Well, today it's just you and I again, Ben. I think our friend Ed said that he was hurtling through the French countryside on a highway of some sort. So I assume that's good news. I don't know, but it sounds like it would be fun. Yeah, drive safer, Ed, wherever you are. That's right. And so I thought there hasn't been a tremendous amount of the type of news we like to comment on. But Ed, aforementioned hurtling through the countryside, he came across two things that that he sent over that I think there's there's some good commentary to be had on that. One of them is a report from from GigaOM from, you know, from Justin, who if you if you follow his stuff, he's Justin. You know, I don't actually know if it's Warren or Warren. How would you pronounce W-A-R-R-E-N? A Warren, I guess. Yeah. OK. Yeah. I just like, you know, I'm I'm aware that I'm I'm and I don't mean this to be self-aggrandizing one of those people who. Well, this may not be true anymore after all these years of podcasting, but one of these people who more or less reads a lot more than I talk. And so like every now and then I'll encounter something and I'm like, I know how I pronounce that in my head, but I don't actually know how it's it's pronounced. And you might have encountered these people a lot. Tech people are a lot like that where they. Yeah. What's going on with your name? Because you've got the little tick over your E and I don't know what that means. Am I supposed to do something with that? It's just it just means instead of saying coat, you say Coté. It means that's that's how I was brought up. And I'm pretty sure that's another failing of mine, as you know, at some point in learning, I assume other ones do this. And I'm pretty sure y'all over there in the UK would do this. But when you learn English at some point, they give you this they give you this lesson where you've got all these upside down letters and weird things and diacriticals. And they're trying to teach you like how to how to pronounce things and how to describe them.
2:01And I think I just must have been hit in the head at the end of that because I don't remember any of it. Like, I think I was too busy chewing the crayons, to be honest, to. Yeah, that's that's probably what was happening with me as well. Anyways. But there's a restaurant near Old Street. Yes. The Old Street office is called Coté. Exactly. And it might have the same thing. I don't remember. It does. But it has a little hat on the O, which, again, I don't know what that is or what that implies about what you're supposed to say. But funny you should mention that I was in a bunch of meetings earlier this week and there was someone who she said her maiden name was Coté. And so when she saw my name there, she was kind of shocked. Also, she was in Austin, which is of all things, you know, where I'm from. The Austin Coté's. Yeah. And she told me she, you know, so we were talking about it. And I said, yeah, there's also a there's also like a chain over in the UK that's called Coté something. And like what I've learned over the years in Twitter is that people love going there for brunch. Like, you know, usually sometimes on the weekend, I'll get I'll get like a I'll get a tweet because they'll say at Coté and it'll be like, oh, lovely meal, you know, with with Josie. And the fellows delicious mimosas or something like that. And I usually write back and I'm like, oh, I'm glad you enjoyed it. And, you know, please come again. I think I think I think one time someone was complaining and I was like, oh, I'm sorry to hear that. But it's not me. Anyways, it's just opposite the the British Library. And yeah, there's there's no not British Library. Name escapes me now, but there's a there's a there's a there's a there's a Banksy like literally away from that place.
4:02Yeah. Well, it's definitely Barbican, Barbican Theatre and Barbican Library. Oh, yeah. Well, there's definitely a lot of them. I see them very frequently when I when I'm over there. Anyways, now that I've made a big to do about it. Justin Warren. He's got a delightful you know, you would like his he's got a consultancy. You should go check it out. Pivot nine. And if you sign up for his newsletter. So not only does he apparently write up analysis is analyses of reports on the Kubernetes landscape. He also has like kind of a consultancy helping people with their their tech marketing. So you can go and read his it's a it's a very common thing with boutique analyst firms or whatever boutique is always such a funny word, right? Like it just means small, but he has a lot to say about how one would do tech marketing in his newsletter, which which is fun. Anyhow, I'm signing up as we speak. There we go. Look out. There's a record. I expect him to recommend my newsletter now, which is available at newsletter.cote.io. I've got to drive some more subs, right? Subs. Yep. Click click and subscribe like. Anyways. So so he put out a Kubernetes report and now, you know, there's actually like he's got a very lengthy ranking of all of the the vendors, as I would say. And the word choice there is the point that I wanted to bring up. And also, I guess I guess the the giga ohm, they don't have a magic quadrant nor a wave, but they have a radar report, which is a little bit like the I don't know it that well. I must admit it's a little bit like the ThoughtWorks radar, which means it's circular. And it kind of it has, let's see, one, two, three, four, five, six, seven, eight, nine, 10, 11, 12, 13 different.
6:02And again, I'll get to the word Kubernetes vendors you can choose from. But I think I think what's what's you can go read it. You can find it. I think I think one of the vendors has licensed it. You can check it out. And it has a good summary of each of them. But the thing I wanted to bring up is that he called this it's it's the radar for managed Kubernetes. Right. And so I'm not going to be Socrates here and ask you questions. I'll just what's the opposite of Socrates, like just a lecture and football tease. I don't know. Hey, wow. That's you're not going to you're not going to come across that kind of wit in the Agora. They would they would run you out and throw empty olive oil bottles at you. Come on. That was very Ted Lasso. I'm on I'm on trend. All right. Oh, no, no. I just meant that those Agora acts, they I don't think they were into to to clever puns and rhetoricians. They were very against rhetoric and wordplay. They in fact, wasn't it wasn't it Socrates who was like, we need to kill all the poets. He was very anti poet, if I remember. And then I think I think Plato Plato later came in and is like, let's let's kind of chill it with being against the poets. Anyway, there's your there's your Coté's philosophy corner for this episode. I love Plato almost as much as crayons. There you go again. What a roll. So I when I see the word managed, I think that that must mean like public cloud, right? Like a managed service, which, you know, someone else running it. But because he includes, you know, OpenShift and Platform9 and our very own VMware Tanzu and talks about hybrid and multi cloud stuff. It was making me, you know, and I'm pretty sure he means like whether you run it on your own or not. And I think I tried to go back and reread this. And the point I'm trying to make here will be moot if I'm wrong, which is the best kind of point to try to make.
8:04It won't really matter if I'm wrong. But like, I think what he's saying is just like, you know, not building your own Kubernetes, just like if you were to get a distro or run it in a public cloud, which I think is. That is like a category that needs to exist. Right? Like, and again, the category is I am not building it my own on my own. I might run it on my own, but I'm not in charge of assembling it together. Right? Because, because the distinguish between a running Kubernetes, using a Kubernetes service, one of your KSs in the big public clouds versus like running a Kubernetes on your own, on your own infrastructure, like those are, those are also two very different things. And like comparing the two of them, like drags on all sorts of other conversations you need to have. So kind of contradicting what I was just saying, like, it's almost like I'm, I haven't really thought this out as you can tell, but here, here's an analogy and maybe you can help me out here, Ben. You remember three, four years ago when we went through one of those phases in the IT world where we were all obsessed with the developer, everyone was talking about how important developers were. And this was right around also when, when Kubernetes was like, you're like, Ooh, that horse kind of pulled the head over there. What's going on? And, and like the, the clever thing to do that you would do if you were Socrates was to be like, ah, but which type of developer? And then, and then you're like, oh no, we're in, we're in the soup now. Like now we're, we're, we're getting off and distracted. The point there being that like, people would say like, oh, you know, there's not just one type of developer. There's what, and, and to cut to the chase, most of the time, I think, and this is still the case when people say developer, most of the time, what most people mean is application developer, right? A developer who is, you know, uh, obviously screen readers and other things aside, uh,
10:08not aside, but like considered if you'll pardon the joke developers who move pixels on the screen, right? Like that's, I think what most people have in mind versus like, you know, embedded developers who work on like, when you press the accelerator in your car, your car goes and, you know, systems developers who work on chips and then our favorite platform developers, blah, blah, blah. Right. And so it also seems like, that's why I started thinking about, you know, Justin's like lumping together of all these Kubernetes things is like, I think it's increasingly important when people talk about Kubernetes to be like, ah, but which type of Kubernetes? Right. Because like, because like all of the, um, a lot of like the, the benefits and the problems that you're going to have with running your own Kubernetes, running Kubernetes on your own, let me phrase that are going to be different than if you're just using a Kubernetes that is run for you in the public cloud. Right. And they're, they're kind of like, there's not, there's increasingly not so much nuance between them. Like, like, it's pretty obvious that they're very different things. So like, for example, you know, if you're running one of the KSs, any KS will do in, in, in one of the three clouds or whatever, in a public cloud, like it's a lot easier. Like there's just like, there's like a whole lot of, I mean, it's public cloud. So there's just like a whole lot of stuff you don't have to do. Like you don't need to have, yeah. Yeah. And not only with installing and upgrading Kubernetes, but also like, you don't need to schedule a meeting with the networking team and you don't really need to like, you know, I was talking about this with one of our coworkers. There's stuff you need to do. It doesn't go away, but it's a lot different than like, you know, talking to the networking team in your own data center for your own on-premises stuff. Right. Like, and, and definitely, you know, there's even like another, uh, contemporaneous topic
12:08that comes up, I think in the, the, the public cloud Kubernetes things where like, it's a lot easier to get runaway costs in that situation. They're not in on-premises, right? Like, because of just the nature of how public cloud works, like, you know, I think it's a little cynical to say they're happy to let you overspend, but it's more just like, yeah, that's just the way it is. If you want limitless consumption possibility as a feature, that's exactly what you'll get. Limitless billing. Yeah. Yeah. Whereas if you're running it on your own, eventually you'll just hit the physical limitation of your computers, right? Like you won't. Yeah. Although you'll burn money in different ways, won't you? Sure. I think a lot of folks underestimate the overheads involved in creating a good, you know, creating a good basic platform, you know? Right. No, totally. I mean, I think it's, it's beyond like CapEx, OpEx stuff that we were so enamored of in like 2009. Uh, but like, no, I agree with you completely. Like, I think, I think maybe there are definitely wasted money in runaway costs. It just like happens a lot slower and a lot more like, not unexpectedly, but I think the point you're making, I cut you off is that like, you'll end up spending a lot more money on like people and like, you know, dealing with it than if you automate that stuff. But, um, you know, so you can end up, let me see if I can phrase this. You can end up being forced to spend more money than you want because you've made this decision and now you need to staff it. And now you need to take care of it, even though you didn't anticipate those. Whereas in the public cloud, you can end up spending a lot. You can end up accidentally spending a lot more money than you wanted to because you weren't, you didn't even know that would happen. Or you end up accepting a lower quality outcome, you know, or whereas in the cloud, you might
14:09have a extremely high quality outcome, but you're not using all the features. So, so yeah, yeah. Swings. Yeah. And then, and then also just to, just to be, uh, just to be clear about it. Right. Like, I mean, I think all these things we're talking about are, are risks, not certainties. Right. Like, like it's, I, you know, I've, I've, I've encountered a lot of like, you know, marketing, like you should like our stuff versus someone else's stuff recently. And like, I realized that's something that like, is people always forget about as like, yeah, I mean, everything we're talking about has like a one to 5% chance of happening. So what we're talking about is just, you've got a different set of risks that you need to manage. Anyhow, I've a bit of lost track of what, what we were talking about here, but that's, that's probably a good sign. But, uh, yeah, so like, I, I, I was, I'm really like fascinated in this, like now new grouping of types of Kubernetes is Kubernetes, whatever it may be, you'll have to see how Greek is pluralized. Uh, but like, uh, because I, I think, I think, I think increasingly we need that in people who are in the kind of business where, and we need that kind of distinction because like, uh, I think a lot of the discussion about, for example, platform engineering, like it's very unclear if a platform engineer is in charge of building, running and maintaining Kubernetes, or if they're just in charge of kind of configuring it and run and, and running their platform on top of it, right. Which is very different things. And like, you know, hopefully it's not the first because it's completely unrealistic that you could both own Kubernetes and own the platform. Yeah. I think, I think they're very different problems. Yeah. The, the, the, the platform engineering for me is, is all about, you know, the customer is the developer and, uh, you're helping them to, by creating automations that, that fit
16:17into the way that they work and, and then how they work and, and save time in the way that they work. But that's a very different problem space to, can I run and manage Kubernetes, upgrade Kubernetes, um, make sure it's, uh, you know, integrated well and make sure we're capturing all the information that's coming in about Kubernetes all the time, all the changes, all the, you know, the, the churn that happens in that particular space. That's, that's a whole different set of problems for a whole different character role in my opinion anyway. Yeah. And there's no way that like a team of people would have enough time in their lives to handle that, like they would, they would burn out really quickly and do a terrible job at all of it. Yeah. Or, or, or, you know, you need to be able to look around yourself and say, actually, this is, this is well staffed and it's well, it's well funded because if it's not, then the consequences are in the quality side, aren't they? And you won't achieve what you want to achieve. So totally right. And, and, and so, so like, that's an example. So platform engineering is an example of where you need this distinction of what type of Kubernetes you're talking about. Right. Cause like the, the kind of shallow discussion of platform engineering, but like, you know, even, even I have had, even I, as if I'm some sort of like saintly person who avoids like a wrong talk all the time. I think it's Coté of Texas. Yeah. Let me, let me, let me, let me go polish off my rhetoricians halo that I've got in my back pocket here. Uh, but, uh, like, you know, I would assume that most of the people talking about platform engineering are not, are probably running on the public KSs, right? Like, like if they're running something on Kubernetes and they're talking about it, improve Kubernetes with platform engineering, which includes, you know, like getting backstage set up and dealing with not dealing with, but getting build pipelines or supply chains, if you, if
18:21you prefer, uh, and like all that, all that middleware and developer tooling, like I, I don't, it would be shocking if they are also not only selecting a Kubernetes distro to get up and running and talking with all your networking people to make sure that works and then maintaining it and upgrading it and also making sure that it runs and they set all the policies, but even worse, if they're DIY building it themselves, like that would just be bonkers. And then that, and then that further implies that like, if I, I guess this is inferring, this is what I've learned from chat GPT is you, you, you say like, I'm just, I'm just a human or a large language model. I don't know, but I can infer. And then you're off to the races. It's great. Just infer. So what you can do is you can just tell it straight away in fur and, uh, isn't, isn't that what captain Kirk would always do is he would say, speculate him, him and Kirk, you'd have to tell, you'd have to tell the computer in the Vulcan, him and Picard. Oh, really? Oh, I haven't seen this for a long time. So I don't know. I guess Picard would have to tell data. He would, he would have to command, command them to like, you know, don't give me these disclaimers. Just speculate. Uh, anyways, that's, that's, uh, there's an equivalent in since you're such a big star Wars fan, remember when, uh, Han Solo keeps saying he doesn't like numbers, that's the equivalent of saying speculate to C3PO. Never tell me the odds. That's right. Anyhow. So, so that would imply that if you're doing like on premises cloud native stuff, right? It's highly likely that you have a whole separate team of people or roles and responsibilities that manage your Kubernetes stuff, because it would just be like insanely impossible otherwise, or, or something else. Let's crowdsource this listeners. You tell us, get in touch, tell us what you think. Are you doing it? How do you do it? Do you, do you have lots of sort of your own data center with a whole bunch of Kubernetes
20:22in it and you still have platform engineers and are still building a platform in a particular way, or is there a survey out there that you've seen? Have you seen a survey that already had some data on this that told us, you know, what the split might be between engineering, public cloud platform, engineering, private cloud, data center cloud. Yeah. I should start asking that in talks that I give. I have, I have a talk coming up at platform con, you know, the online thing, which probably means there'll be a big audience. And I won't remember to do this, but that would be a fun poll to ask people while I'm giving it is like where, where, which KS they have. Do they have my KS or, or, or, uh, one of the, the GAA KSs. I guess Amazon is EKS, right? So you don't have two A's. Yeah. GKE is a KE, not a KS, but yeah. Too many K's. Too many K's. They've all got K's. Highly related. The other one, I just wanted to go on briefly. And you can, I think you can also, uh, I think humanitech licensed this one. So you can go download it if you want. They'll be in the show notes, right? Exactly. Over at tanzutalk.com. But they, uh, I think, I think, uh, they commissioned a Forrester survey. Uh, and it was basically a bunch of questions about platform engineering tools and initiatives. And, uh, you know, just, uh, I think, I think there's, there's two things I wanted to high, well, I guess three things like the first, the first, uh, comment on it is that if, if you, uh, this is an example of, uh, when you encounter charts like this, you have to like put in some extra effort to find value in them. And I don't mean that to be like throwing shade or some offhanded thing, but more of an instructional thing. Cause when I looked at it at first, I was like, Hmm, all of these answers are at, are either 38% or 48%. They're all in this range. So like, if I, if, if I were to say like, I surveyed a hundred people about what their favorite type of sandwiches and like, you know, you had, you had, uh, five types of sandwiches. I can't do math that fast. And, and, and therefore like, you know, you had five types of sandwiches and each sandwich
22:24was rated somewhere between like 25 and 35% of people said they liked it. You would be like, well, I guess people like everything. Like, there's no like obvious winner. There's no, like, you know, you don't have the, uh, you don't have like 80% for pastrami for pastrami on rye, like with melted, uh, Swiss cheese and, and, and, and horser, you know, sauerkraut on top, which is clearly the best sandwich there is. Uh, maybe with some, for what outcome? Early death. This is, this is the, uh, this is the, uh, the, the, uh, the hedonism sandwich index. I think it's, I think it depends which outcome you're measuring anyway. So, so this, this, this is kind of like everyone likes, you know, there, there is no outstanding sandwich type of chart if you look at it, but what's instructive to look at is there's one question in particular. That's like, uh, what benefits have you achieved by like, uh, addressing and improving developer experience, which like developer experience is a fun word, but I just rewrite it in my head as like productivity, right? Like just better tools for developers so they can do a better job. And there's a certain amount to it of just like also developers being happy and having less cognitive load and like, you know, enjoying life more awesome. Right. But like productivity is a little more like a tangible to me. Um, and so you look at that chart and you've got like, there is no outstanding sandwich chart. However, that's a misreading of the question, which is the way when you, when I look at that chart, what I realized is that, oh, when you improve developer productivity or developer experience, everything gets better, right? So like, there's not only one thing that gets better, but all of these different things have a high chance of improving and getting better, which is great. Like that's, that's a great thing you want to know about developer experience because investing in a developer experience is, I always mix this up, is a we would call it a non-functional requirement, which is it doesn't move pixels on the screen
24:28or introduce new features. So it has no, well, look at all these fancy terms I'm trying to use. It has no first order business value, right? Like, so like if, so for example, if we had a, if the team of developers says in this, this, uh, this one to four week iteration, we're going to rewrite how we do our builds. And then the product manager and the business is going to say, so, um, what new feature are you going to release? And the developer would say nothing. And then the business person would say, hmm, how about you keep using the same build and add that new feature? I need to make more money. So, however, you know, so you get in these kinds of discussions, whereas like what you have with a result like this is you show like, no, no, look, it actually does increase customer experience and business profitability and all these other things. Right. In the same way, like the analogy that I use now that I have in my head nowadays is like, you know, if you were to go to a manufacturer and if you were to say, Hey, business wise, if we made your factory run better, you would get that that makes you more money. Right. And they would say, yes, because I'm not insane. I know how manufacturing business works. Whereas like, I don't think we quite have that ingrained in us so much in the, uh, the software world, we don't have like an intuitive, no doubt understanding that if you improve how the tools and the process that developers use, that it will result in a, uh, a better business. Cause cause also that's not always the case as well, since we're not a factory. I know I like in this report, how the, the, the, the using and reusing this phrase, there's coming up more and more the idea of an internal developer platform, you know, so it's not just about the platform engineering, but it's also about the people who benefit from it. Yeah. I think, you know, I think that's good. Yeah. You, you've actually, you've actually fixed a, uh, a word choice problem. Let's see. Do they actually say EDP, E IDP is an internal developer platform or did they, did they say
26:30portal? No platform. Yeah. Oh, good. Because, because originally IDP is like a portal internal developer portal. And no, no, they're going, I think, I think they mean, yeah, I think they mean it as a platform. So, you know, not just covering the sort of, you know, the gooey sort of aspects of, can I see what's going on around me, but more. Yeah. Yeah. But also in addition to that, the supply. And so that's what I like about that is, is as, as I'm, I'm always very open and sharing, I think internal developer portal is one of the worst names we've had in a long time. Like just because like the word portal, I don't, I, I, maybe I haven't, I haven't done a survey with the, uh, generation Z or whatever, but I feel like no developer responds well to the word portal. Like I, I, I, I put the blame for that firmly at, um, web, web logic, web logic portal. You and I can probably both remember that. Yeah. But, but, but I think if, if, if you go through and you make the P stand for platform, now you've got something, right? Like, I think that, that, in fact, that's, I think that's really, that's really good. That's like a five minute, like little lightning talk there, Ben, that you've come up with. It's like, I like to use the bouncy castle analogy, you know, the portal is just your way into the bouncy castle, but the platform is the whole thing, right? It's the thing that gives you the fun, right? Yeah, that's that. Sure. I love it. That's, that's, uh, that's a, that's a minute three of the five minute talk. I love it. Yeah. So yeah. Uh, developer platforms are bouncy castles and pick it to ride. And then anything, uh, you know, listeners can go, uh, go look up the report if you like it. We'll, we'll throw some free lead gen over to humanitech there. See, see, see how they're doing. But, uh, also, uh, you know, there's also a finding that like 51%. So let's just say here, here's how you can tell if someone's like an advertising or not. Should I say half or over half?
28:30Uh, but 51% of the people, uh, are, are planning on investing more money in, uh, in, uh, you know, developer, you know, uh, productivity experience. So, so there's people who actually want to spend money. In time to improve it. Although it doesn't specify if invest means money or time, but whatever, you know, uh, yeah, both, both probably. Yeah. Yeah. I think, you know, and it's, this, this, you know, they clarify this, this point around platforms as well, which is, which is one key point that I think differentiates it from a portal. Right. So, so they say here, uh, in the report, um, reducing the bottlenecks caused by ticket ops, right? You don't really get that from a portal. Yeah. You can only get that from self-service through having services come to you, be able to just commission them yourself as a developer, not having to raise tickets for, you know, firewalls and all sorts of other things, databases, all that kind of stuff. So it is different. I think, uh, I like that. I like this report. It's good. Yeah. Yeah. So, uh, people should check that out. That, that'll be fun. Yeah. But that's what we have this week. Uh, you know, that's all. Well, in that case, I'm a bombshell. I think it's time to end. Right. Well, you know, thanks for listening. And if you like this show, it'd be great to, uh, you know, to share it with people. Just, just leave us some comments or, or whatever, uh, you may like. Uh, and, uh, you can also, you know, find links to those two things. If you go to tonsutalk.com and we'll see everyone next time. Bye-bye.