The Changelog: Software Development, Open Source - Mainframes are still a big thing (Interview)

Episode Date: January 27, 2023

This week we're talking about mainframes with Cameron Seay, Adjunct Professor at East Carolina University and a member of the Governing Board of the Open Mainframe Project. If you've been curious about mainframes, this show will be a great guide. Cameron explains exactly what a mainframe is and how it's different from the cloud. We talk COBOL and the state of education and opportunities around that language. We cover the state-of-the-art in mainframe land, System Z, Linux on mainframes, and more.

this week on the change law we're talking about mainframes my notes right yeah mainframes mainframes are still a big thing and we're joined by cameron say adjunct professor at east carolina university cameron is also a member of the governing board of the Open Mainframe Project. He's a proud member of the Knights of VM and an IBM champion for System Z. If you've been curious about mainframes, this show will give you a lot of information about this area. Cameron explains exactly what a mainframe is and how it's different from the cloud. We talk COBOL and the state of education and opportunities around that language. Cameron shares what the state of the art is in mainframe land, system Z, Linux on mainframe, and so much more.

A big thank you to our friends and our partners, Fastly and Fly. Our pods, they're fast to download globally because Fastly, they are fast globally. Check them out at fastly.com. And our friends at Fly, let you put your app in your database, close your users all over the world. No ops required. Check them out at Fly.io

Our friends at Sentry help teams build better software faster and you can try for yourself today at Sentry.io slash demo slash sandbox. Try a live demo of Sentry right now. No talking to a salesperson. Just give it a try. It is a fully functional version of Sentry that you can poke at.

And best of all, our listeners get the team plan for free for three months. Head to Sentry.io and use the code changelog when you sign up. Again, Sentry.io or head to Sentry.io and use the code changelog when you sign up. Again, sentry.io or head to sentry.io slash demo slash sandbox. All right, we have Cameron Say here, who's a mainframe thought leader and an adjunct professor at East Carolina University. Hey, welcome to the show, Cam. Very glad for you to have me. Very glad to be here. Very glad to have you here as well. We were at All Things Open and somebody walked up to me and she said, have you guys ever covered mainframes? And I said, no. And she said, would you consider it? And I said, sure, we'd consider it. But I have no idea who I would talk

to about mainframes. And she said, oh, I know the guy. And she gave me your name. And I said, all right, I'll reach out and talk to him and see if he'll come on the show. So here you are. Pretty cool. Very glad. I mean, this is very nice. I think that was a protege of mine.

Because I usually go to all things open. I've had some health reversals the last few years, so I don't travel as much. But I used to go to all of them. So Todd's a really good friend of mine. Yeah. Oh, cool. Big fans of Todd as well. Big fans of all things open. I think it's just a strong conference for software at large, but really those who are pursuing, you know, the dream of open source, essentially, whether it's from hardware to software to ancillary businesses

around it. Anybody who really cares about software being open for people to use. I love that conference a lot. So until we're otherwise informed, Cameron, you are the person with the most mainframe students in the university class this term in the U.S. You have about 100 students learning mainframes from you. Is that correct? Let me correct that data right now. Let me correct that right now. Okay. That was based on some projections that I had four sections planned. Okay. Only three of them materialized. So I actually have 75 students. My colleague, Jeffrey Decker,

who's a good friend of mine, he has 88 at Northern Illinois. So he's got the title this year. He's got the title this year, but I want to keep the record straight. I don't want him claiming, but I'm very happy to have 75. And usually I have over a hundred. So that's not an unrealistic number. And that's a lot in the US. That's a lot. Well, either way, a contention for the title is enough for me. I feel like mainframes will be the last thing on the list to try to contend for a title. Right. Exactly. Exactly. That's why nobody knows about the crown. So that's why it's so sweet. It savors so sweetly because nobody knows about it. Right. But it's a big thing. And in this conversation, you'll see why it's a big thing, hopefully. Yeah. Well, let's start right there.

First, give us a definition or a general explanation of what they are just for the people who are completely uninformed. Maybe they just started programming, listened to the show. Mainframe is maybe a term they've heard, but have never even come across a real one or the concept. Define it. And then we'll talk about why it's still such a big thing, even though they seem like they're ancient, but it's still a big thing. Still students signing up for new classes. Absolutely. So the term mainframe used to be a pretty generic term that applied to a whole class of machines that did business computing. The original technology was developed by IBM in 1964, but there were spinoff companies.

Gene Amdahl, the designer of the IBM mainframe, created a company called Amdahl. So all these machines, they did large-scale, high-transaction business data processing. And there were four or five or six companies, Xerox, Bors, Welcome, DEC. Everybody had a mainframe. These were all, in quotes, mainframes, right? Which was synonymous for computer back in the day. Now, with the advent of the PC and the kind of elimination of a lot of those companies, all of them but IBM. So the mainframe, the term mainframe really means IBM System Z. That's what it means today. IBM System Z. That's a very specific technology, a very specific proprietary technology. But the term mainframe used to be a broad definition. So is System Z a hardware? Is it a software? Is it both?

Yeah. When you use the term Z, you're speaking to an architecture, right? About architecture. And as some of your listeners may not know, I'm sure you gentlemen know, when you talk about architecture, you're not really talking about a real thing. You're talking about an idea, a game plan, a plan, a blueprint, right? But it applies to both the hardware and the software, right? So system Z, mainframe, IBM mainframe software, all proprietary, but we started to use a lot of open source stuff. We can talk about that, but that's what we mean when we say IBM system Z, that's what we mean, the architecture. Okay. And so we know there's a lot of those systems out there, especially in the Fortune 500s, large data processors, medical facilities, credit card transactions, etc. Are there new IBM System Z systems and businesses being stood up today or is it all legacy? Well, that's an interesting term, yes and no. But so just so we have a correct

context, IBM makes a new mainframe every two years, right? Every two years, they have a mainframe that does stuff the previous mainframe didn't do. So it's not just a nominal upgrade, it's a specific upgrade. So yeah, those machines are, they're new. Those machines are being stood up now. But the thing is that there is this existing suite of applications. What most people don't know is that probably 90% of business transactions globally go through a mainframe. Somewhere, they go through a mainframe. 90, 95% of all credit card transaction globally go through a mainframe. It is the core and the foundation of the global economy. That's just a fact. That's a fact. And most of those programs

are in COBOL. And that's not going to change anytime soon. So these companies have to, when you say, use the term legacy, yes, but these are the core, it's legacy, but it's actually the core application of core applications of their businesses. You talk about Bank of America, you talk about Wells Fargo, you talk about Home Depot, et cetera, et cetera. If a company runs a mainframe, the mainframe applications are the core of the company's business because the company is using the mainframe because it has to. The nature of his work insisted to use a mainframe. And so those applications on the mainframe are the mission critical applications of the business. That's a true statement. Most people don't know that. is IBM Z mainframe today or system Z now, like in terms of a machine itself? Is it, obviously it's about throughput and transactions and things like that, but like, what is the machine itself? So you make a very, very relevant observation in that. Why not just put a bunch of commodity processors together and do what a mainframe does, a la a supercomputer? It's a

reasonable assumption. And that's the assumption everybody made in the 80s and 90s. I don't say everybody, but a large portion of the people, not mainframe people and not any serious computer scientists who really understood the differences between the architectures. Most people did not understand the differences between architectures. So they just assume you could take a bunch of AMD or Intel processors and do mainframe stuff. For some very easily explained reasons, you cannot do that, right? The mainframes claim to fame is the fact that it can process transactions. Now, let's think about the nature of transactions. They can't be done in parallel, right? Because you could never get the scheduling

so precise so that you can guarantee transaction B is always going to come after transaction A. You may get close, but you're going to mess up sometimes. And you can't afford to do that in banking or insurance or hospitalization. You've got to have those transactions done in sequence on a single thread. That's the key. The stuff has to be done on a single thread. Now, you can use multiple threads if you're clever and you're tricky, but you just can't throw the workload on AWS and say process Bank of America's transaction. It's not going to work because there is no infrastructure that can process a million transactions a second except the mainframe. And if you need a million transactions a second process,

as do Walmart, as do Bank of America, as do MasterCard and Visa, then you need a mainframe. There's nothing else that does it. Nothing can even think about it. It can't even contemplate doing that because the architecture will not support it. I want to throw Ticketmaster into that mix, too, because I tried to get my T-Swift tickets and I just could not do it. Come on. Yeah, I heard about that, man. I mean, now, if there was a mainframe, that problem would not have happened because, see, the mainframe scales vertically. Is that right? Okay.

It would not have happened because the mainframe scales vertically. It doesn't scale horizontally like AWS and Azure. They scale horizontally, and they do so very beautifully. But when you talk about scaling in the cloud world, you're talking about more servers, more wiring, more cable, more space, you know, more wiring. When the mainframe, you're just talking about adding a processor and keep it moving. And worst case, worst case, bringing another box to double the capacity, but you're only talking about the size of a rack. These things are conventionally racks today. So you're talking about a mainframe taking no more space than a rack of conventional servers. So it scales vertically instead of horizontally.

I don't know. That explanation forgot your question. I don't know where your question was. I hope I answered it. No, it does. It does. Because I think, you know, it's interesting to bring this show to the change level because we, I think to talk about mainframes in this day and age, it seems like we shouldn't, you know, like you shouldn't have. Antiquated. Exactly. It seems dated. It seems like, well, isn't that old? Isn't that busted essentially? And clearly it's not, you know, but defining what a mainframe is and understanding that you can't just string together a bunch of PCs to create a mainframe to, it's designed transactionally to be very precise, very, you know, timing and stuff like that.

That makes a lot of sense. The interesting part is, I think is there's two sides. There's the hardware side of it and there's a software lot of sense. The interesting part is, I think, is there's two sides. There's the hardware side of it and there's the software side of it. But then particularly once you get to the software, what language is still relevant? It seems COBOL is still the leader in mainframes. I'm curious why. But from a hardware standpoint, what exactly is a mainframe? Like, is it a bunch of hard drives?

Is it a bunch of CPUs? Is it like an insane power supply? Like, what is a mainframe itself, hardware wise? You ask really good questions. I can tell you guys are a couple of young techies because you know how to find out stuff you don't know. You ask really good questions. So I will tell you, without getting into the weeds too much,

when you talk about a mainframe, you're simply talking about a processor. That's what you mean is a processor. That's where all the magic is, in a mainframe general processor. Now the mainframes key to success is that it offloads as much of the workload as it can to something other than the general processor. So something else, the only instructions, the general processor are running are the ones that it absolutely has to run, right? Like backups can run in the background. All that IO, that networking IO, that goes on somewhere else in something called an open

systems adapter. That's not run on a general processor. DB2, the database, it's a subsystem. I don't know if I use the term subsystem, but that's what these things are, subsystems that run below the general processor while that general processor is doing its work. So it is very, very efficient. And you use close to 100% of the instruction cycles, right? Whereas on a PC, on an Intel, and my lectures never disparage any technology that's not mainframe. This is not a competition, not a comparison.

Each technology does what it does well, right? But from a processing instruction standpoint, AMD and Intel are very, very inefficient. Now, they don't see themselves well, right? But from a processing instruction standpoint, AMD and Intel are very, very inefficient. Now, they don't see themselves inefficient, right? They see themselves as efficient as they need to be because they get the job done. But mainframe people are like, okay, you're wasting 80% of your cycles? Are you crazy? Mainframes don't waste 5% of their cycles. So, it's very, very efficient technology and everything built. And now storage is always a separate issue. There is never that refrigerator box that we call the mainframe. Your storage is never physically on that box. Your RAM is, your local storage is, but your long-term storage disk is somewhere else, usually connected via fiber, always connected via fiber. So that's

what a mainframe is. It's just a processor. A big old processor. Wow. Yeah, a big old processor that has a whole bunch of subsystems connected to it. That's it. Yeah, I got to imagine the reason why it's also probably something where you don't want to waste cycles is because it's so expensive, right? So there's time and cost, right? You got the idea of timesharing back in the day, like when you first came into programming, when you did punch cards like Bill Gates back in the day, like you had a timeshare because these things were just so expensive.

I got to imagine that's why you don't waste cycles because one, you want to be efficient. And then two, it's just extremely expensive to scale vertically. Even if you can, it's just so costly. Yeah. I don't know who told you that IBM charged by the cycle, but that's how they charge. So it's cheaper. Yeah. Is that right? It's efficient. Yeah. But that's by design. It was designed to be, I mean, it was designed to be as performant as possible. Whereas, and see, it's not a comparison of which one is better. Intel and everything else other than the mainframe, its purpose is to be available, to be cheap and to be accessible. That is what Intel processes are created. That is why Windows is an operating

system. It's intended to be cheap and well, you can debate cheap, but available, accessible. Affordable. The mainframe accessibility for the mainframe is an afterthought. We're going to build what needs to be built and then you figure out how to get to it. You figure out how to get to it. You figure out how to pay for it. We're going to do what needs to be done. It's just a different way of thinking. One is not better than the other. It's just a different way of approaching the situation. So how is the main frame world then different or the same as high performance computing? So HPC seems very similar in terms of availability and time sharing and cost and even form factor in terms of being the size

of a rack. They're very similar based on description. Yeah. I mean, that's a common thinking that mainframe, there's some synonymity between mainframe and supercomputer, but they just, they do different things. What a supercomputer does, it takes problems down and it breaks them into digestible pieces that can work in parallel. Mainframe doesn't do that. Mainframe runs on a single thread or a few threads, right? So parallel is going to use as many threads as it needs to get the job done. And the scheduling is done on the back plane. So you put the pieces together on the other end, but they don't necessarily have to be in sequence. I mean, if you're doing some type of imaging, they don't

have to be in sequence until they get to the destination. Then you can assemble them, right? But with a mainframe, you have to do everything in sequence, right? So they have different workloads. Their intentions are different. Their intentions are different. So ballpark for us, we say it's very expensive. Ballpark for us, what it would cost today for a modern, like IBM's latest mainframe to acquire and operate one of those for a year. Like, are we talking hundreds of thousands, millions, 50 bucks? What's the ballpark?

Yeah, you're talking millions. You're talking millions. You're definitely talking millions. So I'll tell you this, the software stack, the IBM answer for everything is it depends, right? Of course, but I'm going to give you a real number. When I was at North Carolina Central, no, I was at A&T,

and we were using the IBM software stack for free, right? So I had a project where I needed some in-kind support, and I needed to put a value on the software I was using, which IBM doesn't want to do. It was like $5 million per year, $5 million per year, just for the software, just for the software. And that was a real number, because IBM doesn't give you something if it's not a real number. To acquire the machine, at least, depending on, I mean, you can trick it out. I don't think you can do any serious, you can have any serious implementation for less than 2 million. I could be wrong about

that. But now understand what you're doing. You've got to weigh the total cost of what it is you're trying to do against another way to do it. So while a rack of service may be a lot cheaper than a mainframe, well, how many racks do you need? I tell people all the time, I can move an inventory of goods from Durham, North Carolina, to Hollywood, to Los Angeles, California in my SUV, right? How many trips would it take me? Is my SUV cheaper than a Peterbilt? Absolutely it is. How many trips will it take me? A lot of trips. How much time will it take me? So where I can take everything

in a Peterbilt. So it depends. But to give you a number, so $5 million was the value of what we were using. We were just using generic stuff. We weren't using anything sophisticated. And probably $2 million to get in

from a hardware standpoint. But IBM, they have a variety. They have an infinite number of plans. But we're talking millions. We're not talking hundreds of thousands but we're talking millions we're not talking hundreds of thousands we're talking millions we're talking millions okay so orgs companies that need a mainframe they already know it like they're well aware they have huge revenues they have huge problems they have transactional problems specifically that a mainframe can solve so they're up and running which like you said approachability not necessarily the point here from From the software side, what's interesting to me is like you have 75 students

this year, you know, your colleague that you named has 80 students. There's people that are still enrolling, taking classes, learning this stuff. COBOL, I've heard, you can confirm or deny this, that all the COBOL programmers are basically retiring or sadly passing away. Like they're aging out. The ones that help get all these mainframes established and build the software in these complex systems. And there's like this growing underground demand for COBOL, which makes it a very interesting language for that one reason.

I was like, you're going to have demand. Is that true? Is that why people are still enrolling in these classes? Because like- That is absolutely true. Okay. That is absolutely true. And it's going to get worse. Now, there is not a critical shortage now, even though it may appear to be one. The people that need co-programmers don't know where to find them, but there are people like Bill Henshaw and other people like that who have identified this horde of people that are available to work, but they're aging and no one is teaching the language. There's 17 campuses in the UNC system. ECU is the only school teaching

COBOL. And my colleague, Jeffrey Decker, who has more students, I'm going to tell him, say he has more students than I do. Other than him, there's just a few of us. I don't know anybody teaching COBOL. I don't know anybody teaching it. Yeah. I don't know anybody learning it until, you know, speaking with you and seeing your students. So is COBOL the only player in the game though? Are there other ways of communicating with mainframes that isn't learning COBOL? Can we have like a translation layer

between JavaScript and mainframes or something that everybody else is learning already? Yeah, well, that's interesting. A mainframe would ask you, what do you need that for? But I get your point. We're doing more stuff on the mainframe now in Python. I'm in good with the compiler team, the IBM compiler team in Toronto, Canada. And well, Toronto, everybody knows Toronto, Canada, I need to say that. But they're younger guys, they're 20 something, 30 something. And they write, they use, can they write COBOL? Absolutely.

These guys are young geniuses. Can they do Assembler? Yeah, we need to talk about Assembler. But we're doing more and more stuff in more conventional language. Also, we're doing a lot of COBOL in Visual Studio and an Eclipse-based IDE called IDZ. Right. So we are bridging that gap. We are bridging that gap, Jared. We are bridging that gap. But I mean, and for mainframe, you said the gap really doesn't need to be bridged.

But to your point, yes, we're bridging the gap. Well, I think IBM wants to sustain their ability to operate at that financial clip, right? So somehow we got to backfill the primary language that drives the mainframe. And if it is true that there is an age-out process happening and a lack of teaching, and then just by nature, less coming in, we're going to eventually have some sort of problem if these machines are so critical to large-scale business. As you

mentioned, I don't know if it was in the show or in the pre-call, but many, many transactions that you're doing on a daily basis are happening on a mainframe whether you know it or not. At some point, these things are critical functions of society. Yes, they are. I mean, they keep the global economy running. And the companies are seeing it now, but they just aren't depending on the universities to fill that gap. And I don't blame them. My students all get jobs. They all get great jobs, but there's not enough of them to go around. So companies are looking at alternate methods and there's some pretty successful methods that they're employing. But the need is there, Adam. The need is there. And companies are aware of it now. They're feeling it. IRS, they want us to train 600

people yesterday. So they need a lot of people. This episode is brought to you by our friends at Postman. Postman is an API platform for building and using APIs. They are most known for API testing, and you may already use them, but they've built a full-featured API platform to help developers along each step of the API lifecycle. But what does it mean for Postman to be an API platform? Well, from API design, testing, monitoring, documentation, mocking, to the sharing and discoverability of APIs,

they've built a full suite of tools to help teams build APIs together faster. Over 20 million developers use Postman to deliver their APIs. Plus, they have a generous free tier. Start designing, developing, and testing APIs. Organize all your API development into workspaces and share those workspaces with other developers. You can create public workspaces to collaborate with the world's developers. You can back up your work to Postman's cloud. You get their core tooling and collaboration for up to three users. Sign up and start using Postman for free today at postman.com slash changelogpod. Or for our listeners already using Postman, we encourage you to explore the

entire API platform that Postman has to offer. Again, postman.com slash changelogpod. So let's say, hypothetically, on this point, if I'm a working developer, like let's say hypothetically on this point if i'm a working developer like let's say i'm a java developer i work in enterprise you know i make my 80 to 150 a year i've been doing it for a while is there a reason for me to go out maybe take some night classes pick up cobalt like would i be able to increase my salary potentially or like really upgrade by adding this skill? Or is it kind of like, well, you're already making that much. You're going to make about that much. What do you think? I know it's different in different areas. That is a very good observation, Jared. And I would not say that somebody that's a seasoned Java or C++ developer needs to add COBOL to their portfolio. But I would say that,

first of all, they wouldn't need to take a class. They could just do some self-study. If they're a seasoned developer, they may need some help on the mainframe piece because you got to know how to use it on the mainframe. But I would just add it to my repertoire. I would just add it to my repertoire. I don't know that it would get you a higher paying job, but it's going to expand your options. And it might, I mean, if the need gets critical enough, because my folks are getting like 65 bucks an hour now. Now, you know, in the Java world, you can get that. You're getting that if you know what you're doing. So I can't say. Now, from a career standpoint, I will say, like, let's de-hypothesize

a little bit. Let's say this is a 26, 27 year old developer with four or five years in the game, right? I would definitely recommend for them to retool, to expand their skillset, try to get on a mainframe career track because I'm pretty sure the upside financially in the longterm is going to be better for them in the mainframe space than it will where they are. I think so. It's just arithmetic. There's more people that know how to write Java than there are know how to write Assembler. So your skills are more valuable. And what I'm seeing is a lack of managerial expertise.

So these enterprise-wide decisions about moving to the cloud, et cetera, et cetera, in quotes, underline, these decisions are made in complete ignorance of what the mainframe is and what the mainframe does. And so people that understand the mainframe can help the company make better long-term infrastructure decisions. The mainframe is not right for everything, but what it's right for, it needs to be used. If you have a mainframe workload, you need to be using the mainframe. No argument about that. Okay. So you have a managerial opportunity there, if not a direct implementation opportunity there. So what kind of decisions are being made out there

that are going towards cloud, I guess, options that are better suited for mainframe? Like what kind of workloads? What I'm seeing a lot of is trying to put things on Azure and AWS, which I'm a daily AWS user. I have nothing bad to say about AWS, but it's just not going to run mainframe workloads unless you have a mainframe there.

So I'm seeing a lot of that attempted. There's a kind of a famous case study now with TSB Bank out of Canada. They tried a migration and it ended up being an utter catastrophe. I think it's salvageable. Yeah, but it costs them a lot of money. It should not have been attempted. So those types of decisions, Adam, that they're being made, I'm not saying, look, if you have an application

that's going to run fine on a cheaper platform that's not as expensive as mainframe, you'd be crazy not to. But you've got to be real careful about trying to put things in the cloud proper that actually belong in hybrid cloud with a mainframe somewhere. So that's what I'm seeing.

I'm seeing a lot of that. What's the API, I suppose? Does an enterprise often own their own mainframe or are they borrowing time? You mentioned before the amount IBM charges per cycle, I believe. Am I essentially renting time or am I buying the full thing? What's the common path? And is it simply an API call to run the workload there and then data out and ingest somewhere else? How does it work?

So both. I see probably many more leasing machines than I do actually buying them outright. But companies do buy them outright. I don't know the cost figures for all. Now, exactly what are you looking for in terms of API? Because do you mean to get to the machine, to run an application on the machine? Is that what you mean? How do you put your workload on the machine, essentially? What's the, I guess, API is a common term, but I specifically mean like, how do you interface

with the mainframe to make the workload go? I know exactly what you mean. So there is kind of a default interface called ISPF, Interactive Systems Productive Facility or something. I'm horrible with acronyms. I can tell you what it it means i can't tell you what the words mean but ispf is like the default now it's very old school it is the green screen although it's not green anymore it's different colors but it's what you would call the green screen uh that is probably still the the interface of choice for mainframe is doing all and that's what i teach but i am seeing now they're doing a lot of there's a there is a RESTful plugin called Zoe, Z-O-W-E, that runs that you can use to connect to ZOS. ZOS is an operating system so that Zoe connects and Zoe is written all in REST. You can just web oriented. You can write to it to Python, Java, whatever you want or simpler.

So you can do all your development in Visual Studio VS code and then run it on the back end on the mainframe. We connect to the machine usually with a 3270. 3270 is a monitor type. We have an emulator for that. So that is our class's connection to the mainframe. That presents us that ISPF interface. But there's all kinds of Eclipse-based interfaces to it now, and the mainframe is very accessible. And I'm on the OpenMainframe project, so we look at using

open source mainframe, because Linux runs on the mainframe great. It runs on the mainframe great. I use Linux on the mainframe all the time. I was just going to ask that. We noticed that on IBM's website, that you can run Linux on it, and I wondered if that was kind of missing the point or not. Like, does it make sense to get mainframe and then run Linux on it. And I wondered if that was kind of missing the point or not. Like, does it make sense to get mainframe and then run Linux on it? Oh, absolutely. Absolutely. I'll tell you when it does, Jared.

ADP, for example, they've got an instance, they have 4,000 servers, virtual servers, Linux servers, doing their business, right? A guy by the name of Phil Scully, he's the manager of that. So yeah, it makes a lot of sense plus and see we don't have enough time to get into the business of this there's a lot of business involved in this for a variety of reasons there's a lot of history this does a lot of politics to it you can run linux on the mainframe a lot cheaper than you can zos that's just because ibm says so

things are expensive because ibm says they expensive. But if I've got an application running on COBOL and ZOS, it's entirely possible I can take the same application, run it on Linux on the mainframe in COBOL, a lot lower cost. I do see that happening. But no, and by the way, just so you know, you guys don't know this, but by far, both of you are familiar with the concept of a hypervisor. IBM's hypervisor, ZVM, is by far the most powerful hypervisor on the planet. There's nothing close to it in terms of performance, in terms of security. Now, I don't say that to Milan VMware. VMware is a nice product, but that's worse than, the comparison is worse than comparing Windows to Linux. It's that great.

ZVM can do things that VMware can't even dream of doing. Like in the mainframe world, they call them guests in the mainframe. They're called virtual machines. They call them guests. One guest is never going to bring down another guest. That's never going to happen. The guest is completely isolated. Now, I know that's an issue in the VMware world. It has been improved upon later, but I remember that used to be a big issue.

One virtual machine could take down a whole sector. So that doesn't happen in the mainframe space. Not at all. You mentioned IBM names their prices, and that's obviously just because the circumstances. Like they are the king, and you serve at the honor of the king, so to speak. Is that ship sale, like is there an opportunity?

Because all you would need is like one competitor. You'd need Pepsi to their Coke to like really keep IBM in check and maybe make this stuff more economical in general. Or is that just never going to happen? These companies tried and they failed and IBM is just going to have that little monopoly and sit on it. What do you think? I'm a tech guy like you. I don't use the term never. I don't know. You know, I don't know. I can tell you what I can't foresee. I don't foresee any competition from a technology standpoint because it would take hundreds of billions of dollars in R&D to do something like the mainframe does, in my opinion. What I do see in competition is competing solutions, right? Everybody's always trying to figure out how to do these transactions. And there is a company called LZ lab that swear to God, they figured it out.

Now they're in court right now with IBM. I was part of the case. The case may have been resolved. I don't know. I've been in court with IBM right now. Cause IBM says, no, you're telling a lie. But to answer your question, Jared, I don't really see a competing technology. I see competing solutions. That's what I see. I see. That makes sense. Well, I asked that question earlier about how do you interface with the mainframe itself to consider maybe AWS mainframe being a thing at some point. I mean, AWS loves to compete with everybody,

eat everybody's lunches. So I got to imagine at some point, they will just give out some sort of interface to a developer or to a team. Oh, wait a minute. I mean, I didn't want to imply, I mean, AWS is involved with several hybrid solutions involving the mainframe. It's just not a textbook solution, right? It's just not the standard, but they are, and a good friend of mine was a senior vice president at AWS. She said, those guys are great. They're not, they're not jokers. Those guys are great. So they are working on a solution where people can,

and AWS is not trying to do what the mainframe does, but they're trying to provide pipes, right? This thing is not necessarily about what you're running on. It's about where it's running, right? Is it on-prem? Is it remote, right? How secure is it? How reliable is it? So yeah, I mean, I do see people coming up. It's not that AWS is not in the game and Azure is not in the game. They're not Johnny-come-lately. They're working on this stuff. And they may find something that takes the mainframe out.

I don't see it, but it could happen. That's why I don't say never. What about the flavor of Linux? Like, how specialized Linux, what distro is it? You know, what distro is running on the system Z mainframe? I'll tell you what I've run on it. I've run Ubuntu, I've run SUSE, I've run Red Hat, I've run, oh, Debian. I mean, I know Ubuntu is Debian, but I've run Debian, pure Debian. Anyone that, I mean, just about every distro, major distro, there is a port for it. It's called S390. S390 is the port.

And you just port it. And so I use them. My favorite one on there is Ubuntu because I like Ubuntu. Ubuntu runs exactly the same. I got Ubuntu on AWS, Linux on AWS. I got Ubuntu on the mainframe. I notice no difference. Wow.

Nice. Which is awesome. I got to imagine at some point the Linux kernel people or those driving those distros are going to want to have access to a mainframe or maybe they do, I don't know, to sustain the support. Well, no, it used to be, I don't know if this is true now, Adam, but it was a fact that 12% of Red Hat's revenue came from their mainframe space. That was a fact. I don't know if that's still true. Is that right? Yeah, 12% of their revenue. I knew the guy who was over there, Brad Henson. They had a huge

mainframe. IBM used to give them mainframes so they could run stuff on it, right? So yeah, so they've been, and I don't know if they still are, but that was part of IBM's willingness to buy them because Red Hat worked more closely with IBM than SUSE did. SUSE in business Linux computing mainframe, in business mainframe computing, which is a thing, business mainframe computing on Linux is a thing. SUSE is a top dog. But Red Hat had courted IBM more carefully. So that's why there's a port for Red Hat.

And that contributed to IBM's $34 billion purpose because they felt they knew enough about Red Hat to make it. That's not the only reason. I know there's a lot of reasons, but that was involved. I had a conversation with both ends. Yeah. Smart by Red Hat. So you mentioned earlier the Open Mainframe project. Seems like we're talking open right now. Can you dive into that? Tell us what that is. Absolutely. Absolutely. It started, I think, in 2016 with several large companies using mainframe. IBM was one of them. CA Technologies at the time, before they had been bought out. ADP, a couple other companies. My university, North Carolina, A&T was involved. So it was just an attempt to, or it is an attempt to promote and report on the use of open source software in the mainframe space. Now, I initially had a problem with that because our focus was Linux on the mainframe, right? Which Linux on the mainframe doesn't run, it runs on a proprietary operating system, ZVM, that's owned by IBM. So that's a conflict right there.

But my out for that is you can actually take Linux out of a disk and pop it in the mainframe and install it. So Linux is truly open source on the mainframe, right? And we get together, we have projects like Zoe was a manifestation of that. I'm a member of the COBOL working group. I'm co-chair of the COBOL working group. We have published a survey about misconceptions of COBOL, et cetera, et cetera. So we're very busy promoting open source software in the mainframe space. It's a lot of fun, too. I'm on the governing board, by the way.

That's cool. I saw your name there, second in the list. Boom. Always second there, Cam. Always second. You're second again. I can take it. With those guys, second is just fine because these guys are really sharp.

I look around what the hell am i doing in this room i don't belong in this room but uh i like hanging out with them they're super sharp guys i feel like that a lot on this show yeah yeah that's very cool so you mentioned cobalt again let's loop back around the cobalt so you said you published something about misconceptions i mean i don't know a lick of COBOL, but I have misconceptions, I'm sure. And it generally is frowned upon by us modern developers. Can you tell us about COBOL, maybe help us fix some of our misconceptions about it as a language? Gladly, gladly. The paper was authored by my first author was Maggie Hall. She's a professor in Austria, but she's an American. And it just listed the, now I don't remember what they are

right now, but they were, I mean, I'll send you a link to the paper. It was like five misconceptions and like nobody uses COBOL anymore. And so many people said nobody uses it anymore or it's hard to learn. Right. So those things, there's a lot of misconceptions. People just don't know. And, you know, you didn't ask this question, but let me fill in the blanks a little bit and explain to you why COBOL is used so pervasively. So COBOL, first of all, it was not intended for computer scientists and mathematicians. It was intended for business people, right? It was developed for back office people, right? So they could write programs. So it uses language, perform, do until, you know, read, write. It's not cryptic at all. It's business English,

right? Business English. Also, its procedural nature makes it very, very efficient to process raw data, right? So if I got a block of data that I got to chomp through, the terseness of COBOL syntax where the processing is concerned, No layers of stuff to go through, just code and data, just code and data. And COBOL is very tightly interwound to assembler. So it's almost an assembler-like language. So it's very efficient. The object model, and people don't, well, why don't you rewrite everything in Java? Well, that's not going to gain you anything because the object model doesn't bring anything to the dance in terms of the nature of the data that COBOL processes. There's nothing to adding classes or the modularity that Java brings to that. It's not going to add anything.

It's just going to increase the complexity of it. So COBOL is very terse and very simple, and it's very easy language to follow, right? Very easy language to follow. So it became by far the language of choice for the entire world for about a 30-year run before Java. So I do have a friend of mine who works at a large credit card processing company that still runs on mainframes, probably always will.

And feeding into this whole, like there's this underground demand for COBOL developers or mainframe developers. He says that there's a handful of people in that. And this is like a fortune 1000, maybe fortune 500. I'm not sure company. He's like, there's a handful of people here who understand this system. And it's not cobalt language.

It's like the actual nuts and bolts of how this business runs through the mainframes. And he's like, maybe there's a half a dozen that might be gracious. Maybe there's three people in this company of thousands of people that actually know this thing. And if they're not available or they're gone or whatever, like we can't do our job lots of times. He's a, he's a programmer there that works on internal tooling. He's not a mainframe guy.

Is that the case at a lot of these big orgs where it's like, gosh, uh, it's the domain knowledge that seems so valuable. Of course you could pick up COBOL, but, and you can learn how to talk to a mainframe, but is it the domain knowledge that's like so gnarly and so ingrained and so many lines of code of it at these large orgs that like are going to keep people busy for years, but also might, maybe the knowledge of how they work might disappear with a bus factor of one. That's my fear. And you guys are really good. Both of you ask really,

really good questions. I mean, I'm very impressed with the nature of the questions. I see Adam shaking his head. I love it. I say three times this time, Cam. You're like our best friend now, man. You're invited back every week. Yeah, but yeah, it's the institutionality because see that code, some of these programs have been around for 30 years and they've not been static for 30 years. We've changed some business rules, changed the program. The logic of the business is intertwined in the code, right? It's intertwined in the code. And you're absolutely right. There's only a handful of people. That's my concern now is I want to produce a managerial class that understands this stuff. Look, emotionally, I'm not emotionally attached to mainframe. I've had a great career because of it.

I'm pretty well known because of it. But if it goes away tomorrow, if it's replaced by something better, I'm good. I'm good with that. I'm good with that. I just don't want to replace with something that's not better. So we need to retain this institutional and see what a graceful transfer looks like. Because the transition may not be from one generation of mainframe to another. It may be one generation of mainframe to something else. But that business logic has to be incorporated in that transition or it's going to be very, very messy and very ugly. So how then do you attract developers into this world then? Like what is some of your personal mission and what do you think could be the mission of others to bring more folks into this fold?

I've had no problem. Now, maybe I'm just super smooth. Maybe I'm a pimp and don't realize it. You're pretty smooth. But I've never had problems. I've never had problems attracting students to this platform. And it could be because, look, but let me tell you a little bit of something about me. This is going to take a minute or so. But I went into education. I was an IT guy for like 21 years in a variety of roles. Coding, networking, I did it all, right? Management stuff, which I didn't

like management because I didn't get to play with stuff enough when I was a manager. But I went into education. That's a long story. We don't need to get into it, but I love it. I love it. And I started educating at historically black colleges because I went into education unabashedly, unapologetically to educate black people about technology. That's why I went into education. Now, most of my students now are not African American and I don't care. If you're one of my students, I don't care who you are. I don't care where you come from. I don't care what you look like. All I care about is you're in my class and I'm going to do the best I can for you. But that type of attitude, plus I'm focused on them getting jobs. I'm obsessed with it. That's how I keep score. I keep score on how many of my students are leaving my class and going to work. That's how I keep score. And I don't know a lot of professors that keep score that way. And they should. They should. Especially in the tech space. But so students know that about me, know I'm about nothing about them getting a job or furthering their career. And look, I don't care whether you work at Mainframe or not. I want you to have a good job in tech because tech can be a rewarding career and it's not for everybody.

But no, so I don't have problems getting students. I don't, why other people do, I don't know. Cause this stuff is compelling. If you teach, if you tell them, look, you're going to get to work on the most sophisticated business technology in the world and you're going to get to make a lot of money. That pretty much is all it takes. What are the prerequisites then for a good student for you? Yeah, yeah, yeah, yeah. So, so that's a good question. My expectations are different for aquisites then for a good student for you? Yeah, yeah, yeah, yeah. So that's a good question. My expectations are different for a college program than for a boot camp, right? Boot camp, I'm going to meet you wherever you are.

You got to be able to read and write English at the seventh grade level. Got to be able to do that. If you can do that, then we can start, no matter where you are. Now, for college students, a good overall background in IT basics is going to be helpful. You really don't need anything to take this course, right? You don't need it. But the more you know, the more it's going to make sense to you, right? But I've had a woman who her degree was in social work. She had never written a line of code in her entire life. She went through four weekends,

four weekend boot camps, Saturday and Sunday. Now she's in a junior systems program in Ensona. So it can be done, but she is unusually focused and unusually mature, but she had no background. She had no background. So that's it. Attitude. But to answer your question specifically, Adam, the right attitude. That's it. The right attitude. And in terms of salary, maybe that's what draws what draws somebody to but i think it's one is the opportunity but then also the financial upside to that too if you're in that critical of a role for a company you've got to be looking at executive level at some point or some sort of ladder rather than just simply you know mainframe grunt kind of thing you know i gotta imagine like that's gonna be a lure to some folks

yeah i mean i've got i've got several students that are vice presidents now, so they're making over 200 grand a year. I've got folks going into sales now, they're doing 200 grand a year. So it just depends on who you are. I tell my students, don't compare yourself to anybody. Just, just figure out what you want to do. But you're right. The upsides are virtually unlimited, virtually unlimited. So very proud of my students. They're crushing it out there. They are crushing it. You said you want to train a managerial class. What exactly do you mean by that? And then what's your strategy to train the managers? Well, see, that's a good question, Jared. I don't really have a strategy to do that because of the nature of the courses that I teach.

Now, I used to teach a graduate level course in mainframe at A&T and the focus was a little different. We looked at case studies, we looked at strategy, we looked at, you know, other things other than just getting the skills. For these undergraduate courses, I just needed them to build some simple chops in this space, right? So I'm not going to get into the managerial conversation, but it's going to be part, I don't, and to answer your question, I don't really know how I'm going to do that now. I do know that a lot of, some of the people, some of the students in a certain portion of them are moving into management. So that's what I want to happen because as they ascend up the ladder, then they take this understanding with them, right? But how I'm

going to do that for my chair right now when I don't teach graduate courses, I don't know. I don't have an answer. I want to, but if I teach a graduate course, that's perspective. I take a strategic perspective when I teach a graduate course. Gotcha. Can y'all tell I love doing this? Can y'all tell that? I love it. I'm eating it up over here. You got some energy. I love it. I love the energy. I'm curious what your boot camp looks like. Okay. So let's talk about the one that the most successful one we had was this summer. And I, I use it a template model because I think this model is optimal. So we had 10 students. We lost two, but we had 10 students for 12 weeks starting in, I think the end of June going through early September. So there were 10 weeks, six hours a day in class, five days a week. Sounds like a lot,

but we broke it up. Three hours of lecture in the morning, three hours lab in the afternoon. So the students, they didn't get burned out. A mix of people, some of them were college grads. Some of them had not been to college at all. And it was the first three weeks were ZOS essentials, right? ZOS essentials. Because we were contracted by a bank to prepare these people to join teams after 12 weeks. So we took that challenge. And then after that, six or seven weeks of COBOL, and then a two-week project. And that worked very nicely. So they come at nine in the morning and go through 12, take an hour for lunch and come have lab in the afternoon. And the labs are open. It's not tedious. The lectures can be a bit tedious, but you got to have lectures. But that, to me, that's the best model. Now the other model I use was Saturday and Sunday from I think nine to four in the afternoon, four weeks,

but you can only cover so much, but that's all the time we had. And those people got jobs too, but they got jobs too. So, you know, there's all kinds of ways to do it. So the eighth has survived. Did they all find work? The eighth has survived the full-time one? Yes. Yes. They all went to work. The day after camp, camp was over on Friday. They all will work that Monday. Nice. How many of you are there? How many, how many Cameron Say's are there? Cause it seems like if we could replicate the education process, we could maybe get this on this growing demand. Maybe we could find more people, more opportunities if there are more of you teaching these courses. Well, I know of one who's a, who's a better Cameron Say than Cameron Say.

His name is Jeffrey Decker. He teaches at Northern Illinois University. And he's the guy that has 88 students. He's been doing this for like 21 years. And he's a wizard. He teaches assembler too. I don't teach assembler. I need to learn how to teach assembler. But yeah. But I don't know. I mean, there's another one of me, Herb Daly, you know, Brit in the UK, black gentleman. You know, he's doing things, but there's not a lot of us. There's not a lot of us. And I mean, I understand why it is, because you've got to be insane to do this. What I'm doing, you've got to be insane. I did not get tenure. I did not get tenure at two universities because of this stuff. Right. Because the departments didn't understand what I was doing. They're like, oh, why are you teaching us?

Because they're coming out making eighty five, ninety thousand000 a year. That's why I'm teaching them this. Go tell Bank of America this is outdated. Go tell IBM this is outdated. Yeah. Go tell them that. Yeah. Yeah. What about the ancillary business opportunity? So there's two financial measures I see happening here. Obviously, IBM selling software. So ZOS, expensive, obviously. The hardware, they've cornered it. They are the mainframe based on your description of what a mainframe is. It's IBM's Z system mainframe. So they own that. They're the largest inhibitor of all this money coming into this world. Obviously, the data centers are raking it in because they've got to put the stack there

and all these different things. And then you've got the other opportunity, which is developers coming in to make money, which, no, knock against the salary, but it's not millions. It's not a multi-million or billion-dollar business like IBM is running. So is there a world where I get skilled enough or a business can be propped up where I buy a bunch of these mainframes and I host them for people? Like, is there any other business opportunities around the mainframe that isn't IBM selling the mainframe or me being a IC in the world? What are you talking about? You talk about a small universe. That is a very small universe. But I know a couple of cats that figured

out exactly what you said years ago. There's this one guy, Sonny Gupta, right? Sonny does this stuff globally. He buys old Z9. Well, that may be a newer machine now, but Z9 was like an old machine back in the day, right? It's an old machine, but a mainframe is still a mainframe. So he does hosting, he does training, and he'll charge you 500 bucks a day to use the machine. And it's not costing him anything because a Z9 is just a glorified electronic paperweight. I mean, a Z9 has no own machines and he runs kind of hosting services on them. So there is a business model. I don't know how standardized it is, Adam, but Sonny Gupta seems to be doing exactly what you said. He has machines all over the place and he'll train you. He does a lot of training, too.

He does a lot of contracting for training and he'll host your application. So, yeah, it is done. So if I'm an enterprise buying in mainframe, I'm probably buying from IBM. I'm not buying from these folks you mentioned. This seems to be sort of a side thing, not the main business of mainframe. Yeah, you're not buying from Sonny Gupta,

but IBM, of course, does have resellers on the retail part of it. So you might go through a retailer. If you buy on a Linux machine, you're going to go through a company called Viacom. But it's not exactly what Sunny does. I don't think Sunny does that. But yeah. And I don't know if IBM sells to anybody directly anymore. I don't know. I mean, I don't know that part of the business that well. I used to know it pretty well. I know people that do. But yeah, I mean, the business model is very intricate.

And it's a lot of nuance to it and a lot of secrecy. And IBM is very careful with their figures that they share. So I really can't tell you what type of secondary aftermarket opportunities there are, but I do know that there are people that are exploiting some. And IBM is the only company making these machines. There's nobody else out there competing on the hardware? Nope. On the hardware, nope. Nope. That's monopoly. What does it look like to get into this industry?

I'm going back to the indie dev, like the person who's like, you know what? I'm going to take this up. Do you have to move to Carolina and take your course? Do you have to be in a city where these large orgs are, like Bank of America, where their headquarters are? Is there remote work? Are there training online? How accessible is it to get in? And what would my steps be if I wanted to get in? So the last, there are three women that I know, none of them are degreed and they all came in

the back door and they all came in different ways. Self-study is a lot of it. There's a lot self-study. You might try to get picked up with an apprenticeship. The apprenticeship model is pretty hot because you come in at like 70, 60 or 70% of salary, but you have benefits. You have benefits from day one. And it's almost like guaranteed unless you step on it, you're going to get hired. So I see that model. But to get prepared for it, self-study, you can do self-study. You're going to need mentoring, right? You're best off if you can take one of my classes so you have somebody like me to walk you through the space introduce you to the space for a semester right but i'm not necessary there's a whole lot of people to get started this game without me they i meet them when they after they get into the game i met these ladies after they were already set up so there's ways to do it

but it it's gonna take some some tenacity gonna, some tenacity, going to take some tenacity. Are there any self-study resources that have your stamp of approval or they're like best in class? Like we could point someone to a website or to a course? Yeah. I'll send you a list, but off the top of my head that the COBOL course at Open Mainframe Project is a great place to start. And then there's this thing IBM does called Z-Explore. That's like the default. That's the starting point for everybody. It's like a series of challenges. It's like a treasure hunt. And it's great. And both of those are free. So yeah, that's a great place to start. They're a bunch. They're a bunch. A bunch of good self-study out there. I'm seeing this now.

This project is an active project. It's COBOL Programming Course. They're at the Open Mainframe Project. And I'm also seeing this mentorship program. Do you know much about that? I do. I know quite a bit about it. So what we do is we take on projects and we give it. That's actually, I don't know who on our team changed that from internship

to mentorship. I understand why they did it, but it's actually an internship program. It was a paid mentorship. So that program is to get the mentors, but they're going to be working

with mentees, right, that are part of the mentorship program and they get paid. So they have projects. Companies submit projects and they work on the projects for a certain period of time, 10 weeks or whatever.

Really cool. A couple of actual distros have come out of that program. Linux distros for the mainframe. One called Alpine. That's right. Alpine Linux. The port. Now, I'm not saying we created Alpine Linux.

Yeah, the port. Yeah, go ahead. Right. The port to mainframe. Gotcha. Yeah. Yeah. It's cool that the OpenMainframe project exists, obviously,

but then obviously having the mentorship slash internship and then the COBOL programming course being there. I mean, you need on-ramps to this. If it's not a university course, undergraduate course, or anything like that, you've got to have a way in. I'm sure there will be people listening to this show, Jared, they're like, you know got to have a way in. And I'm sure there will be people listening to this show, Jared, they're like, you know what? That's pretty cool.

Because how often can you get that close to the center of a business problem? And then fill that vacuum. I mean, you've got a long-term job at that point if you can really get in there. Domain knowledge-wise, I mean, you've got – talk about job security. I mean, that's it right there. Well, this goes back to the three people who have the domain knowledge inside of these credit card companies. It's like they are worth their weight in gold.

And just it wasn't like they planned it that way. It's not like they were being sneaky, like I'm going to create a job. It's just the way that it worked out. And so there's a path to that for more people. Obviously, we don't want it to be three or two or one. We want it to be like a couple dozen so you're nice and comfortable. Yeah, that's a comfort zone. Yeah, we want it to be a few. Yeah. We want it to be like a couple dozen so you're nice and comfortable. Yeah, that's a comfort zone. Yeah, we want it to be a few.

Yeah. We want it to be a couple. Yeah, you don't want to be getting paged when you're on vacation because you're the only one who knows how the system works. But you do want to be valuable,

you know? And see, that's what's happening to these guys now. That's why these guys can't go on vacation. I mean, because they, when they leave something, but I would be remiss,

you were asking about the article earlier about misconception of the mainframe. The woman who wrote that article, Maggie Hall, she also has a National Science Foundation project that I'm a part of called Work Learn. I'll send you a link. Work Learn. You can look it up. Work Learn. And it's a MOOC. It's a MOOC.

Self-study MOOC. Got all my material for both the COBOL and the intro class and the COBOL class, including exams. And this is for the homeless. This is for the financially distressed, including the homeless. That's why she did the project. Wow. So how is that for an on-ramp? How is that for an on-ramp?

Nice. That is an on-ramp, right there. I mean, I'm not homeless, but I would totally take that. Because that's like accessible beyond belief, right? I don't have to leave my house. Yeah, yeah, yeah. You can do that. I love it.

It's got all my material, my recorded lectures. It's got everything. Speaking of the homeless and just the underprivileged folks out there, I mean, like that's half the battle. It's like they're looking for any way that's legitimate. That's not not legitimate, obviously, to get to be in crime or go to jail or something like that, just to find a way back into the financially stable. Even if it's what would seem low to some of us, 80 grand or a hundred grand a year,

like that's a significant amount of money when you have zero and no opportunities. But then 80 grand is pretty sweet for them. It's pretty sweet. Yeah. We know it's not rich. 80 grand is not what it was 20 years ago. I used to tell people, I was making six figures, but six figures was real money. It ain't real money anymore. But yeah, and that's why I went into this also, Adam,

when I went into mainframe, not in the teaching. I was at North Carolina Central University in January of 2005. I will never forget. And IBM came there and made a presentation about mainframe. Now, I had been in IT. I knew what a mainframe was. I had never worked on one. I didn't know anything about it. But I said, man, this is something my students could use because we were having trouble getting interviews because people just don't like to hire from HBCUs. I don't know what it is, but they don't like to hire from HBCUs. Every HBCU I've been has had a lot of superstars. So there's been nothing wrong with my students, but I don't know what the drawback is. But we needed something. I said, this is it. And I've taught this at four HBCUs and it's always been the same way. If you get through this course and you understand, you will get a job. That's just the way it is. And that's the way. I've seen kids, this has changed lives. Some of the kids, when they get this job, they make more money than anybody in their family has ever made, has ever thought about making.

Well, that's spectacular, Cameron. I mean, thanks for doing this work and just keep on doing it. How long can you keep this going for yourself, personally? Yeah. Ask God, don't ask me. I got no plans of not doing it. I got no plan. And do what? And do what? I don't know. Tell me what else am I going to do?

This is fun. I'm going to do it as long as I can. And East Carolina has graciously said, look, as long as you're breathing, you can teach here. So I said, okay. As long as I'm breathing, I'll teach there. It's been a great ride. I get to talk to guys like you. I would never have met you, gentlemen.

I would never have met you all. And it's just been a great ride. My gratitude is daily and continuous. And you guys seem like a wonderful couple of chaps. This is a good gig. Is this y'all's day job? Is this day job uh somehow it is somehow it's become that yeah pretty amazing right yeah yeah todd lewis told me that his day job was putting together conferences i'm like that's a pretty good day job so you guys are good good for you good for you we're

living the dream yeah you are you are. Speaking of Todd and All Things Open, you said earlier you only have recently not gone to the conference, but we're big fans of Todd, big fans of All Things Open. Me too. We were recently there, so listeners, if you haven't gone back and listened to the two anthologies or the several appearances in Change Load News, you should check that out. We were at All Things Open covering the hallway track.

Absolutely loved it. Big thanks to Todd and team for having us out there. And then Cameron for you. I mean, who was the person, Jared, that introduced us to Cameron? She didn't give me her name. Cam, you said you think you know who it is. Yeah, yeah. I'm pretty sure it was Cindy Harris because she sent me an email about it on LinkedIn.

She's a young lady. She was never my student, but she was like a mentee early on. And I helped guide her. And I didn't train her. She was training herself. But she's a good friend. And she's another success story in this space. I think she's Native American. I think Cindy's Native American. I think she's from that part of the world. And this has been great to her. And it's a good story. I mean, there's no downside to any of this. Absolutely.

Well, thank you, Cindy Harris, for mentioning Cameron to us and something that seems to have gone by the wayside, which obviously is not mainframes, COBOL, and the great opportunity here, the teaching you're doing, Cameron, and there are others in the space doing it.

But thank you so much for coming on here and just sharing with us what that is. It's been awesome. And hopefully inspiring some listening to this to take some of the paths you put out there and look into this world. But thank you so much for coming on the show today. It's been awesome. It has truly been an honor.

And I hope our paths cross down the road, gentlemen. You're a couple of fine people, and I really admire what you're doing. And I thank you for taking time with me. Thank you. Thank you, Cameron. That's it. The show is done. So, Mainframes, what do you think?

Are you back on the they are big bandwagon, COBOL as a language? Folks are aging out. Not many are coming in. But it is critical infrastructure for most, if not all of society, according to Cameron and the folks he knows inside this mainframe world. Something Jared and I thought about for this show was just to expose the opportunity here for those developers out there wanting to make a change, a move, go in a different direction. Mainframes are pretty cool. I think they're pretty cool. And if you agree, let us know in the comments.

The link is in the show notes. Again, a big thank you to our friends at Fastly and Fly and BMC. Hey, those beats are banging, banging, banging. Much love, BMC. Much love. And of course, much love to you, dear listener. Thank you so much. We love you for listening to our shows.

Thank you so much. Hey, stick around for Monday. Jared's got some cool stuff happening, but that's it. This show's done. We will see you on Monday. Game on.