Y, Z. Yeah. And if you follow the history of technology period, whether it's, you know, computer technology or not, you build first, then you optimize, right? Yeah. Like that is the process. And so I think sometimes we get shortsighted and we think, oh, well that will never be fast enough. We'll never work x, y, Z enough. Um, and ultimately, like you, you, you have to build a thing before you can optimize it. And, and I, we've seen that happen over and over and over again with, um, yeah. It'll never, you know, 64 K is all you're ever gonna need, right? I mean, I, what whatever the, the technology is, um, build, then optimize and we sometimes give up on stuff a little early. Speaker 2 00:16:28 Yeah, we do. We do. And sometimes we over optimize too, unfortunately. Speaker 4 00:16:34 Over engineer, over-optimize. Yeah. Mm-hmm. . Speaker 2 00:16:37 Yeah. Yep. We build the glass twice as large as it needs to be to hold the water. So Speaker 3 00:16:44 Go ahead. I thought it was just my last note on this. I just thought it was really cool that the, I mean, I really like the writing that these guys do, and I thought it was really cool that they have a book about 5g and a significant part of the book is, Hey, try it for yourself. Yeah. Um, I think that's just really innovative and somebody who wants to, and by the way, if anybody's listening and wants to become an expert in something, the people who actually know how 5G works, there's not that many of them in the world. So, um, you know, that would be a, this would be a great way to learn that if you're interested. Speaker 2 00:17:12 Yeah. Just remember that it's all mobile ip, but they've used different words for everything. Speaker 3 00:17:23 Yep. Speaker 2 00:17:23 And they have special names for every place you can plug anything in. And you will be perfectly fine . I mean that's really, I I really, sometimes I read the 4G 5G stuff and I'm like, that's just mobile AP what you've just done. I mean, I don't wanna fuss or anything, but that's kind of what you've done here. You just called it something different. That's kind of cool. But, you know, I don't know, just change the names to protect the innocent and whatever. So the second thing is off of, um, thinking Sideways and Yvonne tossed this one out, which is, um, using Kanban and fake agile companies. Now, thinking Sideways, um, has been on a thing about fake agile for a couple of blog posts, which I find an interesting concept. Uh, for a long, long time ago, we had Mike Boon come on and talk about Agile, right? I mean, he's like one of the very first hedges we had, we had Mike Buchan come on and talk about Agile and using Agile just as a kind of a, as a, as a workflow structure thing, which seems to be what he's talking about here, right? Like you're just, you're just faking it. You're not actually doing agile, you're just kind of like, yeah, we'll use it as a workflow. Is that, is that kinda your impression, Yvonne? Or is there Speaker 4 00:18:50 Well, I, you know, it's, it's Agile development is more about a philosophy and a mindset than whatever tools you use, right? Mm-hmm. It's about empowering your people. It's about working in short sprints. It's about not clearly defining, you know, like, like we did back in waterfall, like every single feature at the very beginning, it's about changing quickly. It's, it's , you know, shifting. It's about Speaker 2 00:19:15 Shifting, testing left, shifting, troubleshooting, left Speaker 4 00:19:18 Feedback loops, right? Like there, there are all these things that are part of, um, of agile development. And what's happened is we've tried to take the tools that are best fit to agile development and force them into other models and to say we're being agile, to make companies feel like they're being innovative when really they're operating in the same old way, the same old mindset, the same old methodology. And they've just shunted new tools into that process. And so I think like it's, it's important to learn to recognize the two, and when, when, um, you know, an organization says, Hey, well we're, we're going to all Agile development and or you're interviewing for a team and they're saying, yes, we are agile. And they say, you know, we have a ban board and we do this, that, and the other. You know, to, to ask some questions. Speaker 4 00:20:11 You know, you know, who's responsible for determining what the product activities are? Where do those decisions get made? Um, you know, at what point in the process can you, um, change the, the, you know, feature requests? Is it, is it actually every two weeks at the end of a sprint or does it require three levels of management approval to change what the product ultimately is gonna look like? And I think just the concept of a fake agile company and what those tools are, um, and how we confuse tools with culture and mindset and overall processes is important because we, uh, we conflate our terms a lot. Um, I'm a big fan of Kanban boards, um, and this article's absolutely right. You can use a Kanban board for just about anything, just tracking work, right? Um, but there's a set of mindsets and philosophies that are part of agile development that are really where the magic is not so much the particular tools that you use. Speaker 2 00:21:14 Yeah. Yeah. As conclusion, I consider switching to Kanban as the best course of action when working in a fake agile company, because that is a high chance to succeed. Speaker 4 00:21:27 As opposed, when he says turning a fake agile company into an actually agile one is tremend is a tremendously difficult task, which requires buy-in from a lot of powerful stakeholders. Now, he's a, he's a little bit more bearish on this idea than, than than I am. I mean, I think it's, it's possible, but I do think he's right. Like you've got to have buy-in from the, the people who are making the decisions mm-hmm. to redirect them back to the team and to say, I'm not gonna make this decision because we are changing the way we work and these decisions need to be made differently here and here. Yeah. Um, and very rarely do you have a leader who's, uh, got the fortitude to, to make all that happen. But, uh, well, he said Speaker 2 00:22:09 Something, yeah, ahead. I said something up or above where it talks about, um, scrum team members need to respect each other to be capable independent people and are, and are respected as such by the people with whom they were. So Tony p when I talked to him about Agile many, many years ago is he had an entire presentation he gave at the college level about, okay, the reason as agile is not working for you is because your team is not disciplined and experienced enough to really do agile. And so you're trying to push agile onto a team that doesn't have the respect the interpersonal relationships and the respect and the knowledge to be able to make these kinds of decisions to today. I mean, you can't take an organization with a product management team that normally tells the, that normally has discussions with the engineering team and the sales team over what to build and turn it into an agile company. Unless you integrate the product management team with the development team and make them one team, you can't do it with two separate teams. And a lot of companies say, well, we're agile on the development side, but then we have this project PR product management team. Like then no, you're not agile. Sorry. I, Speaker 3 00:23:28 The thing I like about this article is, seems almost tongue in cheek. It's almost like a little cynical maybe. Like, yeah, you're probably not really agile. So since you're not really agile, let me give you some lower grade advice that you'll actually be able to implement. Yeah. Like I , I really like the, the style because of that. Like, well, you know, it does seem a little cynical, but, but it is good to be successful in something and maybe some people will read this and they'll realize that we're not actually agile and maybe that'll help people. Speaker 4 00:23:58 Yeah. When I, I recently had a conversation with somebody who has worked in an agile kind of environment for, for a long time, for, for several years, relatively senior person really knows their stuff and kinda got thrown into a team situation, um, where they were using some of the tooling, but the team philosophy didn't have an agile mindset. And it was, it was really troubling for this individual, right? Like, nobody wants to hear what I have to say, nobody cares. I've worked in this industry for, you know, a few decades and I'm not being treated as a peer and equal member. And I think like it's, it's that, that cultural piece is incredibly important. I if you wanna get the best out of everybody on the team, and yes, cross-functional teams are incredibly important, having all of those things. But, but you're right, Tom, I also appreciate the, the, the sort of tongue in cheek and yeah, you can, you can use your to ban board, but don't, don't confuse that with thinking that you're actually agile. Speaker 2 00:24:56 That, that you're actually agile. Yeah. Yeah. And, and I think that's, yeah, this blog, by the way, um, thinking Sideways tends to be a little snar gear than a lot of the blogs that I follow, which is fine, I I don't mind snar. Speaker 3 00:25:11 Um, it, it make, it makes me wanna write like a network automation thing in this style . Speaker 2 00:25:15 Yeah. Oh yeah. Like, Speaker 3 00:25:16 Yeah, you're probably not really gonna do it. But here's a couple things you could , Speaker 2 00:25:20 This kinda reminds me of a talk that Eric Osborne and I used to do at Cisco Live that was IP based traffic engineering versus, um, MPLS based traffic engineering. And basically the whole talk was Eric and I sniping at each other and cutting each other on like, no, that's not true. Or whatever. And uh, yeah, it was a lot of fun, but in the old days we could do that kind of thing. Speaker 3 00:25:47 , we should, we should have him come back. I haven't talked to Eric in like five years myself. I know. We should have him come on the show. Speaker 2 00:25:53 I haven't either. Yeah. I'll, I'll see if I can get in touch with him. I actually don't know what he's doing now. So third article is about network spending priorities, which is, you know, usually very boring for network engineers, . And the basic takeaway from the article I get is by Tom Noel, by the way, is that security and operational stability and application delivery performance are the big deal right now. That's where people are spending money. I don't know, does that match anybody's reality out there? I don't know. What do you think, Tom? Is that where you're seeing ? Speaker 3 00:26:33 Uh, no, I mean I, infrastructure is always aging. Um, all kinds of infrastructure. Um, and network infrastructure never stops aging. And so, I mean, I, there are plenty of people that are ripping out all of their, all of their on-prem stuff, their on-prem routing and switching and I mean that, that's happening. People, vendors are still making big bucks on that. I don't think it's, I don't think that thing, I don't think that's stopped be for anything. I, it, a lot of things paused because of supply chain, but the demand is still there to say that nobody's spending anything on that I think is not correct. Speaker 2 00:27:09 Hmm. Yeah, cuz I mean, he says in here, networks are network operators are trying to cut their cost even though they are seeing slightly increased spending year over year. And, you know, they are trying to reduce the cost of network services and they would like to refresh, refresh, stretch, refreshing of old gear. Now that sounds great on paper. I'm just gonna tell you my experience with, with, with stretching the refreshing of old gear is it's really not a good cost benefit trade off. There's not a good ROA on that in the long term. And I've seen it time, Speaker 4 00:27:50 It's like keeping your car a year longer, but you're putting, you know, several hundred dollars a month into repairs. Exactly. Right. I mean, and you're dealing with the headache and of, of having an unreliable vehicle, right. It's it's exact, it's really exactly the same thing Yeah. When it comes to infrastructure hardware. Speaker 2 00:28:06 Yeah, it is. And so I always struggle with that. I mean, I used to, I used to to talk to people who would say, well, you never see new equipment or, you know, you never see new equipment in a, in a, in a fire sale company that's going under. I'm like, okay, that could be either cause or effect, Speaker 3 00:28:27 Huh. Speaker 2 00:28:28 Right. Don't, don't, don't imply that just because, right. It doesn't, it could be they're going outta business because, and I know of a bank who bought another bank because their network was so old and they felt like it was cheaper to buy the bank, another bank with a modern network than it was to build a new network. I mean, that's, that's interesting. That happens. And so like that's crazy. But that happens. So I don't know, I mean there was a lot of talking here about ESD WAN being a big winner here. I don't know, I have the funny feeling that ESD Wan's Day is kind of gone is is come and it's kind of peaking and we're gonna see it replaced by things like network as a service and stuff like that. And, and I could be wrong, but that's just my, my feeling about where things are going because why bother with an SD wan n when you can do a network as a service and you don't have to do any of it. I mean, well Speaker 4 00:29:26 I have, I have believed for 10 years that ultimately in the long term, the private network is dying, right? As services move to public internet as we talk about zero trust, that's based on identity. Um, and we're securing our applications differently and not relying on the network to do as much of that which is happening. It's happening slowly, slower than I want it to or think it would, it's gonna be another decade I think before we get there. But I think like, you know, the, the line between internal and external networks is, is vanishing. And I think in the long term, the need for that, just like we've seen a continual degradation and the need for VPN tunnels, right? Yeah. Do people still use them? Yes, they're still out there, but are there other mechanisms we use instead of a site to site tunnel anymore? Yeah, there are. Um, and, and I, I think like technology's gonna go that way ultimately now it's, it's got a long tail on it, right? I'm not saying that, that you just throw your sdwan away or that active project that you're doing right now is a waste cuz it's not. You're gonna need it for a bit. But I do think, you know, the long margin of time is going to move the technology further and further away from a private versus public, um, paradigm for networking. Speaker 2 00:30:45 Yeah. There could be, I could be wrong. There could be pushback on that though. There could be in the long run pushback on that. Some people in the industry are starting to get a little like, upset about it, but I don't know what they can do about it. But I'm just saying, you know, again, that pushes towards centralization and a lot of people really don't like centralization right now. That has become a, that's become a big deal. And so yeah, I don't know. I mean, the other side of that could be that we see enough open source stuff out there like the five, the the uh, the 5G open ran stuff and maybe Sonic will actually mature at some point. I don't know, maybe it'll become a, maybe it'll grow up and be a operating system and we can actually deploy it and have people actually download it and deploy it instead of, Tom's been spent two years trying to get it to, or a year and a half or whatever, trying to get it to run and compile and find all the bugs and everything else. Speaker 2 00:31:40 And it, it still feels very much like a a do it yourself project with, but, but you can't buy consistently sized two by fours to build your walls out of. So you gotta, like , you gotta like put little washers and stuff in to make all the two by fours the same width so you can put your drywall up and the drywall comes in weird little pieces and you gotta put it together like a Rubik's cube or something. But anyway, I mean, I'm gonna try to be mean as that's, I don't know, it just feels that way a lot of times. Speaker 3 00:32:10 Well I think, I think there's a hook in here from the first article we talked about. I, you know, SD n is actually, I'm not gonna say it's easy to build. I know a bunch of people would get mad at me if I said that, but, but it is, it is fairly straightforward to build tunnels over the internet with open source software and, uh, I I question the need, uh, frankly for the feature set, the giant feature set in the SD WAN appliances anyway, like you, you can, you can build with a lot simpler, you can build a lot simpler networks on open source. And I think some people will do that. And the other, and I think that many more people though, to your point Yvonne, I think many more people will start to see the network, the private part of the network is simply an overlay that terminates on end user devices. Speaker 3 00:32:51 And then once that becomes, you know, network as a service, like once that, once that is seen, because cus companies have to do this anyway. They have to manage all their endpoints, they have to do all this ridiculous stuff to secure computers that they issue to employees. One more thing, one more smartly provisioned tunnel to network as a service is not gonna, no one's even gonna notice it. And so why would you build a private network when you already have to manage all your end endpoints anyway? I, you know, so I think, I think there's that and I think open source will make it possible for people to hang on to what they wanted to do. Um, so Yeah. Speaker 2 00:33:21 Yeah. Well I was gonna say, I think that a lot of the cloud stuff is gonna push open source more to an easier to manage thing in the long run. If, if the open source projects want to survive and that's gonna like, cause a blending of what the way people manage networks in the cloud versus managing networks internally so that it may not make that much of a difference from the operator's perspective, whether it's private or public cloud or privately owned or whatever. It's just gonna be, it's a network and I just manage it, you know, I have access to different things and that's just what it works. Um, I, I do personally see the danger in over centralization myself. I do that, that worries me a lot cuz like this openness, you're talking about time where people now know how to do network operating systems, give it 10 years and if everybody goes to the cloud and there's no private networks any longer, that knowledge will go away. Hmm. That's interesting because you won't be buying vendor hardware so it won't, the knowledge of how to build a routing protocol and how to build a network and everything else won't be tied up in the vendors. It'll be tied up in the cloud providers. And unless you go to work for a cloud provider, you'll never learn how to build a network at scale. Speaker 4 00:34:48 You know, that's really interesting that you say that cuz uh, there's a, there's a young man that I know who's interested in technology. He just graduated high school and he was having trouble with his pc and so I went over to try and help him out with it. And, uh, we, we've had lots of conversations about AI and cybersecurity and, and, and higher level kind of topics, but as we were working through like troubleshooting his pc it dawned on me how little fundamental knowledge he had about what all the components were and how they worked together. Whereas in, in one of my first jobs, my boss bought us all components and had us build our own PCs and part and some of that was part of our learning experience. Right. And, and it dawned on me then that, that, oh, like there's a generational shift happening where even for, you know, technically minded home PC gamers, like there is not this fundamental understanding of how the components work now, how much of that are they gonna need? Speaker 4 00:35:53 How much of that, like, you know, I we don't need as many farriers anymore right? In a world where we all drive cars and don't ride as many horses. So I think like what that balance is gonna work out to be is interesting. But I do absolutely see that generational shift happening and in, in, in the skill sets based on age and just, you know, when folks entered the technology arena. But I also don't think it's fair to expect them to go back and learn 75 years of, uh, of technology knowledge before they can start contributing. Right. We have to figure out what that balance is. Speaker 2 00:36:31 Yeah. But I also think, I don't know, I have a, I just have a sense, like Tom has talked about this before. Most of the people who are really good network engineers that I know came out of some other background. They didn't, they, they didn't come out of network engineering. It's not where they started. I started in electronics. Right. I mean, so I had this interesting experience this week while I was writing the Csst book, which is done, by the way, should be published in the next couple of months I think, depending on Pearson. Yay. So you can point your high schoolers at the Csst book by the way, because I didn't, I did not write it as a Cisco book. Oh, I shouldn't say that in public. Should I, I wrote it as a Well, I mean the good thing is the blueprint for the exam is very generic technically. Speaker 2 00:37:16 So it's not a, it's not a let's go figure out how to configure things on Cisco boxes from day one. Like the CCNA is the ccs or C SST is much more of a explain to me how beam forming works. So anyway, I forwarded a chapter about beam forming and wireless and stuff. And this person was all fascinated that I would actually know all this stuff. And I'm like, well I worked on an airfield, I worked with beam forming, I worked with, you know, not, not on a wifi router but on a localizer on a, on a airfield system. I worked on beam forming and stuff like that. So like, you know, this, this is not, and so I really think that we underrate or we undervalue the broader knowledge, the broader background knowledge of how things work in our world. And, and I find that very sad myself. People are very unidimensional now and surfacey. That's a general like broad philosophical statement I just made there. But anyway, I don't know. So maybe Yvonne has a . Speaker 4 00:38:26 No, I mean I think like, I, I remember being a young technologist and looking at people and thinking how do they know so much? Speaker 2 00:38:33 Mm-hmm. Speaker 4 00:38:35 Again, some of that is just the march of time, right? Yeah. It's um, you know, and, and Nick Russo says he's quoting me when he says this. I feel like I heard it from him. So I don't know, it was probably some other third party who's the original source of it. But like you can, you can, you can have 10 years of experience or you can have one year of experience 10 times. And so if, if in your career you are consistently having that, you know, every year is a new set of experiences. Eventually you, you just learn stuff, right? And then you're able to, to knit concepts together in ways that new people can't cuz they just don't have the historical background. And and that's what you're talking about here, Russ, Speaker 2 00:39:19 With building a mental map. That's, that's in the philosophical world we call it build a mental map, right? But I think, I worry that in a world where nothing is hard, you buy the computer from the local store, you plug it in, it just works and you connect to the internet and all of your services are out there and your company doesn't have anybody in it on staff because why bother And nothing ever breaks that. Or when it breaks you throw it away that we are always pushing people in a position of always of, of the 10 years of one year experience. 10 times. Like we're not giving them the opportunity to like build the knowledge that they need to be excellent engineers. And, and maybe we just don't need excellent engineers anymore. Maybe we just don't. But Speaker 4 00:40:12 I think we always will. And I think I'd, I have very little confidence on humanity's ability to eliminate hard problems. Like I just don't think we're ever gonna get there. Right? There are always gonna be hard problems to solve Speaker 2 00:40:24 , but I, but I have, I have a very large confidence and mankind's ability to believe they have solved all hard problems and therefore eliminate the training and experience needed to have people who can solve the hard problems so that when the hard problems happen, nobody knows how to fix it. Speaker 4 00:40:43 Enter the COBOL program. Speaker 2 00:40:45 . Exactly. . That's exactly what happened. Yeah. That's, Speaker 3 00:40:50 That's, I don't know, I don't know if, I don't know if you guys know, I wonder if there's a, sometimes you can find trends, you know, earlier in history. Like there are things that we use today that have been around for much longer than networks have been around for, you know, machines and, and things like that. The required engineering. Are there any, can you guys think, I can't think of any. Can you think of any examples of um, I don't know, like the, like the car maybe or I don't know, something that has been around longer than networks that people figured out how to optimize and still know how to fix that We could Speaker 4 00:41:24 Throw lessons for the thing with car, like cars have gotten a lot harder in the last couple decades. Um, yeah. You know, like for a long time you could just take the darn thing apart and figure it out. You can't do that anymore. Speaker 2 00:41:35 Which I, which I've done a few times in my life. Yeah, yeah, yeah. So Speaker 3 00:41:39 Yeah. But we still have 'em, they haven't disappeared and people still have uh, expertise in engineering ability around cars. So I don't know. It's not all doom and gloom. ? Speaker 2 00:41:47 No, I don't know. Maybe it is . It's more fun if it's all doom and Glen, don't don't, don't mess with our pity party here. Speaker 4 00:41:55 Tom. I'll leave your rest. Speaker 3 00:41:57 . Speaker 4 00:41:58 Optim. I enjoy optimism. Sorry, . Okay Speaker 3 00:42:00 Good. I was gonna say apparently I'm the optimism here on the hedge . Speaker 2 00:42:05 No. Okay, well I think we've hit all this and we probably need to wrap up cuz I know Yvonne needs to run from the she shed and go do something. Yep. Plant flowers or something. I'm sure. Speaker 4 00:42:18 Ah, husband's traveling this week, so I'm pulling double duty. Okay. Yep. Speaker 2 00:42:23 All right. Well, okay, Yvonne, where can people get in touch with you if they want to? Speaker 4 00:42:28 Yeah, still on Twitter at Sharp Network or you can reach out on LinkedIn. You can find me there, um, at Yvonne Sharp. So, Speaker 2 00:42:35 All right. And Tom, Speaker 3 00:42:37 I have something shocking for you today, Russ. You can actually, you read a blog this you can refine something I wrote. It didn't quite rise to the standard of a blog. Well actually it kind of did. Um, Ivan, Ivan Pep and I had a conversation and he took a big chunk of what I wrote and stuck it on his blog, so. Oh, Speaker 2 00:42:53 That's Speaker 4 00:42:53 Really cool. Hey, that you've arrived, man. Yeah, that's good. You just skipped a whole phase. Speaker 2 00:42:59 That's Speaker 3 00:42:59 Right. That's good. I was, I was, yeah, I felt really good about that. I, I'm glad that he liked it, so yeah, Speaker 2 00:43:04 That's excellent. Yeah, Speaker 3 00:43:05 Go to go to Ivan Pepper's blog. You can find something I wrote, but, but I'm otherwise IP space.net. blog@ipspace.net. Yep. And uh, yeah, I'm on uh, Twitter and LinkedIn. Just search for Tom Amon. Speaker 2 00:43:16 Okay. I, Ms. White, you can always find me here at the hedge. I write regularly at Packet Pushers. Now I think you, Yvonne does as well. So you can go find Yvonne on packet Bushers and I write here and there other places, who knows. But anyway, um, I do try to point to all that from Rule 11. And you can find me at Rule 11. You can find me on LinkedIn, you can find me on Twitter, you can find me here at the hedge. I don't know where else, but that's good. It's enough. All right. Well thanks everybody for paying attention to our ramblings for 45 minutes as we wander around the networking world and, and talk about stuff. We appreciate your attention. And thank you for joining us for this episode of The Hagen. We will catch you next time.