Everyone’s Getting Better at Software, Highlights from SpringOne Platform 2018

Episode 101 · 48:33 · SRT · archive.org · all Tanzu Talk

Well, I'm here in London for dual purposes. I was at the FutureStack conference for New Relic on the partner panel. There was actually a fun discussion of, I thought we were going to have to talk about how great our partnership is and everything, but they just wanted to know about DevOps. So that was a lot easier. Here's a pro tip. I don't know if you do this, Richard, but you should ask for the questions they're going to ask ahead of time or the themes, which thankfully I got late last night so I could think about these DevOps-y things. Yeah, you were well-prepared. Always, always. I prefer to think of it as I was optimizing my margins by just-in-time preparing, as it were. But I got them last night, so I had some time on the plane to sort out the answers. You know, it was the usual exciting things. What do people start with, automation or development first? You know, of course, the answer is both. Otherwise, I said, you know, automation, because if it takes you like six months to ship, all of your great development is for not using sort of like lean gold-ratany in theory. Gold-rat-ian? Anyways. And then I think the other one was like, what metrics do people use? And I've kind of forgotten the third one. I should look that up. Anyways. But that's not what I was going to talk about. I'm also here for a meetup that if you want to go to, you know, if you listen to this show right when it's released and you live in London, it's tomorrow night, August the 2nd. On metrics, what Coté has figured out so far. Yeah, October the 2nd. Isn't that tomorrow? I said August. I thought you had a DeLorean. No, no, no, no, no. Man, that would be nice. You know, that would be a fun episode one day is if you could go back one month, what would you do differently? I guess that was two months. Anyways. But anyway, so while we were there, some sort of door or fire alarm went off. And I recall the last, one of the last times I was in London presenting at the register conference during my talk, a fire alarm went off right before the best joke in the whole

presentation. And we all had to leave the building. And then as we were starting, I looked up the TV in my hotel room here and it says, please be advised that today at 4 p.m. we will be undertaking our weekly planned fire alarm test. So I don't know what the deal is with me in London, but I really bring the fire alarm testing when I come to town here. So people should just prepare. So if in about 20 minutes you hear a fire alarm, everything's fine. That's as they say, this test is an essential requirement in order to guarantee your health and safety with health and safety capitalized for some reason. So that's just a warning for what's going on. But last week we were at, we were at SpringOne Platform 2018 in National Harbor. How did it go for you? It went well. We didn't have any fire alarms, but there was a bat in the keynote hall. So that was arguably the most exciting thing for me for the week. Really? Yeah, I saw it. Like it was pretty awesome. Like that bat's flown flying overhead. That was, there was no mistaking it. That was a bat. You know, you remember that, you remember that story, that story, that movie, American Beauty, where there was a guy with a camcorder and plastic bags and how artful they were. Like I think, I think, you know, a lighter version of that might be like birds in grocery stores. Like I'm sure that's a phenomenon you have where there's just like a bird in a grocery store all of a sudden, just kind of sadly flying around, trying to figure out if it's going to like live there for the rest of its life or like not kill itself bumping into things as it flies out those relatively tiny doors to the rest of the thing. I was obsessed with his backstory. Like I really wanted to know, yeah, how did he get there? Was his future just going to be in that conference hall, just dive bombing speakers? Like that'd be an amazing life for him. Hopefully he learned about digital transformation. Well, I have to if you're going to listen to the SpringOne main stage. You're left pretty inspired. Exactly. So it was a good show.

You know, in the moment, it always feels so frantic and manic and I can't wait for it to kind of be done. And then by like three days later, I'm like, that was a really good show. I had a great time. But it takes like being away from the show to actually appreciate it. It's true. It's true. Yeah, there were several, you know, we had Pivotal Cloud Foundry 2.3 announced and Pivotal Container Service 1.2. And there's, as always, from your team, there are extensive write-ups of everything that's running around in there. I was looking at the PCF releases and among many other things, you know, back long ago when I developed, I did J2EE back when there was a 2 in it. And as I was reading it, I was occurring like, oh, there's all these services in the platform now. Like when I joined Pivotal like three years ago, I think a lot of PCF, there were services in it. But a lot of the discussion was, I don't know, operational, sort of like core compute and hooking, you know, a little bit of the networking and hooking components together and building the containerized apps and things like that. But in recent times, in addition to like, you know, there's a whole slew of security stuff and there's always like some isolation and managing your sort of multi-tenant from the perspective of, you know, spaces and orgs and stuff and PCF. But each release, there's a lot new features that are in there. You know, I'm sorry, services that are in there, which, you know, are essentially to some extent middleware that you might have. But there's also things like, you know, what set me on this, this trail of thinking was that there's a, to disambiguate from container land, there's a task set scheduler in there. Which I remember, you know, it'd be old man Coté. I remember back in the day of J2EE, you know, scheduling things was a, it was a difficult but fun task to use. Like, I think we used like courts or something. I don't know. I forget what like the big open source Java scheduler was back then.

But it was like understanding how two-phase commits worked and then also understanding how to tune a JVM. Those are right there with like understanding how to schedule things and figure out that working. So anyways, I mean, that's, that's sort of like my major takeaway without detailing the PCF release so much. Yeah, that was good. I'm going to be endorsing you on LinkedIn for JVM tuning. So that's going to be, it's going to be a big day for you today. Thank you. But yeah, there was, there was a number of good PCF announcements. As we talked about, PAS 2.3 or PCF 2.3. We had, we've had Jared on talking about some of these releases in the past. But, you know, the CredHub stuff, as you mentioned, the services, like I went to a few CredHub talks and like that's pretty sweet stuff for storing credentials, using it for your apps, you know, even generating certificates now that you can use. So like those are ancillary services you might not think of that you have to bolt on. But like, that's just all in there or service discovery or Spring Cloud Dataflow keeps doing some really cool stuff in PCF itself and HealthWatch. So yeah, it was a nice highlight of that. And then we had the PKS announcement as well with Amazon Web Services support and multi-master support and a lot of good stuff there. So yeah, the kind of flagship announcements around PCF, but there were a ton of Spring announcements. We have a link in the show notes to Spring Tools 4. This is if you like building Spring apps, but don't want to just use Eclipse. You can use Atom. You can use Visual Studio Code. That's all in these various marketplaces and things. So a lot of Spring-based announcements, Spring Framework 5.1. You know, a lot of great things are socket kind of, I saw a story written up today on InfoQ about our socket and we announced that this was this new protocol level thing that works kind of at layer seven, but for more reactive systems developed by Facebook and Pivotal and Netify and kind of these companies coming together to build maybe a more efficient communication protocol versus just straight HTTP. So a lot of really neat things came out last week that I think surprised some folks. Yeah, and as I was going over stuff, how is the .NET stuff doing in PCF?

Since you're a Microsoft MVP, you must have an opinion. Thanks for reminding people. We did do another Windows update, Windows Server 18.03. It's not just the versions anymore. Now they more or less ship Windows Server every six months. And I believe that the support is only like 18 months. So you're getting these Windows shops into this world where it's being updated all the time, which is weird for companies still sitting on Windows Server 2003 or 2008 or these versions. So the fact that we bake it into the platform and just roll it out there with no downtime is nice. So a lot of good improvements there. I actually moderated a panel with a woman from Fiserv and a fellow from Magenic, the consulting firm, kind of talking about their transformation with .NET. And the Magenic fellow was dogging me a bit that there wasn't even more .NET content at SpringOne. So we've got to up our game to have even more talks. But there was a great number of talks this year, more than ever, on .NET, around Windows. So you'll probably see more and more of that as we think that's a big bet for us. And our customers seem to like having a decent Windows story that's all managed by the platform. And, you know, to add to the reasons why I should be endorsed on J2E, you know, I probably already have an endorsement for that being the fancy pants that I am. We'll see. It's probably right next to like Styrofoam peanut construction from that era a couple of years ago where people would just make up crazy LinkedIn endorsements. But anyhow, yeah, I mean, I think one of the other things that's notable is that, you know, of course, the embedded operating system was updated. And in PKS, the Kubernetes instance was updated. And, you know, the following kind of falls into a little bit of the, let's call it EBC sort of like, you know, fear sort of inducing stuff. But there is a good example related I was reading to the need to patch things and keep things updated. And you might remember last year, maybe you don't, but well, you probably do because you would notice this kind of thing like me that there was, I don't know, I think 15 hours where you couldn't file your taxes in the U.S. with the IRS.

And I came across, I think it came out a couple of weeks ago, I came across a, I don't know, a postmortem. It's like this fun, well, fun, if you enjoy reading about train wrecks, it's this fun postmortem. And basically, there was this known firmware problem in a storage array they were using. And I think they had sort of been notified like a couple of times. But more importantly, like any large organization, especially in the government, they had outsourced it to someone else. And for some reason, the outsourcers hadn't updated it. So, you know, you should patch these things. I think it was in some sort of deadlock situation. But it is also, you know, on this little digression here, as I was tweeting about where one puts this kind of commentary, there's kind of a depressing, blameful postmortem in it, where several times they're basically just like, here's the people who screwed up, and we should punish them by extracting cash from it. And I don't know, it's kind of a read, a good read of like, how not to do things. Not even the patching part, but there must be something a little better that the analysis they could have done. So, you know, it's good to make sure you patch things. Yeah. So some other things that we had in the news were, we mentioned it last week, but Amazon shipped a beta of their new service brokers. They built one that helps you connect your PCF apps to, I think it's 18 different Amazon services from machine learning to data services, things like that. So it was a great partnership. They were out there in force as well, SpringOne, supporting that and talking about that. So between PKS and the service brokers, some good stories for Amazon customers using Pivotal technology. Yeah, and then also, we may have other partners that I'm excluding in this area. But because I'm at the FutureStack thing, New Relics Conference, they had a little, I don't know, thing they released last week that's basically, you know, if you're familiar with Pivotal stuff, we have this concept of the five S's, which are a way of grouping together the sort of performance and business metrics that, you know, you could call them KPIs.

But the numbers that you're interested in, why you would redo things a pivotal way and what you would track afterwards. And I, well, I will embarrass myself by saying why I'm not going to embarrass myself. I don't really have them all memorized, but they all start with S. Maybe you know them better than I do. But anyways, I'll let you skirt by by not making you recite them. But there's a blog post, which I'll link to. And New Relic did an interesting, they have an interesting, I don't know what they call it, a package. But basically they have five dashboards and I think a roll-up dashboard that's oriented around metricing and reporting on those five S's, which is great. It's a good reference of how you would implement these things. And of course, visuals, which I am planning on stealing from my presentation tomorrow to go over briefly in addition to our own methodology. And then finally, as people, a few people kindly told me, we're going to have a conference in Austin next year on October 7th and 10th, which is exciting. You know, as often happens, I no longer live there. So, you know, it won't be a great advantage, but I'll get to go visit my storage lockers. I think I'll just sort of like open them up and maybe lie down and just like feel my stuff, see how everything's going there in storage locker land. I had a little bit of a panic. I didn't quite remember where I put the keys for my storage lockers. But then like a good boy, I had put it in my backpack, which is sort of the equivalent of like the purse that I never lose. So it's in there. I was smart. I knew I would lose it. So I put it in the place I knew would always exist. Always thinking ahead. Well, what else did you see happening last week? Yeah, it was a big news week in general last week. So it was tough to keep up with everything. But Microsoft had their giant conference, actually flew down Thursday after SpringOne and went and spoke at that one on Friday. I saw that. Yeah. Yeah, because I make good choices. But that was they made a number of announcements.

The ones that caught my eye, the one I pointed out here in our show notes was their Cosmos DB. This is their planet scale, NoSQL database. They made some cool updates to that, actually adding multi master, meaning you could have this database striped across, let's say, five or six geographies. And each one is actually a master to read and write. So you're not just having the sort of lag of like, well, we only write in the US and it eventually replicates to Europe. No, I could still write it in Europe and it's actually still considered the master. So some really neat stuff there from a database perspective. And then they added an SLA that's actually five nines for read and write, which I don't know if I've seen that in a cloud database in terms of that much availability for both read and write. So just doing some good stuff in Microsoft land, if you're looking at distributed databases, you know, PCF plug, if you use the Azure service broker, you can connect that easily from your apps, .NET or Java or whatever, from anything running a PCF, which is pretty cool. You know, on the topic of cloud databases, which comes up every now and then in our literal view, you know how people say that like releasing to production should be boring, by which they mean you should automate as much of it as possible with not only automation stuff, whether that be a platform or other things. But then also standardize on your practices and standardize to the extent that maybe you don't even think that they are – that you're having to choose something. Like I was reading some – I forget what it was, but I was rereading some old like, you know, how to do software better stuff. I think it was in the Puppet DevOps report or maybe it was the Dora one. And they were saying, you know, it's okay to be sort of lockstep about things you should do because that's what a best practice is, right? Like if you develop a best practice, there shouldn't be any argument about like, for example, we should probably just not write a file system and use the file system that comes with it. Anyways, with all of the sort of cloud databases out there, do you think we either are or will quickly approach the point where like databases are no longer like hard or difficult?

Like in the same way that in theory, like deploying to, you know, infrastructure isn't that difficult. And I ask because it seems like databases have always been like this mystery and this like difficult thing. But maybe if you remove a lot of the tension of having to stand them up and manage them, then they're not such a big deal. I don't know. Yeah, I mean, it's interesting thought there. I think there's two parts to it though. I guess there's the complexity of managing them and using them, you know, standing them up, maintaining them, patching them, which has clearly gotten simpler in a cloud world. I mean, you and I can go provision a global scale database in the next three minutes. Like that's amazing, right? It's automatically patched and it's replicated and backed up. Like those sort of traditional DBA tasks in terms of operations have largely started to disappear and even on premises. Now you still need certain things. You still may need an index. You still may need, you know, certain components that you operate. But that sort of stuff has gotten simpler. I wonder on the app dev side though, like is that still, you know, it still may be tricky on what's the shape of my data. Should I write a stored procedure or put my stuff in code instead? Or, you know, how do I change a data field and go from full name to split it to two fields to first name, last name? How do I refactor that in my data? Like that stuff's still not trivial. There's patterns for it. But I haven't seen that get easy yet, if you will. No, yeah, that's a good cutting through all my babbling there. Because that was kind of like the analogy I was thinking of is like you could say like with all your cloud stuff that doing all of your, as I would call it, sort of like more lock level like management and operations and release management. You can remove that from the overall like software delivery pie of suffering, right? So you can kind of like, you can kind of like take that effort out so you can spend effort on other things. And then, however, it's still difficult to like design good software and make good product, right? Like there's no, I don't know, we don't have a build pack for that.

That's just like pops out good software on the other end. And it seems like in a similar way, like just as you were saying, as there are more public cloud databases, or I guess like managed down in the private cloud. Let's just say databases that are managed in a cloud way, wherever they are on the side of the firewall, like it removes all of that nonsense from our pie of suffering. And then you can worry about like schema design. I don't know. Whatever the equivalent of design is in a database, and you can spend much more effort on that. But, you know, I'm not a database person. We should get those data cow people or like, you remember, you ever read that, all the agile database stuff that Scott Ambler wrote? That was kind of fun. Yeah, I attended a talk of his back at the agile conference this summer, and he was talking actually about refactoring databases and not using that as an excuse to slow down velocity. So there's a good movement at hand, I think, to make sure we demystify kind of agile data as part of the pipeline. All that automation you just mentioned, like it doesn't matter if I automate all the app delivery if I'm still waiting for a DBA team to run a script to update a field. Like then I might as well not automate the rest of it. What's the point? Exactly. I think that's still my constraint. So, yeah, it's good to see more and more about that, although I don't see a ton of simple patterns always evangelize. We could probably do better there. Yeah, you know, I think his book on dad or disciplined agile deliveries, the only book I've seen where there's mind maps sort of printed as the overview, which is nice. Speaking of, how's your book going? The first draft I delivered the Friday before SpringOne. Wow, what a machine. I'm a monster. Luckily, it's one of those smaller O'Reilly books. So I'm not taking credit for like a 500 page monster. It's like 100 pages. Yeah, that's good. That's good. Okay. But yeah, so I'm waiting for the first brutal reviews back. And then I got to, of course, make like 30 updates because technology keeps changing. Thanks a lot, technology. Yeah. Brutal updates. That's always exciting. Well, you were suggesting, rightly so, that we should give a summary other than sort of the releases of like some of our highlights and, as they say, takeaways from SpringOne platform last week.

Now, I didn't do, despite being here at the FutureStack thing, I didn't attend like the partner award thing or sort of conference. Although, I did get a challenge code from Nima, our partner person, a challenge coin. So I'll have to do something with that. It seems like a good paperweight, but I'm not sure a situation is ever going to come up where I need to pull out a challenge coin so I can sort of pass muster. We'll see. I should have brought it to this event, I guess. I've failed. Come on. But how? Yeah, I can bet it's not in your man purse backpack there. There you go. Fire alarm. Fire alarm. So while I mute this, why don't you tell me the partner's part? Sure. Yeah, so this SpringOne, we had our first ever partner awards. So we gave out a couple dozen or so awards to different partners. You know, SI partner of the year, do the right thing. A number of things to just demonstrate things that have worked well with our partnerships, kind of customer impact and the like. So we highlighted a number of those, sent out a press release, put together some great material. And it seemed to be well received. We have an amazing partner ecosystem at this point of system integrators, of software vendors. And so it was good to highlight those. We had a partner off-site on the first day of the conference, or I guess day zero on Monday, just bursting at the seams, kind of working together. How do we help these partners accelerate what they're doing with us? How do we learn what they need from Pivotal and from customers? So that was extremely successful. It was rewarding to see all the hard work that our partner team puts into building a great ecosystem, kind of come to fruition at the conference. The show floor was bustling with all the partner booths who seemed to be packed at a lot of times. And it was, you know, I walked around, I think the last day of the conference, just to kind of shake hands and talk to each person, find out what they had kind of taken from the conference, each of the partners at the booth. So it was just a good partner show, and I think it was probably the best we've done so far and a good indicator that this thing might actually stick. That's right. Well, you know, the conference is actually, I guess, did it start on Saturday?

I forget, or was it on Monday that it started? Pretend started on Monday, obviously. That's right. Pre-team got there Friday. But yeah, Sunday was there. Monday was the real first day. And, you know, sometimes there's like workshops on Sunday or workshop days. But on Monday, there was an analyst day, which I have to admit, I did not attend all of. But the two sections that I made sure to attend were the customers talking, which is always interesting. And then the, you know, the executive Q&A is always fun to see what analysts ask. And then I guess also what my coworkers answer. That's fun. But I think, you know, some of the things are under NDA and not in an analyst event. But later on, there was a keynote by one of them, the Rabobank guy. You know, now I'm much more familiar, a large Dutch bank. And, you know, I think what was interesting about his keynote and what he was talking about in the analyst day was the sort of like general cultural sort of path that they went down to have things to get pivotal Cloud Foundry in there. Like his his his his posing of it was that essentially they were convinced of like doing Agile and DevOps and they kind of had that under control. And I don't know if it's unlike a lot of people that we work with, but it's certainly not maybe the majority of people, at least that I know of. But is very they were very interested in just looking at pivotal Cloud Foundry as just like the tool they would use to accomplish this, which which I think was an interesting pull in that they were doing. And also, you know, I'll have to I'll have to go look at look at his his keynote again to make sure it matches the tone that he had in the analyst day. But I would say if they're you know, there are a few better representations of like the Dutch humor I've encountered than the one he was doing. He's very like, very not like dry, like English people, but he just sort of like marches through some jokes without necessarily waiting for you to catch up.

So you got to be pretty quick that like, oh, he's joking right now. And there were some good zingers in there. That's funny. Yeah, I thought he had a great talk. And it was you mentioned this elsewhere, too. Like, again, this idea of just tools, I think I just heard over and over again, the U.S. Air Force did this, a bunch of others of just the sort of like we had to do a few things to make this successful. This wasn't just any product. And like, although that like no one product fixes anything, that's insane. No one platform. And it was that way of working sort of stuff. I actually had jotted down. I was reading transcripts today and there was a Boeing discussion on kind of moving fast and automation. One of the panelists said, you know, quoting here, the hardest part about the transformation was actually the fact that the users we were interviewing couldn't comprehend a world of actual automation and self-service. Like it was so beyond there to think, wait, you're going to do all this stuff like you had to even just sell this idea of working in a different way, ignoring the technology. So I thought that was always a good undercurrent, that this is a complimentary set of pieces of how do you run technology well? How do you rethink how maybe you develop software and do product design stuff and automate the path to production? And yes, there's tools in tech that makes a ton of sense there. But if you don't come in it with the right mindset or have the right empathy for your customers, it's all just going to be disappointing shelfware. And it was nice to hear that. Yeah, yeah. I mean, that was a theme that came up here and there was the, to put it in an extreme way, like once you get things optimized as far as like not, like kind of we were talking about earlier with databases, analogously that you don't have to worry about automating this or releasing it or whatever. There's often a phase you go through where no one believes you. And so, and so you've got to like, you've got to like emphasize to them that it's sort of like actually real and, and these things happen, which I think, you know, you, you, you mentioned the Air Force. I mean, they had, they had a pretty good update to their, their ongoing story. What, what, what, what, what did you think about their talk? I thought that was, you know, interesting, inspiring.

It's fun to see how they're trying to improve, you know, again, having them on stage talking about the main stage, talking about their progress and how well they're shipping software. And for some reason, I don't usually take pleasure in the pain of others. Like that's not my, my thing, but it was during that talk, it was fun to see somebody tweak. I'm like, oh my gosh, like the Air Force ships more often than we do. Like we suck. And I like that. Like you're looking at these companies going, wow, like they're doing this. And I think it was also this reminder that better doesn't equal good. Like I could be better than I was last year. I could be now shipping every six months instead of every 12 months, but I need those reminders that that might not be good yet. I need to look at what good is. And in this case, Air Force was saying, look, here's our journey. We're now saving half a million dollars a day on fuel because of this fuel tanker improvement thing. And hey, we're shipping code so often now. And then I went to one of the U.S. Air Force breakouts where they were actually putting this in so many PCF and so many remote sites and doing this faster than they'd ever done. But again, it was still this journey of like, OK, now we have to operate this thing. What does good look like from a day two perspective, not just getting it up and declaring victory? So their whole thing at the show, I thought, was fascinating as they talked about what they were trying to do, how they were using software to get better, what that actually meant to their own people, and how excited their own people were about doing some of this work and getting better at software and not fighting bureaucracy. So I thought the Air Force team did a really nice job putting their work into perspective as well as what the outcomes were. They weren't done, but they were showing a nice progress report. Yeah, I particularly like that because of, I don't know, for numerous reasons other than it's a good idea, like have studied the Air Force case quite a bit, like finding all the little bits I can and tracking their projects over the past couple of years. And so, one, it was good just to get an update of what the numbers are and things like that. And then, as you pointed out, so the original number, essentially, you know, the most famous one they have is this, like, Project Jigsaw, which, of course, is all capitalized and probably stands for something.

They also have, like, great names like Raven and Marauder and these other things. Marauder is one of the underappreciated words in the English language, I think. Like, do you remember when there was a muscle car in that tragic period in the late 90s when all American cars looked terrible called the Marauder? It was basically, like, kind of like a Monte Carlo that was, like, carved down. It was quite the automobile. Anyhow, they were going over, originally, when their tanker planner thing came out, just a plan where tankers fly around to refuel jets. You know, the number was, like, they were saving $284,000 a day in fuel because they optimized the analytics behind it so much because it was no longer a whiteboard. It was software that they didn't have to put an extra tanker up to account for errors in the analog computation that was happening. And then, like you pointed out, I think there was further, I'm assuming, further features that they added and sort of, like, an evolution of the problem where they, I can't do the math in my head, but they around doubled the amount of savings that they had. And if I remember, they also, like, reduced the amount of, like, operators or people needed to run the program. So maybe they talk about that longer more in the session that I didn't attend. But I think that kind of story, that sort of year-over-year thing is useful for looking at, that it's not just sort of, you know, this is the other side of the tools thing. That installing, like, a Pivotal Cloud Foundry or whatever tool you might have, it's not just sort of, like, a one-time boost that you get. But it really puts in place this thing, as in the case with the Air Force, where every week you can, like, release a feature or two and study it and, slowly is the wrong word. But to use the idiom, slowly but surely improve the business process that you're applying it to. Yeah, I love, in one of those breakouts, what they had mentioned is, it was funny the impact that had on culture. Like, first of all, yes, that they weren't just big banging, releasing something, all moving on to something else, that software's always just shipped.

It's never done. But what they know, this is that when their other business clients, if you will, I guess there's business clients in the Air Force, that when there's other clients, that when they saw that software was just kind of constantly iterated on, they would actually call up and pull over an airman or an engineer going, like, hey, this isn't working. Can you sit with me? Can I show you this? Because they know it's actually going to get fixed. It's not like, let me just send it into some faceless help desk where maybe four releases from now in three years, it'll get fixed. It was, hey, let me call you over or let me give you a feature idea. Because, again, it was this constant just sense that I will give you new ideas, and those new ideas, if they're good ones, will get reflected fairly quickly here. And that was completely new to them, but it created actually a new partnership where it wasn't just to get IT in the business, which are terms I hate. But it was the idea of, like, we're all in on this, right? We're all trying to make this better, and you've proven to me, like, you've proven that you can take my input and turn that into software. I thought that was really neat. It was a cool switch to see them flip. Yeah, yeah, yeah. I mean, I think, you know, there's, building out that there's two other things I noticed a lot in the discussion with, let's call them end users or not pivotal people. At the conference, one of them was, you know, you're kind of alluding to this, is that one of the needs, and the Air Force is, in the talks I give, are always very clear about this because of the, I think it's because of the comp structure that they have and the hiring that they have to do. And there was some discussion from one of the civilian heads who was speaking on their behalf. But basically, hiring is really hard. Like, people who do technical things, you know, the kind of people you would want for programming and operation stuff, it's hard to not only get them, but because of various weird reasons, it's hard to retain them. Even retaining them beyond just wanting to be in the job, but you can only, like, work for six months or something. It was one of these situations where it was like, you know, you can't fight City Hall or Congress in this case, just some weird regulation.

But this similar thing came up, you know, whether it was people talking at the analyst day or just customers I would talk with, you know, over lunch or in the hallway, which was, and I mentioned this because I don't take it too seriously, is like, we got to figure out how to, like, hire and retain people. Like, that's really difficult. And so part of the joy, I guess, that comes with not doing more Lockheed and Drudge work, the toil, that's one way of making people happier, to retain them and then kind of tell them when we're trying to recruit them. But also, just showing up to events like this and talking about what we're doing and how we're helping people out is a good recruiting mechanism as well. I mean, everyone laughs in a keynote when they do this, but, you know, of course they do it. They're like, we're hiring. You should come work for us. And I think that, you know, the other thing that I noticed that kind of you see in the Air Force example and that came up with a lot of the talks and the conversations that I had with people is basically, I mean, I don't know. This is kind of an obvious thing, but talking with people made me realize it a lot more is a lot of what these organizations are doing is basically becoming a software product-centric company, which I think, you know, forever we've been talking about how, you know, everyone's a software company or whatever, which I'm sure I've said that here. It always, like, mentally, like, grinds in my head because I'm like, I feel like unless you're selling software as your product, you're not a software company. Right, exactly. In the same way that, like, you know, if you're delivering packages, maybe you're not a transport company. I don't know. Or I guess the equivalent would be like, you know, we realize that we're an experienced delivery mechanism, but really you're just like the catering for airlines or something, which nothing wrong with catering for airlines. It's just like maybe you're just catering for airlines. So anyways, but I think the kind of toned down version of that is a lot of what people are talking through is like, basically, we need to become a product company, which has two effects.

One, you know, we were kind of stumbling through this with like, are there business people in the Air Force or whatever? The phrasing for this always gets fraught when you cross between for-profit enterprises and mission enterprises. I don't know how to categorize public sector. Um, as I just proved, it's hard to use the words. Anyways, uh, but you know, what you're getting at in both of those organizations is that we're delivering a product. Uh, and so we should do things in a design oriented way and all the usual agile stuff we talk about. And one of the things, uh, I was talking with, um, a couple of people about, um, is that also on the other side of it, like if we go up to the mythical C suite or, you know, the, the, the agile side, like we need to get to the point where they understand, um, the tools that are available to them to change their business. And there's a couple of examples you could think of, like in retail, there's, uh, probably like omni-channel stuff you should look at. Right. So on the non-IT side of the house, people should think about like, well, what if we had a mobile phone to buy groceries or whatever, which people do, but it's just a simple example. Um, but then I think, I think the, uh, you know, the analogous thing I was thinking of was like, probably at some point, like maybe, I don't know when aerosol cans were invented, but maybe in the twenties or something, I'm sure all the way up to like in the shaving care industry, like up to the board level, there was probably some presentation where they were like this new technology called aerosol spray cans. And, you know, eventually we should get to that point where, uh, the sort of like leadership in a company is understanding somehow this, this, what this technology can do for them as far as like doing weekly releases and doing better design. And, uh, hopefully over the next few years, that will be sort of like a, uh, I don't know, a more pervasive thing than we have currently. Cause we're still kind of like, you know, the IT more locks, uh, it seems like to the vast majority of people. Yeah, that's good.

Yeah. Speaking of the IT more locks, I was, uh, as you think of the other sort of things that we hope become ubiquitous all the way up to that thread, I was reading again, a transcript this morning from the Wells Fargo presentation. They, it's great to see all these new customers on stage, StubHub and DBS bank and Wells Fargo and Dick's sporting goods and, and others, but the Wells Fargo team was talking about repaving. And I'd like that that's starting to become more of a thing. And so they're saying, look, we've got 15, 18,000 VMs in our overall, all our PCF environments. Our mandate by the end of next year is to repave daily. We're doing it now every two to three days. So they're just talking about how do we rebuild our entire environments every day with gold images, always patched, all up to date, no downtime. And just making that like a thing saying like, Hey, if we want to be secure, that means having a much more aggressive approach to how we build infrastructure and rebuild it. Like that's just so neat to me that something that we've talked about for a bit, hopefully that keeps moving up the chain so that these maybe executive suite folks go, why are we rebuilding VMs every day? Like that seems like a lot of work. Like, no, no, this is what we're getting. Automation makes it possible. And this is sort of security profile and results that have come from that. And like, that's a technology thing that someone even who might be a line of business lead goes, yeah, I get that. How do I get more of that? So I like that there's some of those even technical concepts are moving up into leadership teams to other people. And I think that's pretty inspiring stuff. Yeah. I mean, I think security is a good example of what I was grasping at to have the, you know, move beyond more locks knowing about stuff, which is, you know, I think most people in a company, especially the leadership, whether they're technical or not, know that security is important and it's something they need to worry about. And unfortunately, I think their knowledge is motivated by like negative forces, like, like it's something that we need to survive. And, you know, maybe it would be great if what we in the software world could offer was a more like positive force. Like here's, here's something you don't do because you will get fired if you don't, but here's something you can do that's like a positive growth mechanism that you can take advantage of.

So, so we'll see what goes on there. And then speaking of that was my, my, my keynote thing to pick the, I need to go back and rewatch it, not to look up if there was extended, an extended session. But I liked the, I don't know, maybe the last, I think second to last penultimate for those who like dictionaries slide that the Dick Sporting Good guy had. And it had the usual sort of like, and I mean usual in a good way, usual sort of like things we learned from our digital transformation or whatnot. But there, there were, there were two at the bottom that were, I think they were things that you normally wouldn't see people admit to however many hundred thousand or so people, hundreds of thousands of people who are in the room and will watch the videos. And the first one was, uh, you know, the cultural change of, of, of changing the skills and the norms and the structure. I mean, he was pretty blunt. He was like, we're going to lose some people. Like, they're just not going to like it. And they're either going to like sandbag and like not go along with it or they'll quit. And, and that was, um, that's usually not something that people like to admit. But I think, I think it's helpful to say these things because it's, uh, it's sort of confirming and soothing that like, if you're trying to do this and you're having difficulty retaining people, like that's totally normal. Like that's unfortunately going to happen. It's not because you're necessarily doing something terribly wrong. It's just that, that shift is difficult. And then, and then the second thing he said, which I mean, I, I, this is, this is remarkable to me because I don't know if I've ever seen someone say this is like, oh yeah. And then we're light on design talent, which, which is another thing that I imagine given, uh, that most people are not product centric around aerosol spray cans, uh, metaphorically speaking. Yeah, probably designers are something that most organizations don't have. And so it's a, a, um, uh, a function that you'll need to be on the lookout for, for hiring around, uh, not the aerosol spray can engineers, but the software design people.

But he had a, uh, he had a good and, and again, sort of, uh, Frank talk. And also he was hiring as I recall. Of course he was. Yeah. Again, I thought Vincent did a great talk, the Robo Bank guy. I thought that was, that was good stuff. Were there other main stage talks that jumped out at you? I know for me, I, uh, my first get was Neha from Confluent back, I think in February I asked her and it was a leap of faith for her. So I, it really meant a lot to me that she said, yes, probably just unwittingly. Like this seems interesting, but we weren't even really partners yet with Confluent, the creators of Kafka, but you know, our Neha is one of the co-creators of Kafka, but I thought her talk was just outstanding. It was a good walkthrough of like why events matter, how to think of events in the data-driven world as you start to almost turn the database inside out with some of these event streaming platforms. So I saw it was a really great talk. She just, just nailed it. And so it was one of my favorite talks of the show. And my other technical talk I really liked, I mean, all the spring team always does a great job. You can just tell how much they prepare for this. And that, again, I respect that a lot. Michael Manella had the unenviable spot of like closing the show and talking about batch and like, and I'd asked that team, I said, look, give me kind of an oldie, kind of golden oldies talk for the last day. Like help us remember something that's actually really cool in spring, not just always like latest and greatest, sometimes just like what's good, like what's already mature and awesome. And so I just thought he really did a wonderful job setting up why batch still matters. Here's an example of using it with things like data flow, even a newer technology. But so, I mean, that was one of my favorite technical talks. I thought Michael really took a tough slot and did something really interesting with it. Yeah, yeah. And, and yeah, I mean, I think to, to, to go back to the, the Kafka talk as it were. Yeah, I thought, I thought she did a, a great job presenting it because it was, it had, it had that always difficult to achieve mix of like, it had some interesting new like ideas about databases. And I guess Kafka people get upset if I called it a database, but whatever your, your, your data, your place where data is that you based this, your data on, uh, you know, like going over kind of, as you were saying, like how, how things change and why events become important.

And then, and then, and then the other thing, again, that most technical people don't always has is man, she was like super confident. It almost like was part of the, the, uh, the content on its own. Uh, and you know, most technical people, including myself are just sort of like, eh, take it or leave it. But, uh, afterwards, since I got, I was like, man, I got to use this stuff because it's very important and it works. So that, that, that, that was a good talk. It was, it was, uh, it was nice to have like some data talk up there from, from someone who was, uh, outside of, of our overall community. Well, so, uh, you know, we usually put these videos up very rapidly. When, when do you think, uh, we've got some keynote ones up, uh, that you can go see. I think, I think the, the Dick's one is up there and probably all of the, all of the keynotes, but when, uh, what's, what's, what's the trickle strategy that we're going to have? Yeah, we've worked with our, uh, our vendor to actually render these super fast. So yeah, all the keynotes were online, like by the end of the day of the keynotes. And then I think all the breakouts should be online this week, this, you know, this last week of our first week of October, I guess we're in. So people should be able to watch those. Um, you'll probably see us starting to recap some of these in different blog posts as well, as I don't expect you to watch a couple of hundred sessions. Unless you have a lot of free time. So we'll try to do our best to, uh, to summarize some of the greatest hits across some of these different, different talks. But I know I, when I look at the schedule, I was like, I maybe watch 5% of the show. Like when I look at all the breakouts, all this stuff. So I'll be immersing myself in a lot of stuff just to just kind of hear what customers are talking about here. A lot of the engineering talks and some of the things coming in, you know, this is still not just customers talking. This is still how engineers explain what we're building and what matters and what's next. So a really good mix between the two. So tons of stuff to, to watch. And then hopefully people attend SpringOne tour for kind of that little version throughout the year until, uh, next year in Austin. Yeah, yeah. No, I, I, I, I agree. There's a lot of them and it's hard to see everything, you know, and, uh, I need to redouble my efforts. I mean, given, given what my job is and what I do and things I'm interested in, uh, I need to try to actually watch all of them.

Cause I, I pull a tremendous amount of content from things I talk about from these, uh, these SpringOne platform videos and some CF summit videos here and there. But I mean, as an example of how difficult it is, I was, uh, there were, there were some, there were a couple of people from, uh, Mercedes bins.io sort of like a, uh, consumer facing kind of browsing cars and scheduling things. You know, part of, part of, as we would say, the funnel of getting a car. And I think some, I think some post, uh, buying activities as well, but, um, I was, uh, I'm going to have a podcast with them sometime. I don't know, in the next couple of weeks or time going over that. And, uh, I went back and watched their 2017 podcast from the presentation from SpringOne platform, which I hadn't seen. And it was awesome. And it, you know, it reminded me that like, there's a, there's an endless well of these things. But I think, I think if you're interested in, uh, whatever digital transformation is and agile and DevOps, it's, it's well worth your time to allocate some time to sort of at least find three or four of those that tell these extended stories and kind of extract out, uh, some, what would you call them? Anecdotes as helpful tips, uh, that people have gone through. And then I'll, I'll have to get those transcripts from you and then, uh, I don't, I don't have to make transcripts on my own. I know I've got a secret drive, which I may share with you. Secret drive. Yes. I'll, I'll have to, I'll look up in my little book, how to ask things nicely. Then I will also endorse you for that. Oh, can I get skilled and asking things nicely? Well, uh, you know, if you've listened to the past few episodes, I had, uh, I had a interview with, uh, with a couple of people with Kyle, uh, who used to be at CSAA. And now works at Smarsh and then also, uh, Thomas from, uh, West. And, uh, this idea of product thinking and what culture is came up in both of those conversations, which I'm sure a listener you have listened to in detail. Uh, but those, those are actually fun episodes to do there, uh, on the, in the lobby of the hotel.

Well, I think there was some Gwen Stefani as the soundtrack when Thomas and I were talking. Um, and I think there's some follow-up interviews that I was, I will have like with the, uh, Mercedes Benz people. But, uh, if there, if there's any of the talks that you, the listeners saw, and you think that would be good to have those people on here for extended conversations, you should send them to me. And then our sister podcast, Pivotal Insights, we'll beat them to talking to customers that, that are interesting. They're always much better at doing that. So I'm sure they'll have some people on as well. And, uh, as always, thanks for listening. This has been Pivotal Conversations. If you want to get the most recent episodes and you're thinking, oh, I should go listen to, uh, those old episodes. Cause, uh, I was a teenager in the nineties and I love Gwen Stefani or, you know, or the early two thousands. Uh, you can go to soundcloud.com slash Pivotal Conversations, and they're all listed there. And then we post the show notes about every Thursday on pivotal.io slash, uh, podcasts. Um, and, uh, I'll put a link to many of the things we mentioned there and maybe things that we didn't. And with that, we'll see everyone next time. Bye-bye.

Tanzu Talk · 🍕 Get more pizza.