Techstrong Gang – November 15, 2024

Hey everyone. We're still a cube con and you are still not, you're watching Text on Gang. Hey everyone, it's Alan Shimel and we're here at Q Con wrapping up a full week of open source Cloud native Kubernetes Joy, or maybe it's Joy.

We're gonna discuss that. Let me tell you who we've got to discuss it. We've got some of our core gang members who've been here all week.

Let me introduce you to them. Well, you probably know most of 'em, but if you watch the text on gang at all on my far left, he's usually high up in Colorado. Don't, don't put the pause in that so much.

He's at a higher elevation here in Salt Lake City. But anyway, it's our, uh, fu of VP of research for DevOps and some other stuff. I think even maybe cloud native Mitch Ashley.

Hey Mitchell, welcome. Great to be here. Great to be on with the game joining us.

He's usually in Texas, but he's also here in Salt Lake City. I came to find out his family lived here in the fifties. So he's he's he's Heritage.

Yeah. I wasn't in that family yet, but yes. Oh, okay.

Well, I was gonna say you were like a legacy now. So from now on, he'll be known as Flounder. No, but Guy Currier Guy.

Welcome. Thanks. Good to be here as always.

And on my right, taking time out of his busy schedule, scouting minor league ball players for the Yankees here out west, our chief content officer, Mike Vizard. Mike, welcome. Happy to be Back here.

First time in 25 years, 20 years for me. Yeah. Yeah.

We were talking about it. So let me put forth a proposition. What, four scoring?

No, not that one. But, and, and then I'm interested in your takes on it. You know, they, they estimate there's 9,000 people at CubeCon.

Mm-Hmm. I estimate 1500 of them are, are working at the 180 booths that are here on the floor. So there's probably 7,500 people attending Salt Lake City Q Con.

Not a number to sneeze at, but still not as big as it was pre Covid North America. Not as big as it is in America. You've got your keynotes, you've got 24 tracks going on sessions.

You've got, you've got an expo hall that goes on forever in every direction that you can look at as far as the eye can see. Is it too much? Are people enjoying themselves?

Are vendors getting their due? Who are paying good money to sponsor? Is their learning going on?

Is it, is it too overwhelming? You, I was telling do we need to reorganize Con and if we're gonna reorganize Cube con, do we need to reorganize the whole sort of open source community way of doing things? I've got a lot of thoughts On this CS there or something.

Well, I don't know if you need to reboot it, but do they need to look at rebooting these, this event? Hmm. And how we organize this open source thrive.

It is a thriving community. Yeah. But is it ever going to get to that next place?

Well, just to jump in, if, if you, you know, the number of people that have contributed to, um, OpenTelemetry and also the Kubernetes, you know, a lot of those people are here. Not, not all of them, of course, but over time, that community has grown up. I think that's a big part of what's here, is people that are contributing to open source.

So it's as well as the consumers of it. Okay. 2% of a community or less contributes to open source.

So you're telling me this whole show is for those 2%? That's not, no, No, I don't think that's all who's there. Because the other part of it is, I mean, I spend most of my time talking to the vendor community, not going to keynotes and, you know, developer meetings.

And they're, maybe they're just being positive, but I, they seem really happy with the show and they're getting good leads out of it and connections. And so there are, it's not the looky-loos like we used to see at RSA, like, I'd like to get into security. Which Husky do you have?

You know, There are other mitigating factors in work here. One is, you know, as much as I like Salt Lake City, the reason they're not at their pre covid numbers is, you know, they went to places like Salt Lake City, Chicago, and Detroit. And so they're not really gonna pull that kind of people that you would get into Vegas or Orlando and whatever else.

So that's part of the issue. The other thing is, we've talked to some of the vendors here. They didn't even bring their sales teams.

They're here to kind of do some bonding with some end users of their technologies. And they're hoping that that build a relationship and maybe sell 'em later on something they can upsell. Because the open source business model is much more of a, you know, let's get everybody in the community and then start peeling people off to see if we can get them to buy something later, kind of thing.

That model, whether it works or not, is debatable. Well, that's the key word there, isn't it? Community.

That's just what I was thinking about. And I mean, you'd have probably better perspective than I would, but, um, the open source has been about community from the beginning. It's also been about commercially taking advantage of community.

And I think what, what you're highlighting is that kind of imbalance because the vendors didn't come here for free. They paid for it. And I have the feeling that Linux Foundation, cloud Native Computing Foundation sold as many booths as they could possibly sell.

So there's a huge demand from the vendor side to be part of this community. Mm-Hmm. Because cloud native in particular, Kubernetes in particular is clearly at a market peak right now.

And, and I I could make an argument. I agree with what you said. I could make an argument that Dev Re is now expanding into platform engineering.

There are companies here saying, I have to figure out my dere strategy platform Engineers. What debre Well, every Dev Rel person I know got laid off over the last few Years. I'm just telling you what I heard.

So, Yeah, no, I'm, but yeah, I'm telling you, real life. We know a lot of Dere people. Well, And it may that that's changed what we thought of derel two years ago or a year and a half ago.

Maybe that's not what Derel looks like now. I don't know. Maybe the Debs are not as relevant as they used to be in terms of making those platform decisions.

Yes. Distributed. We've been talking yeah.

About this pendulum swing on a lot of our engagements that the grand pendulum swing that we've been experiencing since the days of the mainframe between everything is in the hardware and in the infrastructure. That's what the developers need to code to, to developers need to be served by the infrastructure side. And, and it's the software development that should be the center of gravity.

I think we've experienced five, six years now with containerization, Kubernetes, all this focus on infrastructure and cloud. We've experienced a good five years now of very infrastructure centric development in the industry. And maybe, and I think that led to the kinda layoffs and the kind of concern around, uh, investing in the developer community.

But it's just gonna swing back, isn't it? It's just gonna swing back. Look, I've been involved in open source for 25, 30 years.

So have you fundamentally, there was always a question in the open source community, what's a successful business model? And what did people tell you? Look at Red Hat.

Look at Red Hat. Red Hat. The, the poster child of open source success.

IBM paid, what was it? Whatever, $12 billion Red Hat changed their licensing on their flagship product. What does that tell you?

HashiCorp probably one of the most successful open source companies out here. Change their licensing. Yes, we forked Open Tofu, but HashiCorp still sold maybe less billions than it maybe would have.

I'm telling you that fundamentally these companies give you that tappy talk or where, from George Orwell in 1984 about, we don't want salespeople here. We want a kumbaya with the community. B******t, b******t.

They wanna sell product. Are, Are you suggesting that open source maybe is an unsustainable welfare state? I, um, I think you said they're at the peak.

You know, pat Riley was one of the greatest basketball coaches in the world in history. Great teams at Los Angeles. The Knicks, the Heat.

Why did he go to all those different teams rather than staying like red aback at the Celtics or John Wooden at UCLA? Because you can only stay on that peak so long before you fall off. And I think, and I'll call it, I think the last five years have been a, a golden era.

Maybe the last 10 years even have been a golden era of open source where we saw extreme experimentation on open source models. We had open core, we had foundational open source where foundation supported these things with multiple companies, you know, participating where we, we invested a lot of people who wore funny hats and went around with purple hair talking about Kumbaya and delivering delight in their software. And that would magically lead to dollars at the end of that rainbow.

And here's what happened. Companies had to change their licensing. They found out that those people in the funny had some purple hair, didn't necessarily translate into paying customers that the open core model does have commercial Mm-Hmm.

Application. But that the fundamental numbers haven't changed. Out of a hundred percent of people in the community, a handful, not even 2%, a handful actually contribute code.

If a KI 2% maybe will contribute other ways, they'll contribute bugs, test beta, do stuff like that. And in an open source community, if you get three to 4% of your users converted to paid customers, you gotta successful business. But you, it's very hard taking that to 10% or even 8%.

It's hard to present that to your startup board at the sea round and say, keep doing that. That's it. That's the issue.

And that's fine when it's all a cottage industry. There's no longer a cottage industry. What, and so it's a pretty big cottage if you Yeah.

Well yeah. It kind of like on Jekyll Island, if you've ever been there. But there's Another trend that's playing into this in a couple of conversations I had with some, um, a few of the senior IT folks who are here and they're basically saying because of, you know, incidents that I happened with Log four J, they're not buying anything or not allowing anything into production that doesn't have a sizable community around it.

'cause if it's only supported by a handful of people, they're like, that's a recipe for disaster. So they're starting to put some different guardrails in place for a lot of these projects. And a lot of these projects aren't gonna meet that metric and they're gonna be, you know, not gonna survive.

Probably becomes when those smaller projects that don't have the full support Become orphan Well, orphans. Yes, exactly. Or, and or they're still embedded within the things that do have a lot of support.

That's what happened with Lock four J. And I'll put forth, it's not just the smaller projects. I personally know of some very big projects that the companies who originally Oh sure.

Managed, hosted them, would love to get out from under 'cause they've moved on. Well, I'm, I'm getting confused. Is this podcast about open source or is it about CubeCon?

This is the text junk gang. We talk about whatever the hell we want. But I thought we were okay.

So we just veered off in, I mean, I'm kind of well, Well pulling the time out. Your argument is, is also we're saying that for all intents and purposes, the cloud native foundation now is the leadership open source, Open source community. This is where open source lives.

Yeah. Yeah. Okay.

That's what I'm Saying. Okay. Sure.

And I, I get the feeling, the funny feeling that there are other things that live here besides open source Money commercial. There are very Few commercialization money, if you like. There are Very few things that live here that are not open source.

Right. Or that are not, uh, dependent on open source or that don't contribute to open source. Or At least, at least, Least I Pretend to Be.

It just, you know, uh, the origin of open. We just come a a long way baby from the origins of open Source. Yes.

We have, you know, this is not the, The, I think your fundamental point here, Alan, is like, you know, this, this, you started off with this, uh, true statement. Just look around, look at the preponderance of vendors and products and pitches and messages and all of that other sort of stuff in comparison to practitioners and collaborators and contributors. That's the market space we're in right now.

We haven't mentioned the elephant that came in and trumped all over everything. Namely, it starts with an A and n. So then I, and doesn't have any letters in between.

I feel like, you know, that huge investment, that huge concrete global it boost is pla being played out here right around us right now in so many ways. I I think 80% of these companies on our Commercial we're, we're not immune, sorry. We're not immune in an open source community of collaborators trying to build, uh, uh, value through that collaboration.

We're not immune from that grand commercial Course. Use the ecosystem word, but I mean, that's what's around, Right? Yeah.

We're not immune. We're, We debated this a little bit and I think this is where you're going. But if AI becomes the new mechanism by which I engage with all software at some point, all those things in the backend converge because I'm just gonna call them and I'm not gonna care whether or not it was Prometheus, Kubernetes, or Grafana or whatever the hell else was on the backend.

It's just all gonna be one API call motion that mixes and matches as I go. So, you know, am I gonna have 175 projects or are they all really features of something? Yeah, exactly.

Is it all Not only projects, if you walk down here, you know, those 10 by 10 boots are fast and furious and there's probably, there's probably 120 of them over here. It's overwhelming actually. How many of them are viable companies?

I don't know. You tell me. I could count it on my hands and toes Mean I'm, there's you're Yeah.

Yeah. It's one outta 10. Yeah.

It's one outta 10 viable. I think it may be one outta 50, You know? Yeah.

What, what split the mill. Let's say it's one outta 25 or one outta whatever it is, right? Five.

And out of the those that are remaining, how many you're going going to be subsumed by some other brand? Well, They all get subsumed, right? Yeah.

But, But just to put a a point on this dot, right? So a couple years ago it was like, oh, well you gotta get to the developers 'cause they bring you in and they're the big powerful influencers, right? Not with the hot hotel road then.

Well, I don't think, I think that's fallen off. I think what we're seeing is a shift back towards platform engineering. Centralized it, standardized platforms, and those guys are gonna decide what the thing is.

So You telling, There's two martini lunches back. Maybe There's a shift back to the Monolith golf. What times our tea time?

There's a shift back to the monolith. I've heard that at least five times at this show. Have you?

Yeah, Yeah, yeah. Well, I've heard it, I've heard that as well with the nuance to it. Whereas the monolith has got some APIs wrapped around it and it's just a big ass microservice.

But, you know, other than that, it's the same thing. A Macro micro, a mega microservice. Like what is that?

There's a variation on that. What I hear is the target is enterprise, large enterprise. Yes.

Not mid-tier, not small. Mm-Hmm. It Is doubt Everybody's going after the scale.

Well, because, so here's the thing. What we've, you, you gotta look at this in, in connection to the macro world. The tech industry dollars of dried up from VCs.

VCs are saying, show me traction. Yeah. Show me traction.

Show me paying customers. And I think what a lot of these companies have discovered is they've gotta be in that enterprise market to make that kind of money. And So they're, they're who have the problems that appear at scale, right?

Versus this is new interesting technology Use. And all the tech startups that we're buying or using a lot of this open source software themselves have disappeared. Right.

Or, or their budgets are tight as well. Their budgets are tight as well. What's interesting is I don't see a lot of security people here.

Not as much as in past years. Yeah. I, And, and isn't that so, you know, you've got sonatype, you've got a few folks, folks that are in that area, but it doesn't have a strong presence as I thought, by the way.

I had one person say, I came here. I can't believe how much AI there is. I didn't expect that.

I'm like, where Have you been? You didn't expect the AI here? Yeah.

Where have you been? I don't know. I had, I had a weird conversation went like this.

It's like, yeah, all the things you guys are saying are absolutely correct, but I'm still building a monolith. I said, why? Because monolith is easier to build.

And if the monolith is successful, then I'm hoping to have all those other problems and I'll get to 'em later. There, there is that have Lifetime employment for all the problems I created in my monolith. Okay, Let me, let me just play devil's advocate 'cause we're almost outta time here.

The fact is, this is a huge Petri dish. All these companies here, the issue is yes, only maybe one outta 10, one outta 25, 1 outta 50 million fact make it. It's knowing which one Sure.

Yeah. Is is the hard part, right? And everyone here, you know, you could take 180 odd projects, but everyone takes a different view and uses it in a way maybe that hasn't been thought of before.

And, and it's almost like evolution right before your eyes. And, and so maybe new things are synthesized out of this, but my, I I, my closing argument is I'd like to see the CNCF reorganize cube con. I don't, I think the average person, that's too much for him to wrap their heads around or her to wrap their head around their, at this show.

I, and I don't know what the answer is doing micro shows or, or stuff Like that. Let me, let me take the counterpoint just to take The counterpoint. Go ahead.

Come on, Jane. Um, rather than a futurey dish, which I get, it's also the sorting hat, right? 'cause every one of these companies, not the ones that are, well, well, more established in the bigger boots, I would say, like Intel.

Oh, Maybe not. Yeah. Those guys, Um, you know, they do what we did didn't start up land, which is here's a problem, let's go, you know, see if the market's got that problem and build a product around it.

The problem is they're a big enough market. You know, that's the next challenge. And that's the sorting hat, right?

Is there enough people that want network figured out and security as we interconnect, um, you know, multi-cloud, blah, blah, blah, blah. Is that, is that valuable enough to pay money? That's the sorting.

So, and I would argue that the sorting hat has worked pretty good. Okay, so Let the sorting hat sort, But what's the mechanism of the sorting hat? I mean, I don't want to stretch your, The market.

Uh, No, I hear you. I don't wanna stretch your metaphor too far. Okay, Go for break it.

You can do whatever. Well, yeah, thank you. I, one of the things that has struck me about, about all of these messages and stories, whether it's keynotes, vendor meetings that I've had, a few customer discussions I've had, is that there's a whole lot of talk about what this widget or this tool or this contribution, um, what it allows, what the outcome is, a technical outcome.

Now we can scale to 10,000, or now we can, you know, handle a model of this size or whatever it is. And there has been almost none about how they accomplish it, how it's done. It's, it's, I don't wanna say that it's like empty stories and messaging because there's clearly substance behind it.

But why isn't anybody talking about that? It's almost as if you're just supposed to trust us and this is how it works. Here's our demo.

That seems to be the level of excitement and interest that, but Isn't that an indictment of society in general? No one looks, I don't wanna get back On you. I cracked over the door.

Totally An accident. Forgive me. All right.

All I heard was Petri dish and now I need a hazmat suit. I think the Petri dish idea is a little bit more compelling. I'll stick with sorting Hat, I guess because they, because you know, there are colonies that are gonna grow and there are those that are gonna die, Die.

It's gonna throw a net, be through a natural comp competitive mechanism. Alright, we got one minute left. We're gonna do a little speed dating your big story.

Favorite thing at Q Con this year? Mitch Kubernetes everywhere. It, it is the dominant, it is the operational workload operating system.

Your favorite. I've been impressed by the dark horse of this conference, which is Web assembly, wasm, Wasm. Good.

That's the dark horse. Dark horse, of course, the dark. We're gonna talk about this probably in the next show, but my favorite is patent trolls.

Man. They're after all the open source guys and there's high drama and conflict. What more do you Want?

What patents? We don't need no stinking patents. Look, for me, it's the viability of open source business models.

And I think, I thought for a while that we'd answered that question, but I'm beginning to think we have it. That coin flipped. I agree.

All right. Hey, we hope you enjoyed this. We're gonna be back Monday with a fresh tech, strong gang.

But for now, as I said, we're here in Salt Lake City. You are not. This is Alan Shimel.

We're out.