The intersection of kubernetes, Edge, AI, and tuna pizzas, with Derrick Harris

Episode 125 · 55:46 · SRT · archive.org · all Tanzu Talk

Well, I think I might have a tapeworm or something because over the past few days, and I want to ask your advice on this, Richard and guest, like I have just had this insatiable appetite. Like this morning, I had an extremely large breakfast. I had a two egg omelet with a sliced up bratwurst in it, and then two pieces of toast with butter and strawberry jam, and then also a small orange. But I feel like I feel like that's an inordinate. Usually I have like 50 to 100 grams of steel cut oats. That's all I eat. So I don't know what's going on. And then I had a lunch that was like some smoked fish. And then I had a bunch of chips. And then I had I don't even remember what it was, but it was just intense. And then and then yesterday for breakfast, I had half of a leftover pizza. So I mean, I know I'm not pregnant, so we can eliminate that. But I feel like there's something going on that I need some help with. This isn't sustainable. No, no, I mean, I haven't seen you physically in a little while, but hopefully I can still hug you when I see you in a few months. Yeah, it could be a growth spurt. Yeah, I think you're fun at that age. That's what I was thinking. You're gonna be like seven to the next time I see you. Uh huh. Uh huh. Yeah. I mean, that's not unknown to happen, maybe in the entire history of humans for someone in their early 40s to have a growth spurt. I guess it could happen. It could be that you when you move to Europe, your body was like subconsciously wanted to grow larger than to meet the northern European. I think you're right. I think you're right. I'm trying to fit in with the the the descendants of the barbarians. I've got to make sure I can I can do it. That's that's that's uh, okay. I'm I like that theory because it's positive. So I'm gonna order a couple pizzas for tonight and tomorrow morning. Yeah, you're you're not leaving anything out, right? You're not like running marathons for fun now or just, you know, no, there's reasons to be that hungry. No, yeah, yeah, I yeah, I don't know. I can't figure it out now. So this reminds me of something I was realizing recently speaking. I was I was eating some pizza. I spoke at a meetup a couple nights ago.

First of all, they gave me a bottle of champagne as a speaker gift, which, you know, all meetups should aspire to be so polite. But as with all meetups, they had they had all sorts of pizza and I and I spotted this one pizza that I was getting a slice of and it was like it was like mushrooms with like some shredded pork or maybe beef on top or something and I had some other ones. And so I got around to eating the shredded beef one and it was the strangest tasted shredded beef that I've ever had. And then I realized it was tuna because they love putting tuna on pizza over here. Now, in fact, our babysitter, that's her favorite. You get tuna pizza, which I mean, you know, I don't want to tell people how to live their life, but it made me realize that I've always had slight issue with with European pizzas like they're a little different. And I think that the issue is that Europeans, they think of pizza as a sandwich. They don't think of it as a distinct thing, because the other thing you'll notice is there's basically lunch meat on pizzas over here. At best, you'll get like prosciutto, but they don't have like pepperoni, you know. So I think this is my new operating theory is is in Europe, a pizza is really just a sandwich that's been flattened out. Man, that is a definition of a hot take right there. I mean, I thought you were going to go further on the hey, I thought I was eating beef and I realized it was tuna sort of switcheroo, which I would have compared to like eating a muffin thinking it was chocolate chip and finding raisins. But no, you took it as a sandwich route, which was, you know, it's it's it's I I because I feel I am the one responsible for having eaten a tuna pizza because at this point I should know better. Like I should have spotted that and immediately been like, nope, I don't like that. That's tuna pizza. But I was I had my guard down. Did you not think it was weird that there was shredded beef on a pizza to start with? Well, I mean, see, exactly. Exactly. That was my failing is I should have. I mean, I think I was a little excited, but I think I think maybe I had that, you know, kind of pre speaking, like not really thinking about things straight, trying to trying to like socialize with people there or find a corner to hide out.

And I just I just wasn't on my pizza game. It was it was upsetting. But, you know, speaking of American pizza, where where was the Microsoft Build conference? Was that where was that located? It was here in Seattle, which I was also having a team off site in Seattle this week. So there was a lot of things happening here. Oh, very nice. It wasn't in Orlando or in Las Vegas. That's great. They had they had it all there. Think of the savings they had of all the employees not having to travel anywhere. That's good stuff. Yeah. Well, there was there was there was a lot of announcements going on there. And as always, you being a Microsoft MVP and having recently written a book on modernizing dot net applications, I'm sure you are the best informed to to go over. But it seems like my summary of the announcements, I think that that we and listeners would care about. There's there's a lot of Kubernetes support. So lots of things working around in there. There's some part more partnerships with with other people as far as things you can run on Azure. And then my broad, open ended question for for for you was like reading through all the SQL Server related announcements. It made me and, you know, this sounds slightly purposely absurd, but it made me realize that like SQL Server is kind of a big deal. Like, am I wrong? Like it gets used everywhere. Right. Like it's but people don't seem to talk about it very much. Like, I don't think I mean, maybe it's just like my I think you call it a selection bias, but it's not like over the past 20 years, SQL Server has been like up in my face all the time, like my SQL or whatever a Kafka is or all these other things. No, I think you're right. I mean, it's a really bizarre phenomenon because you're right. Sometimes I'll see it. I tweeted something yesterday about a new SQL Server library that's getting consolidated between .NET and .NET Core thinking like, yeah, everybody's using this. I forgot about that. I just looked at the DB ranking and like Microsoft SQL Server still number three, like Oracle number one, my SQL number two, SQL Server number three. Then you have Postgres and Mongo and fun stuff like that. So it's everywhere.

But you're right. I just it doesn't feel like we're talking about it very much. And maybe just because it's simply boring, established tech that does evolve, but not at the same rate as the other crazy tech we all use nowadays. So, yeah, I know it's one of these most used, under talked about things. Yeah. And do they do they still like build a lot of things into it or is it kind of just like a relational database still? Like what's the what's the brand umbrella for SQL Server now? No, it's still been beefed up to, you know, as usual, all the analytics, all the data movement, you know, storing different types of data now. They're, you know, JSON stuff for JavaScript and it runs on Linux now. It's one of the more popular container images that you can run on Linux, at least for developers, you know, millions plus pulls from the Docker Hub. So people are running it on Linux. You're running it on Windows. You're, you know, running it in the cloud on Azure SQL. So it's the brand is now kind of cloud on prem, you know, big stuff, little stuff. So you've done a good job with it. Maybe maybe you and I are just missing everyone else talking about it. Could be. Yeah. See, like, for example, just just right there, like everyone's always excited about like, you know, I don't know, Windows running on Linux or Linux running on Windows or whatever. But like SQL Server runs in a Docker container. That seems like something you would name check all the time as far as like, you know, the world is topsy-turvy for us olds. But but not. But not. There you go. Well, what else? What else caught y'all's fancy from from Microsoft build? Yeah, right. I mean, you know, a bunch of serverless announcements, new improvements to Azure functions and stateful functions and stuff at the edge and some new developer tooling was nice. And I think it's a private beta, but an online version of their Visual Studio Code sort of editor. So, you know, Visual Studio Online sort of stuff. So, you know, as usual, it's an absolute grab bag of things from cloud stuff to AI stuff, some new machine learning libraries shipped and things like that. Every service seems to try to get their announcement in there. It's it seems that we we're all doing conference driven development as an industry now. So everyone anchors their big announcements for conferences. But yeah, no, it was it was a great grab bag of stuff.

I don't know if anything stood out to me. It's like here was the one crazy thing they announced, Derek. I don't know if anything stood out to you in that respect. No, I tried to I tried to follow up, follow some of the news and I found it nigh impossible to keep up. So I'm glad I'm glad it's not my job anymore. But yeah, I mean, there was a lot of stuff and nothing. Yeah, nothing. Nothing was mind blowing. Right. Nothing was. Wow. This is this is like this is a game changer. It was kind of like, yeah, just a lot of improvements, a lot of improvements. They did announce a partnership with Starbucks, which I thought was cool just because predictive drive through ordering seems to be the the what's my my words escape me sometimes. But that's the that's the that's the that's the thing of the day in restaurants or fast, fast service restaurants. They say McDonald's is doing it. Just bought a startup and now Starbucks wants to do it. So, you know, that was a I guess Microsoft's going to help him with that. So that was, you know, what is this predictive drive through? Is that is that for Starbucks or for me, the customer to figure out when I should go there? Yeah. Yeah. Yeah. So apparently when you drive up, it will somehow be able to know who that you're driving up and get your order ready or suggest or suggest on a digital screen, like the handful of things that thinks you might want to order. That's I'm guessing. I'm guessing if you have the Starbucks app, it would clock that you are in the vicinity. Yeah. Yeah. That would be that would be some exciting machine learning stories for the next three years. Right. Like like it'd be a good basis of like I tested this out and here's what Starbucks that I wanted. And then and then, you know, you could try to, you know, fool it by training it and ordering the wrong things and absurd stuff. Or I know, you know, if you're playing a prank, you could like borrow someone's car and dress up like them and start ordering like, you know, double large peppermint frappes or something. And just see if you do that five times when they drive up, they'll be forced to get one of those. That's right. I don't know how it works, but I don't think they're scanning license plates.

That might be a privacy bridge too far at the moment. Yeah. But just keep going up there ordering tuna pizzas and then that one time you actually show up. Stupid. Why are they giving me tuna pizza? That's also, you know, there's been a lot of I haven't really studied this very followed up on it or whatever. But, you know, there's all these like ethics boards around AI stuff. But there's sort of an odd technical complication of like the simplest thing for what you're saying would be to just scan your license plate and be like this license plate wanted that thing. But that culturally bad idea. So you've got to like hack your way into all these other ways of identifying someone. Yeah. Well, you know, my car wash did that without without asking me as far as I can. But I do get free car washes every 10th watch. So I'm always surprised. Yeah. I hope that's in some subsection of GDPR that I can finally agree to is like if if if requester is being given a free car wash, you do not have to ask them to click an accept button. And that would be nice. You're right, though. It's funny how we did that with with data was like, well, we're going to. Yeah, this is this is the whole story of big data to begin with. Right. Companies will just ask you for what they wanted. They would have to get it. They would have to identify you in all these other ways. So it's collected everything you did to build a profile like just ask. Yeah, maybe maybe probably they'll tell you. Yeah. You know, I mean, coincidentally, so there was also a one of the build things was that there's I think I think it's either you can run or integrate with or whatever, like some Kubernetes stuff on an Azure edge thing, which I assume is just like a small device somewhere. But it reminded me of last year, the the Chick-fil-A people were talking about how they do their in-store mini Kubernetes clusters. And I went back to go look at their engineering blog, as it were, and I found a post that I'd missed from, I don't know, sometime in the fall. And it goes over. I think it's machine learning in so far as just image recognition. It's fascinating to read it. It's kind of like it's kind of like what you always hope one of those Raspberry Pi write ups is, like something that's actually useful, I guess.

And it's this whole overview of how the problem they were solving is like you set these timers, not only for how long you cook something, but how long, you know, like chicken can be left out. I mean, I guess obviously chicken, chicken can be left out and like setting the timer is cumbersome. So they they it's kind of like remember that interview, that podcast we had a long time ago about a food services company that was optimizing all the software in the kitchen. It's a similar thing where they they switched over their timer to a touchscreen and then and then they found out another problem was like it would be nice to automatically time how long the chickens have been out. And anyways, it involves like a camera and some machine learning and barcodes. And it's it's a fun story of actual like doing digital stuff in the kitchen Monday to Saturday, as it were. Yeah, I'm kind of similar to offering coffee to buy. So then also also in in Richard related news. So .NET Core is becoming all of .NET. Right. What did it would explain to us what that means? Yeah, I mean, that's had confusing branding for a while because you had .NET Framework, which refers to kind of the classic .NET that runs on Windows. And then you had what was really a it wasn't even a fork. It was just like a different version of .NET that was built to be cross platform, multi language, kind of built from the ground up to be more modular. Things like that was that's .NET Core. And so those two things kind of live side by side. They had different version numbers, you know, you have one going up for four, four, five, four, eight, another one going one, two, whatever. So this announcement instead of announcements was, look, .NET 5 is the new thing. And by the way, .NET 5 equals new .NET Core and .NET Framework is kind of frozen at four, eight. So once your brain kind of digests all that, I think that the story was, look, any new project should be .NET Core. Your existing .NET Framework stuff, it's not like it's retired. It's just we're probably not adding any more to it. It's done, if you will. It's kind of feature complete from our perspective. So I think that just means for most organizations who have a lot of .NET and, look, it's still all over the place, as you'd make sense, you know, as you'd expect.

It's still, you know, Windows is still half of the server environment. So most of that's probably running .NET stuff. So it doesn't mean you have to retire it tomorrow, but it does mean things like, hey, what we do with app transformation and things like that is really valuable stuff because you do want to put yourself back on some sort of mainline path where you're not worried about being on what will inevitably be something that's unsupported. Yep. Yeah. And so this is like, as I think is increasingly becoming the case, the Windows news update. Have you noticed that? We're always talking about Microsoft and Windows, coincidentally. But also over in part of Pivotal Cloud Foundry, the PaaS section, somewhat delightfully, confusingly named. There's also a Windows Server 2019 support that we've added in. And so, you know, and I was just realizing, do we call that like Pawswa? Like we need to come up with a way of pronouncing P-S-A-W. I was thinking possum or Pawswa, but there must be some way of saying that. Yeah, the S is silent. It would just be paw. Just paw. Right, right. Like an animal paw, not like a hillbilly's, you know, dad, basically. That's right. Not pawpaw. Right. So, yeah, so we did ship PaaS for Windows, as we call it, to five. And I did look up last night. And the last kind of server OS count was at 49.6% of server OSes are Windows. So, I mean, it's still all over the place on the server. And I say that because all the apps running on those things, especially when they're custom apps, need a better place to run. And I'm willing to go to the mat with anyone that still what we do with PaaS for Windows and Cloud Foundry with Windows is still the best way to run custom built .NET apps, especially in the enterprise, probably in the cloud as well. So, if you're looking for a better place for your .NET apps, the best place keeps getting better. That's right. And then, finally, there was a post from our head of product, so to speak, Ansi, where I think he was doing a very concise sort of roadmap, if you will, of things coming up. And, of course, there was continuing our thread, continuing Windows support, including in Pivotal Container Service being worked on.

And it seemed like, from what he was saying, sort of a unification of using all of the Envoy and Istio-based things for all your routing and wiring together stuff across whatever the OS and things may be, which is what you would want in a cloud environment. And then sort of just like a continual evolution of the idea of this is the platform for if you have an application and you want to deploy it to whatever type of cloud infrastructure scheme you want, it should be pretty easy to do in this setup. Now, I'm giving a painfully concise version of a concise overview, but what were the more detailed things that you thought were fun to look at there? Yeah, it was a good post from Ansi, who heads up engineering. You know, I think the main takeaways were, you know, A, Pivotal always does a lot of open source work. The numbers last week, Matt, as they were sharing it on Twitter, I think Pivotal was the number two by percentage of employee contributed to open source, which was, you know, it's amazing stuff. And so we always build a lot of open source. We integrate a lot of open source for a lot of companies that de-risks them because they're not the ones stuck, you know, stitching all these things together to try to build a platform. So all of this open source for years, just a reminder that, you know, we've swapped container orchestrators once, we've changed container runtimes three times, we've changed the networking stack in Cloud Foundry, we've changed all these things over time because that's what good platforms do, right? You know, Amazon swapped out hypervisors a few years ago. Did anyone realize that? Probably not, right? Everybody always evolves the underlying pieces of the platform. If not, you should be really concerned about that platform because that means they're probably getting stagnant. So one was, hey, this reminder that we're going to keep integrating all this great open source as it matures, Kubernetes included. At the same time, we're always innovating on top with great new things, cloud native build packs, things like that. And then the future is going to continue to be componentized to some extent of different best parts of PAS and Cloud Foundry and Pivotal that you can run kind of on any sort of infrastructure as well as that vertically integrated stack when you just want to go fast on any sort of infrastructure.

So it's nice to see that evolution continue, the innovation continue, and I thought Ansi did a nice job of putting it in context. Well, and then finally, as just a point to, if that's a way of saying it, there's last week, I believe, in The Economist, there was basically a special report this week, not on, you know, some obscure, at least for, you know, us Americans, some obscure country embedded somewhere and how their, you know, central bank is just driving fascinating things as they want to cover. But it was on essentially like, to use our tech world terminology, disrupting in the banking and financial world. And when, you know, other than the sort of cases that they go over, right, of course, there's kind of in a nauseating way, it's like millennials and what they do, which is always like fun. And I always like to think that, you know, they're probably like in their 30s now, so they're probably getting to be more and more just like the rest of us, beset with life and all that excitement. But on a brighter side, one thing that's kind of entertaining in the coverage to notice is like at the very beginning, they're like, well, we're going to cover like crazy, interesting things happening in banking, except in the U.S. Where there's a bunch of regulations and it's annoying. And so there's just like a whole, a whole major market of banking is just like knocked out of the analysis, which is, which is curious. But I think, I think what's the, the more interesting part that they go over that, that which I think is, is worth looking at is how essentially a lot of the, as we used to call them ride sharing services, because in various parts of the world, they essentially have financial access to you. I don't know how else to put that. Like you pay people and get paid through them, that they are extremely well positioned to basically do banking, which I think, you know, also you can apply that same strategic idea to things like Apple Pay and other people who are just like figuring out being middle people in the flow of cash everywhere.

But there's a nice little package to look at with some delightful economist charts and funny little graphics if you're into that thing. Well, I never asked you to, uh, introduce yourself guest. So do you want to introduce yourself? Sure. I, I assume people just know, but of course you get a very distinct voice. They're like, Oh, I know who that is. I know. I know who that is. Yes. I'm Derek Harris. I work at pivotal. Is that, is that a good enough introduction? You want the full, that was great. That was, that was like a, that was like a pre 240 character, 280 character, a Twitter bio. That's good stuff. Did you see that they let you, uh, as, as a footnote side note here, do you see that they let you put images in to retweets or something? I mean, I almost put that as the news item for this week, just cause that's really going to have a big impact on my life. Yeah. What, uh, how is this going to affect us, us, us three here and people like us? What, what do I need to be training for? Maybe this is why my appetite is increased. Just the anticipation. Yeah. More, more blood to your brain is needed. So I don't know. I think it just, uh, you know, if you want the retweet with commentary that just includes a picture, like no words, just something from dodge ball or, you know, whatever you have choice. I think that's going to be a power move now, not just the retweet alone. And I even just with some pithy words like, no, here's just a picture. I see that is. So basically there's going to be that like fry narrowing eyes animated gif. And then that Homer backing up into bushes is just going to appear everywhere now. Like, that's right. Okay. Okay. Along with like, along with like honey boo boo or stuff I don't know about. Yeah. I'm just going to spend more time coming up with, not with pithy words, but instead just pictures that represent my feelings. So I'll say even less. Got it. Got it. Okay. Okay. Well, back to you guest. You just keep calling me guest too. That's fine. You know, it feels, uh, it's fun guest, a special guest. You could even have that. Uh, so what, uh, tell us a little bit about, uh, I don't know what you've been doing over

the past few years. Uh, like what, uh, what's, what do you, what do you like doing with your time professionally? So, so many things, so many of the things. Um, yeah. So let's see. So I assume, I always assume, and this could be the time travels faster now, I think. So I, I assume many people know me from GigaOM where I wrote, if anyone remembers the, the original incarnation of GigaOM, right. Where I cover, but I wrote a lot about cloud computing and big data and artificial intelligence. Um, I helped put on and plan and interview people at the, the various structure conferences. So structure and structured data and all these things. So, so I mean, that is, you know, that was probably the, the, uh, the, the most memorable thing that people seem to connect me to. Um, however, in, in the meantime, GigaOM went out of business in 2015 and March. And since then, let's see, I've spent some time working at Mesosphere. I've spent some time working at a small startup called Replicated. And I started my own newsletter, podcast, news site, that's what's called a publishing empire called Architect. Uh, as I always remind people, spelled T-E-C-H-T as in technology, architect, with an H through at the end of there. Uh, frankly, if the correct URL had been available, I just would take a mat, but you know, if you do what you, you do what you could do. Um, and, and so, so, so, so yeah, so that's what I've been up to. And I joined Pivotal in July. So, oh yeah, almost a year, almost a year. And I'm doing all sorts of fun, uh, editorial and content. Yeah. Yeah. You, you, yeah, I think over the years you've always had, uh, one, one of the, uh, I mean, I don't know what the ordering of the top three or five or whatever it is, but one of the better, like roundups of all the, uh, the, the cloud. And as you say, AI and stuff like that news. Like I remember, uh, wasn't it GigaOM that was like a quarterly, like, I'm kind of exaggerating here, but like 50 page PDF summarizing all the cloud news of that quarter that you did.

Well, as briefly. So yeah, so GigaOM had, you can, you, we can get all into the, the, the, the Giga, the demise of GigaOM, but, uh, GigaOM had an editorial business and a research business and right for a while I worked on the research side and yes, once a quarter, this would have been back. I'm sure someone else kept doing it. I did it maybe in 2009, early 2010. And it would be like, it would be, yeah, it would be like once a quarter, you would just go through. So I'd be collecting stuff all quarter and then go back and like do this highlight, this big recap of like, here's everything that happened now. Of course, that would be 5,000 pages and you could do that. Like, so when I started doing a newsletter, I was like, I think I have to do it daily. Because of the things that, when I see things I think are interesting, there's, there's at least 10 things a day. So totally. There's so much stuff happening. It's crazy. Yeah. So what do you, so I mean, as you've watched that whole journey of cloud from just the thing that, you know, people talked about, but no one was using, or it was just kind of completely bleeding edge to where we stand today, where you have to have a legitimate reason to not be using cloud computing. I mean, what do you think of that history? What have you seen? Are there any sort of mile markers along the way that you thought caused maybe a big lurch forward for everyone in the industry to kind of start adopting certain things? What have you kind of noticed from a macro perspective over that time? Well, it's funny. I mean, I think the number one thing I've noticed from a macro perspective is that it's kind of come full circle in the sense that if you look at cloud today versus whatever, moving to AWS launch 2006, technically, if my, if my memory is correct. But, but, but it used to be like, now it's just kind of, it's table stakes, right? To use an overused term, but it's, it's what's expected, right? And there are big, there are big fat multi-year contracts. There are a handful of providers, you know, to me, well, that's probably the weirdest thing is that it's just, it's, it's the, people are lifting and shifting applications there, right? I mean, that's a big, that's a, that seemed to be like Google's big push recently.

Like that, it's, it's kind of, I mean, on, on the one hand, as a business, you go cloud computing. That's not like, why are we still calling it cloud computing? It seems like just enterprise IT at this point, right? So I think that, I mean, on the macro level, that's probably the biggest thing. But of course, all along the way, there have been these things, you know, these, you know, there have been these improvements to the way IT functions, I think that, that have propelled it to where it is, right? I mean, just the speed of provisioning infrastructure as was kind of eyeopening, the programmatic nature of it, the, you know, security and reliability and all these things. And honestly, I think people were maybe rightfully afraid of in, you know, 2008, but now they've realized, yeah, these things are, uh, some recent history excluded like last month or earlier this month, uh, these things are pretty reliable, like much, like, like, you know, like, I think a lot of the fears have been ameliorated. Now it's, now it's a matter of, okay, how do we get there? Cause we're, we're, we're, we're pretty convinced this is the right way to go. And we trust, we trust these companies. So would you have, would you have anticipated that we'd be dealing with a few mega vendors now versus just some sort of garden of providers that we all would have picked for vertical needs or for geographic needs. So was this pretty inevitable in your mind or is where we stand today, a surprise to you? Well, you know what, when I first started, I was definitely in, and it could have been just Nebitae. Um, I, I was pretty, I was pretty convinced there was going to be a handful of small providers and men, maybe that's just because that's what the small providers were telling me. Right. And I was like, okay, if you say so, you must know more than I do. Um, because it seemed like, you know, cause again, Amazon came out of nowhere to go, okay, well, who are the, this was a bookstore. And then, but right for a while, I mean, if I look back, there was, you know, if I look back at all the cloud providers I've taught or anyone who called himself a cloud provider that I've, even if it was a glorified hosting provider that I've spoken with over the years, I think, I think, right. You'd be forgiven for thinking, oh, some of these are going to last. And it's going to be, you know, why wouldn't there be a cloud for, you know, the, the film industry for rendering that just specialized on rendering?

Why wouldn't there be some high performance cloud? Why wouldn't there be all these things? Well, it turns out it's just easier for the, for the big cloud vendors to, to do all that. So, but I think for a while, just if you looked at history, it was like, yeah, of course, why wouldn't there be, I think, I mean, I think platforms have always been a thing, but I don't know, maybe it's, maybe it's, and maybe my history is incorrect, but just when you see the rise of platforms everywhere today, from your phone to your social media to everything, it's like, oh, okay. It would, this was probably inevitably going to be the, the way it worked out. But yeah, for a while, I really thought it's like, there's going to be Google and Microsoft and Amazon. I actually Google for a while wasn't even on the radar. And then there's going to be rack space and go grid and God, or some of the other ones that I've talked to that I used to talk about. It's like, you know, they, they allow the space. Sure. Why not? Maybe not. No, I mean, the Gardner magic quadrant would have put 15, 20 providers in there for IAS and everybody was upstarts. And now I don't think they're really doing that, that MQ the same way anymore, just because it's not as, not as meaningful, but have you, uh, what do you think of the lock-in concern? Is that just overblown? Is that us forgetting history? Is that us realizing that this is different than maybe the giant providers of the eighties and nineties? Is that something anybody should really care about? I, I mean, I think there's something there. I don't think it's, I would say it's, it's, it's probably different and you probably have a better take on this than I do, but it seems different in the sense that, you know, we're not tightly coupled to hardware or, you know, we're not definitely not tightly coupled, you know, in many cases, like an operating system, or I think the level of lock-in is definitely different, but on the other, but, but it's, you know, it's kind of an API lock-in. And, um, you know, there's, I think so much stuff is built on open source today though, too, that theoretically you can put together a stack of things that are running, that are running based on open source components. And there's, there's less of a risk of lock-in, right. Um, I don't know my, my thoughts come and go on this on the one hand, I think, well, you know, if it's the, you know, if the switching cost is not insurmountable, then I don't know.

I mean, sometimes it seems like the switching cost from Oracle is insurmountable. So people don't do it right. I, I don't, I don't know the reason all the time, but, um, yeah, I think, I think it's a concern, but I also think, you know, depending on the level, right. I think to, to be obsessed with lock-in is probably, it's probably going to impede progress at any level, right. Whether it's the ISV level, whether it's a cloud level, whether it's whatever, like you're going to accept some of it. And anyone who thinks that we're in this world where you just get to use like upstream open source and, and everything's going to be better and nothing, you're not locked into anything is just wrong, right? So I think you have to accept, like, there's going to be some, you just have to pick and choose your places. Yeah. Yeah. There needs to be some kind of like a three hour, uh, panel discussion about like the new lock-in or something. Cause it is, uh, you know, just like we were doing earlier, lock-in used to mean, uh, basically like, uh, I'm running my stuff on windows and, and that was like lock-in as far as I could tell. Cause if you were running your stuff on open source back in the early two thousands, there was no lock-in. Everything was cool. And so it was just, and just so much of it felt like it was packaged stuff. Like if you were locked into IBM or you were locked into some of these Oracle or even Microsoft at the time, it's like, you're not just going to suck your stuff out of SharePoint and run it in some other CMS. Like it was just completely coupled to it versus most of the software today that I write or I buy really can run kind of anywhere. Yeah. Yeah. So, so, so, and, and, you know, so, uh, there's something that people think about lock-in as, and, uh, I don't know, it'd be nice to redefine. Maybe, maybe now that you don't have, Gartner doesn't have to track the magic quadrant of the three major cloud providers. Uh, they could, they could come up with some, uh, they get those Burton group people to write some treaties on lock-in or something. That would be a fun read. 80 pages on lock-in looking forward to it, but yeah, you know, uh, so, well also, so related, you know, we were joking about how, uh, uh, you know, used to do a quarterly and then, and then for some reason you decided it was a good idea to do a weekly news roundup of, of, I guess,

cause like you said, it would have to be 50,000 pages if you did it quarterly. So you got to do it every, every couple of times a week, but now, uh, at pivotal, you started a new newsletter called, uh, intersect. So what, what's this letter all about? Yeah, intersect is a, a newsletter and a website, uh, pivotal.io slash intersect where, you know, targeting, I would say CIOs, I'm gonna be a little specific, but you know, anyone kind of in the C suite and the executive suite at a large company, just, and just trying to give, give some perspective for what's happening, right? So the newsletter is going to catch you get a little analysis, it's going to catch you up with, if you're reading it, it's going to catch you up with what's, what's happening in the week, not following every piece of industry news and every piece of software vendor news or cloud provider news, because again, that's, that's probably overkill. And if, you know, you, you, you, you, honestly, I don't think that's a level where most folks need to be keeping up, but, but more keeping up with like, well, like what are trends, right? Like what are some big, some bigger things or some use cases or some applications of, of these things to keep an eye on, right? I hope, I hope it's a little more useful in the sense of, Hey, what's this company doing that might be in my space or, you know, how are, how are, how are these companies using AI, right? Or what, Ooh, what's this big security, what's the latest security breach I need to worry about now? You know, are we protected? Like, like that sort of thing, rather than like Microsoft, you know, running down the, uh, the 57 things Microsoft announced this week. Yeah. And it does something that like, it's, I think it's hard for a lot of, uh, new sites to do, which is, it not only has, you know, there's not only like, uh, I don't know what do you call original content, like, like the, its own content there, but like, you also list, uh, related things that are happening off, off of the URL as it were, which, which exactly serves the point of what you're saying is like, you know, here's, you can keep up with things that are relevant to this, uh, this space and, and, and beyond that you might be interested in. And, uh, I don't know, it's, it's, it's, uh, there's not a whole lot of like in the, uh, I don't know what to call it, enterprise infrastructure kind of space.

Uh, sort of coverage, like, like, like that. A lot of it is just, you know, what's the latest odd Facebook thing or what's up with the pixel phone. So it's nice to have a, a reconcentration of that, which, which is like, I think, I think you've always provided lists of stuff like that in the past, but you sort of, uh, tirelessly have kept that up. Yeah. I hope people appreciate my, you know, 600 words of opining at the beginning. I, I, I, I think I like to think that I've, that I've, I figured out something over the years as I'm pushing 40 and have, have some, uh, if I have some insights to share, but, but yeah, and I think you're right. I mean, just curating, curating the fire hose of, of news is like, right. You know, you can't, you can't play the horse race game of like, you know, keeping up with every single thing I was, as I was looking this week, it was like, there were about every new site, at least some of the, some of the ones had like, you know, 15 headlines with Microsoft is the first word and announces probably as a second word. And then another 15 for Google, which is all about smartphones. And then a couple about the TPU clusters that you can rent from Google cloud. And then, but it was like, there's this big stack of things, right? So there's that. And then, yeah, there's, there's trying to keep up with that. And then there's trying to keep up with the, the, you know, just, or there's trying to break out of, right. Like the Facebook, Netflix, like the thing discussion and, and what's, what's this kind of, there's a very, uh, very fun and interesting, uh, discussion at the, I guess you would call it the enterprise IT layer. That's more, more impactful than, and more meaningful. I hope than just trying to keep up with the news. Cause again, who, like if you were to choose a cloud, if you were thinking like, you know what I'm going to do, I'm going to follow Microsoft build, and then I'm going to decide what my cloud provider is, right? Like that's not your plan. That's not a lot of use. Cause Amazon will hold a pop-up event next month and announce 15 different things. Right. And it's just, you can't, you can't play that game. So it's like, let's try to be a little more, a little more purposeful. Why don't we do that? Yeah. So that's the goal of intersect. I mean, the best part of the newsletter, at least, you know, it's, as somebody who now reads it all the time is again, I do like how you're trying to throw things in context, right? It's not a link farm.

It's, Hey, here's a really, you know, it's a thoughtful opening. Here's like three things you should take away. Cause you and I've talked about this, Derek Coté, you and I too. I think I, I just still don't know how people filter the unbelievable amount of things happening in our space and make good choices. Like, it still fascinates me. So I think we're trying to, you know, with this all started, cause Derek wanted to kind of help make a difference and helping people focus on things that matter and then try to sift through all the craziness. But, you know, do you see that as a part of the responsibility, Derek? Yeah, that's, that's absolutely it. I mean, I thought it occurred to me, right. What occurred to me pretty early on was like, there's a value. And I think, you know, part of it is like the navel gazing of, of someone with a journalism background, like, well, you know, what's the future of journalism? How do you make that work? And then newsletters always pop up as the thing. And then you start to think like, why is that? And frame, well, one of it is like the personal connection. And the other thing is if a curated newsletter just does the job that you don't have time to do. I mean, if I open up Feedly and go, holy crap, there are over a thousand unread things just from this week, I got some work to do. Right. And I don't, I don't, I'm willing to shoulder that burden and hopefully put, you know, give, give the, you know, give readers the output of that. Right. And what happens in the middle in that black box of my, my judgment is, is work that they don't have to do. So. Yeah. Yeah. And I mean, you're, you're, you're, you're hitting on something that like, you know, also I should mention, I need to go dig this up, but we have a, a lost interview episode that we did way back when from my other podcast. I need to finally post that. We talk about all this stuff for like three hours. I think it's actually like 40 minutes, but it was seemed like three hours. It was packed with so much information and analysis that it seemed like it had to have been more, but it is like, uh, aside from those, uh, uh, sort of like overeaters, like I think the three of us are when it comes to like tech news stuff, like, man, no one's really like figured out how to curate that stuff. Like, I think the, the, the thing that I found that works slightly okay is like, uh, like

you ever use nuzzle, like that's kind of good for, for doing this stuff. But like, then it's also, you know, has the annoying, like what everyone else wants to read thing. It seems like, it seems like there's an inherent, like, um, I don't know, business problem in trying to like be a robot that's curating these things is like, I always want to get my like very, uh, low value as far as advertising, I guess, I don't know, news stuff and read all that. And then there's always some tab that's just like, here's like the big gigantic mainstream news that we would rather you start reading. And I don't know, it just never works out very well to have like a nicely curated set of news, or maybe it's just hard to actually find it. And it takes a lot of time. And, uh, yeah, I, I found, I thought you can pick your sources and that's, that's good. But if your sources like tech crunch, good luck or Politico kind of like, right. And then, cause, cause they don't sub curate and then some places will actually let you curate, which is, I think is the way most publications should do it. If they're producing a lot is to curate by topic, which makes it a little easier. So I can have a feed for like X publication, X, Y topic. That's super, that's, that's useful. But, and I found the algorithm, the strictly algorithmic tools. I never, I like, I tried to went back in the day called prismatic, which doesn't exist anymore, but they tried to do it. They tried to do it based on a shout out to Bradford cross. He's a smart guy. Um, but, uh, they, yeah, they tried to do you based on your, so based on your social network and based on other things you liked, you know, try to do a algorithmic way of doing it. It was, I mean, you would find some stuff. There's like that serendipity, but it also was, and then as a professional reporter, I was like, well, I'm afraid I'm missing something. Like what are the, like, if I use this, what if I miss something that I should be seeing? So yeah, that's kind of like, uh, sometimes it's the manual grind is the, for me is the only way to do it. Even if I'm, you know, you know, putting a lot more work on the plate. Yeah. Yeah. As you, uh, as you chew on all this, as you're just kind of absorbing all of this news and information, can you point out, you know, one or two things you think are actually under

hyped right now? Like we're not talking enough about X and maybe one or two things you go, I feel like we're talking way too much about this. Anything that fills that for you? That's a good question. I mean, I don't know. I, I want to say that I want to say that Kubernetes is overhyped just because like there's, there's so, so, so much talk about. And I would argue, you know, it's not what I'm going to say. I mean, it's, it's, it's, it's, it's, it's crazy. It's, I think the, I think people have convinced themselves it's the future of everything they're going to do and like, that's what I need to focus on. And there's probably some truth to that. I just think that it's, as, as we're definitely seeing it's early and that this is going to be that what you're seeing today is going to be how this looks even, you know, five years from now or fewer is, is something I wouldn't bet my life on. So I think, I think people, I think what happens now is, is folks are rushing to get ahead of or get onto something because they don't want to miss out again. I think a lot of people felt maybe a lot of people felt like they missed out on cloud maybe, or they got in too late, even though it's still by all accounts early in that space. Like you just don't want to be seen as a laggard. So, so, so folks are rushing to, to get in on that, but I think it's not, you know, I think, I think that's evolving so fast, that whole space that we're not quite sure how it's going to play out, right? Serverless might be one of those things as well. So like, so everything is kind of puts us in just the future of, but I, you know, it's not, not entirely sure. It's not entirely certain that's going to be the case. You know, the other thing, but like cloud, I just, just, I'm rambling here, but like, but cloud computing has changed a lot. But one thing that hasn't changed is there are still virtual machines, right? And there's still IaaS and like you get storage and database, like that part of it. There are these things that aren't going anywhere, right? And everything above that is kind of up for, up, up to evolve. I would say underhyped, underhyped is tough.

I think I would argue that, you know, I, because I spent a lot of time thinking about writing what AI, artificial intelligence, if someone didn't get that acronym, I think that that's definitely overhyped, but I also think there are some really underhyped aspects of that, that don't get the attention, you know, the overhyped stuff is like the cloud, you know, I think some of the facial recognition stuff, although there are privacy concerns, gets a lot of attention, you know, the existential threat stuff used to get a lot of attention. It gets less so now, but, but on the other hand, there are these, when, when places write about like the latest deep mind model that can beat Starcraft or that the, what was it? AlphaGo Zero, which was trained in like a day and was better than all the other Go playing systems they built. There are some really, uh, like really big applications for that sort of thing that, that people tend to overlook cause it's complicated. But I remember when the, when the AlphaGo Zero came out and there was this, like the line on the end of DeepMind's blog post, which was like, you know, you could use this for materials science and, you know, drug discovery and, you know, like a lot of hard scientific problems could be run through a deep learning system like this and really do something powerful. That's much more than speech recognition or, or object recognition or something. And I, so I think there are these hard problems with AI that, that tend to get over, that tend to get overlooked. Um, but if we, if we focused on that, then maybe when, when someone says AI is naturally into the economy and do this, like we would have like really, where's the details? But if the details was because look at what we're doing, but that, you know, that I don't think that's what anyone is thinking when they're writing those, those analysis or those stories. So I don't know. I think edge computing, frankly, might be under hype too, if that's possible, just because you, if you think about it as servers at the edge, like the Chick-fil-A thing you mentioned earlier, like that's one aspect of it, but when you start, when you start thinking about 5g and, you know, interconnected devices, personal devices, and how, how our like personal communications

and networking and stuff is going to play out, like there's a, there's a lot that's going to happen there that I spoke with, um, on my, on my podcast, which, uh, I, I'm not doing any more actively, but I spoke with Jason Hoffman. He actually, he co-founded joint back in the day as cloud or cloud, another cloud provider that, that, that still exists, I guess, but, and not, uh, you know, not light in the world on fire. Um, and, and he's, he's at this company getting the startup now called mobile edge X, and he's a really smart guy about networking and computing stuff. But one of the things we talked about was like, you know, if you're, if you think about virtual reality or augmented reality or, or any of the, any of the new ways of interacting and even mobile payments, right. If you think about the ways we're doing these things now, there is a strong component of like, what's the network going to look like that supports that? What do the devices have to have that supports that? How does this all look? And I think that part of it gets, it's overlooked. If we think about edge computing as simply like there are driverless cars and small data centers at the edge, right? Like, or, or servers in a factory. There's just a, there are a lot of pieces there that might, that might be able to do a lot of really unique things that we've only talked about up until now. So I think it's just, I think the, the underhyped parts typically tend to be in my opinion, but it's just like people willing to go a step or two below the surface to talk about how this actually pans out. Yeah. Over, over the, the three things sort of like a cloud, cloud slash Kubernetes and AI and edge stuff. Like there's at least, at least as I'm thinking, as you're talking, there's a, there's a running thing of like, so what do, what do we actually do with this stuff, right? Like, like what are the actual features and things we can do with it? Right. Like, like as far as, uh, you know, I guess it's cool to be able to beat someone at go and, and jeopardy or whatever. That's certainly, I mean, you know, entertainment is a valuable part of our world and a big industry. Uh, but, but on the other hand, just as, as, as you were talking about, right, like we got to know when that chicken expires or like what kind of coffee you want, just sort

of like boring AI stuff. And I always get excited when I, uh, excited, I don't know. I guess I do get interested when, when I hear about like just some boring novel use of machine learning or AI for something that, that would, uh, I don't know. Cause a lot of that seems to be stuff that would make my life easier. Right. But boring novel uses or amazing, you know, potentially world changing uses. Right. But it's, you just, you have to do the digging to get there. Exactly. If more people thought about this stuff application first, you know, then I think you would, we, we, we might be onto something. Yeah. Yeah. It's right. You think like, what would I build, what would I build to do? Like what, what's the application I want? What's the thing I want to do? And then, and then, and then work back from there versus I want this thing. Oh yeah. No, that's, that's, that's, that's, that's a good like mental framing for like this, this little discussion here of overhyped underhyped is like, instead of thinking of the technology, like, as you say, think application first and see if you can work backwards to I don't know, Istio or whatever. And then, and then, so, so, so, you know, as, since you bring it up, like I've noticed some, some, some agencies and analysts, people are spending a lot of money on this phrase edge computing. And so like, what, what is edge computing? Well, I think it's many things to many people, you know, Gartner was at a Gartner conference a few months ago. Gartner certainly has its preferred definition of it, which I can't recite off the top of my head, but I, I, I just think of it as, and maybe it's just, I would say it's, it's, it's smarter things. I mean, I think the definition of it is basically like not a centralized cloud, right? If we look at it like that, but also it's not just, obviously not just like server closets and whatnot, right? I think it's, it's, it's that, it's that distributed, those distributed computing devices, plus the network connectivity, plus probably some level of intelligence, right? Yeah. So I think, I think that's always implied, even if it's not expressly stated. So I think if you combine some sort of machine learning, plus like low latency networking,

plus, you know, small, you know, at least, at least much smaller than cloud scale, uh, down to the device or sensor level, um, you know, computing element. Yeah. Yeah. And then I think, I think to layer on your framing, right? Like I think, so first of all, that all, that all comports with what I've stumbled over. That makes sense. Basically like instead of internet of things, it's like internet of computers to be kind of silly about it. Uh, but like, I think, I think, you know, with your framing, the thing that is probably more interesting is like, so what are you going to do with that? Like, what's the application for, for edge computing as, or edge stuff. And I think that maybe that's, maybe that's what I've been missing a lot in the conversations is like, you know, sure. You got all these whiz bang, you know, boxes hanging out in a dark room somewhere doing something, but like the, something that they're doing is what's actually interesting about what edge computing is. Cause then like you're saying, it's like, what's the difference between just like a couple of machines that are blinking in a closet, you know, versus edge computing. And, uh, I guess, I guess what they actually do and what applications they, they go after is what makes the difference. Otherwise it's just like, you know, I don't know the equivalent of netware in 2020, like not, not really that big of a deal. Yeah. I mean, if I were being a complete cynic, I might say that a lot of, a lot of vendors that, that started up trying to push edge computing from the infrastructure side, we're like, well, we can't compete in the cloud anymore. So how do we make it look like cloud providers aren't, aren't the thing, right? Like cloud is the future. And then, by the way, there's no saying that the cloud providers aren't going to just dominate the, the edge computing future that it is indeed. Because to your point, they're all involved in it too, right? Like, so, right. They own the devices, by the way. Yeah. Yeah. Which almost, almost turning, uh, your cynic point to, you know, your cynic persona to an optimistic thing. If like, if everyone in, in the, the market is kind of going after it, then there actually is something there. It's not just a, um, defensive posture against something else.

Yeah. No, I, I mean, I think there's something there really. I mean, but it's probably things, what I, what I found is that, you know, there are lots of things we used to talk about. I remember covering big data back in the early days and it was like, you need all the data, right? The, the unreasonable effectiveness of, of big data, right? And you, so you need all that. And then, and then there's all the stuff that's only happening today was like some pipe dream of that was 10 years ago, but it's only happening because we have the networks we have now where it's only happening because we have the smart devices in our pockets that we have. Like, like all these things are happening today. And I mean, and it's, it's basically because of the things that I think are that we call the edge today, right? So the data processing, all these things are part of it, but it was totally reliant, like any sort of, you know, real time and accurate geo-fencing or, you know, it was always gonna be like, you're going to get a deal when you walk into your phone and knows you're in Seattle or something like whatever, whatever it was that people were like, this was mind blowing. Like now you can do that. And it's only because of, and it seems mundane, but it's only because of these, these bigger infrastructural pieces that have been put in place. So, yeah, yeah. I remember, I think I forget if I mentioned this on some episode a long time ago, but to your point of like in-store stuff, you, you probably talk with, had similar conversations, but I remember talking with some, I don't know, some Nordic or Northern European person who worked at one of those companies that like, uh, offers you coupons in a store and like the, um, the way they had to implement it, this is kind of like the Starbucks thing. It's like, they had to go through all these crazy hoops about like basically identifying your Mac address on your phone. Uh, you know, if you logged into something, then like, it's not a problem, right? Like, like if they could scan your license plate and you held your license plate up in front of the camera, or even better, you walked into the store and typed in your license plate number, it wouldn't be a big deal. Right. Uh, but really, you know, they can track your Mac address and a couple of other things. And then, so they can kind of like track that over time. But then he was telling me the annoying thing is like, I forget versus Android or iPhones or

whatever is that oftentimes the phones will just like change their Mac address. And so there's like all this stuff you have to put up with and, and to your point, right? Like there's a whole lot of like maximal data collection and then analyzing things that you can do, but like you need that capability, uh, in order to do it in the fast networks and all that stuff, uh, behind there. So apps. Yeah. Yeah. Yeah. You need the actual, uh, you need the actual thing that does the thing. Exactly. Not just a box. Well, great. Well, uh, I'm glad we could finally get you over, uh, get, get you over metaphorically speaking over here, but, uh, as a guest on there. So, so, uh, if people want to follow up, uh, tell, tell them where the intersect thing is again and, and your, your Twitter handle and all that kind of stuff. Where, where can they, they, in a very non-creepy way, follow you around the internet? Yes. Most importantly, yeah, you can go to pivotal.io slash intersect and you can read there. You can also sign up for the newsletter, which I would recommend doing. It comes every Thursday, ideally to, to your inbox and yeah. And you can also, you can find me at Derek Harris on Twitter. I'm not very active these days, but maybe that'll change. So again, you can, you can find me there. That's right. I think the idea of cloud bursting has moved to Twitter. There's Twitter bursting just every now and then there's a lot of activity. You wouldn't, you wouldn't just have bursts there. Well, as always, this has been a Pivotal Conversations. If you want to get the, uh, the back catalog of episodes, several years of it, you can go to soundcloud.com slash Pivotal Conversations and also about every Thursday, not this week. However, uh, we post the full show notes glorious as they are over at pivotal.io slash podcast. And with that, we'll see everyone next time. Bye-bye. Now this reminds me, speaking of pizza of one last thing I wanted to get in here is, is, uh, uh, my, my, I don't know what's up with my dog.

She's upset about pizza. Probably hungry. Yeah. I'll have to do some editing. Two bugs. Hey, let me see. Let me make a note of this. Two, two minutes, dog barks. This is terrible. And now the dog thinks I want to play with her. Two minutes, dog barks. Okay. Don't worry. Don't worry. Uh, all right. Uh, and then around two 30. So this reminds me of something I was realizing recently speaking. I was, I was eating some.

Tanzu Talk · 🍕 Get more pizza.