John Mertic on the Open-Source Renaissance
Reg Harbeck talks with John Mertic about the history, future role and community impact of open-source technology.
By Reg Harbeck08/03/2020
John Mertic: Project.
Reg: Okay, and I guess maybe part of it as I ask you to introduce yourself and tell us how you ended up in such a position. I know your background technology and such. Also, explain the difference to us the difference it being a project or a foundation or just what the nature of it is. So, John, tell us about yourself in your role.
John: Sure, so my role is to help support, facilitate and be a part of the driving of success of the Open Mainframe Project. You know, my career has been one that has really started in the software engineering realm. I was a software developer for many, many years but also at the same time I was starting to spend a lot of time in open-source communities and I really was drawn to the dynamics and the people and really the great innovation that was happening in them. I really started pivoting a lot of my career to spending more time within there from being a contributor in that space to helping be a leader in that space while at the same time helping build a lot of my acumen around partner relationships, ISV management and business development. Which kind of all came together in my role at the Linux® Foundation, which I came on board with just under five years ago helping lead a number of our open-source initiatives and in particular the Open Mainframe Project which is, since it's a collaborative effort in one of the well over 300 different projects that we host here at the Linux Foundation. Our projects are of all various natures from very small unfunded collaborative efforts to huge large scale ones like the Linux kernel and Kubernetes and Node.js and even more vertically focused ones in the energy, motion picture industry, automotive industry, so on and so forth there, so it's a real, real mixed bag of stuff we have. A lot of our role—we sometimes joke about calling ourselves the janitors of open source, but I think really what it comes down to is we're here to help our communities be successful. They know the technology area. They're subject matter experts. They're really great at a lot of those pieces. We come and help fill in those gaps and add that extra support and services that you don't think about needing an open-source project needing to be successful but it's really, really a key piece of it, so things like ecosystem development, things like events, things like legal and governance, things like finance in a lot of cases. All of these pieces and more are just really crucial to a community operating, especially one in a vendor neutral capacity and that's really what we're here to do. We're here to do that. We're here to serve our communities. We're here to ensure that they have the tools to be successful and we're here to facilitate that along the ways and in a lot of ways we try to teach communities to fish VS be the fishermen for them.
Reg: Now of course, open source and open software of all kinds is such an interesting topic. Because on the one hand, too many people mistakenly think it's free and of course even Linux isn't free anymore if you want to get a distribution, but the history of open-source software goes way, way back to before the IBM System 360 mainframe even with the SHARE operating system, for example. I gather you've had a chance to really immerse yourself in that history. How would you characterize its role in really the history of humanity?
John: You know, it's a really interesting way to look at it because I think by very nature our society has advanced the most when we've had the avenues to collaborate and the abilities to collaborate. Even if you go back to the Renaissance, what really kicked that into gear was the printing press in a lot of ways, and that ability to share knowledge broadly and immerse yourself into it, but also build collaboration within it, and that's always what has kicked major shifts in our society forward. As computing and technology came to the forefront, we've seen that start to happen again and I guess the early bits you had back in the 50s with SHARE where these new things called mainframes and these big large computers that would take up entire rooms or floors of an office building. People were just wrapping their heads around what to do with them and how to make them productive for them and events like SHARE were just these opportunities for people to share tips, share ideas, share code back and forth. That spirit started there and really, the modern open-source movement, we point back to that as that's really the pinnacle starting point. You begin to draw that entire line to where we're at today, through in the 80s of the rise of free software in various academic circles and driven then from that perspective to open source coming in the 90s is coined as a term; Linux is something bridging sort of between both of those. The idea is how can we speed up collaboration, how can we get a larger group of people contributing, getting ideas, driving it forward. Because I think the one thing we've noticed in human history is when you have more people at the table, the outcomes are better and that even stretches across where you see a lot of the big focus is not just in technology, but I think doubled down in open source even around diversity across a number of different aspects. The more we know of having different people at the table, the ideas are better, the outcomes are better, and we're able to drive faster as a society and really build some amazing stuff. There's a lot of people that have compared to where we're at with open source now is to the next Renaissance, which is kind of a fascinating sort of way to think about it in the context where you said that global, that historical aspect.
Reg: A Re-Renaissance.
John: Yeah, yeah, I mean, think about it like we've had this much technology capacity at our hands. We have an access to this much knowledge of materials and just sort of building blocks. You can get up and going if you have hardware, you can build an entire open-source stack of software to build pretty much any sort of modern infrastructure. If you look at the so-called ”FAANG” [Facebook, Apple, Amazon, Netflix, Google] companies out there, their infrastructure is built upon open source. That is what accelerated all of that growth was Linux and various open-source technologies that over pinned that as well and that is what has pushed those companies so quickly forward.
Reg: That’s interesting, and it's easy to imagine an open hardware platform like Intel® which doesn't require Windows®, or even better Raspberry Pi. But then you think of more closed proprietary platform like Apple or especially like the IBM Z® platform and I'm curious: How do you see them coming together, because clearly that's right exactly where your sweet spot is?
John: Yeah, it's really interesting because I often, being involved in the mainframe community, I've been able to take an appreciation for where lineage of these architectures begins to come from. One thing that has always struck me about the mainframe architecture: It is designed with the same design principles that it has been since the first of the modern mainframes rolled off the line in what? 1964, something like that? That same principles of we need a computing platform that is secure; it is reliable; it is scalable; you know it is high performant; you know it is all those things that you need from something that is going to be a critical part of your society. But you just don't need two of those; you don't three of those; you need all of those and you need all of those turned up to 11. Not to go all Spinal Tap on you here, but I mean, as I talk to folks about mainframe and say it's not what you saw on the Space Ship Earth ride at Disney. Yes, that is what it had been but if you look at the modern one, it keeps a lot of those same design principles along the way. Security, scalability, performance, availability. That's what these are all designed so if you look at it from a heritage point of the architecture, everything falls in line in the same way in Apple. Again, Apple has moved to a lot of more commodity infrastructure although they have a very tight set of how they pull all this together.
Reg: They're running on Intel too now.
John: They run Intel too now, and rumors of going to different architectures, but the design principles that they have struck around are all about empowering the user, making it as simple and straightforward of an experience, design simplicity, to really sort of attract those professionals that are looking for that. They're not looking for something that they necessarily need to look like a big bizarre-looking brick or has to have a whole bunch of things sticking out. Yeah, they look for those very aesthetic design principles from the hardware to the software to the alt pieces. So, they've designed along that principle. Mainframe is just the same thing. They designed along the principle of, you need all those turned up to an 11? This is the hardware on the planet. This is the only hardware on the planet that's going to do that. Every time a new Z box comes out, all of those principles, that's how everything is judged against, which I just find a very fascinating thing. I think that's a unique thing that the mainframe community can hang its hat on.
Reg: Yeah. Well, it is sort of interesting looking at the history of open source on the mainframe itself. On the one hand, we have the shared CBT tape that had all this shared software and then you have MVS right up to MVS 3.8 when it went object code only, and then suddenly we got the turn of the millennium and you have Linux arriving and first of all running bare metal, then under VM and now in containers with Kubernetes. Where do you see that happening next?
John: You know what? There’s a slide that we have around Linux's dominance across multiple industries: smart phone, share of supercomputing. I think every one of the top 100 supercomputers are all running Linux. On the mainframe side we say, I believe the numbers from IBM have been somewhere around 90% of mainframe customers are running Linux to some degree, so you see Linux dominating across there. But I think the overwhelming trend is open source is empowering because a) you know this is, we're not the state-of-the-art modern technology is being developed. It's developed in open source and it's made available with a DevOps mindset, putting the developers first. Which is just a mindset difference that you might have turned back a couple of decades and not seen that per se being driven in some sort of circles there. But I think also open source has this ability to draw together different silos of computing architecture that one might be using to really power the modern enterprise. Smart forward-thinking enterprises are always understanding that their best advantage in an environment is something that's going to be heterogeneous, meaning they're not going to buying all of one hardware. They're not going to be putting all their stuff in AWS. They're not going to be really putting, so to speak, all their eggs in one basket.
John: But they're going to recognize that all of these things have different strengths, different weaknesses. How can they use all of these to their advantages? How can they use all of these to be compellingly different? Even where we see companies will use that almost a differentiation, VS, their competitors, their ability to have a technology stack that enables them to execute and serve their customers better, is a competitive advantage. We see open source as kind of as little bit of that leveling appeal. It's enabling people to get to that point faster than they ever had before. You don’t need a vendor to be that person. Even legacy organizations and companies have turned themselves into software companies because open source has opened that door for them.
Reg: Well, yeah. Talking a bit about horses for courses, as you look at that. Now one of those horses of course is proprietary and another one is open source. It sounds to me like what you're saying is two need to work together almost like bricks and mortar. As you think through that paradigm and project it 10, 20, 100 years into the future, where do you see us going with that paired journey of open source and legacy and closed proprietary as we build the future?
John: So, I think the one thing we've always seen is the layer of where commodity is always something that's coming up, right? Decades ago, it was at the OS level when you had several proprietary UNIX® operating systems or BSD, you had Windows, you had VMS and you had, even earlier, Mac OS. Then over time as Linux came along, it became compelling enough that that was a great sort of commodity base to begin building your next layer up on, right? We've seen this a little in the web browser space. We had IE and Netscape and all sorts of different engines out there. They sometimes had proprietary rendering engines, and then we've seen Web Kit become a sort of de facto rendering engine that so many web browsers are employing from Safari, Chrome, Opera, even Edge using derivative of Chromium there. And then from there, you're seeing the level of innovation is happening higher in the stack. We've seen that on the Cloud front; we've had all of these Cloud vendors that originally were building a lot of their own stacks. Then you saw some convergence around technologies like Open Stack and you then saw it come up the next level of the stack to an infrastructure, the higher service as a service where technologies like Cloud Foundry, and now you've even seen it go to the level of where containerization and service deployment and management where you see everyone you know rallying around Kubernetes. I think what becomes really interesting is you see natural places in the market where there's innovation happening in different pockets and as there is convergence to bring this into communization, that becomes sort of that base level layer of open source. That's frankly where we see open source being the most successful. It's not when it's out there having to be at a higher level, a higher value end of the stack where there can be a ton of differentiation, but it's that base level of the stack that it becomes really, really, really critical for folks to build around. We see that time and time again out there as that commodity layer moves up, the open source is the key area, the commodity that everyone rallies around. All of these solutions are built around, all the collective effort is going in there and then that's becoming stronger. Then we just naturally see that level up and up, so I guess if you look at in that perspective, there always ends up being a space for commercial software, because at the end of the day there's specific value add that you can deliver in a channel in that way. One of things when I talk to companies around open source and what they should be open-sourcing, what they should be leveraging out in open source; sometimes the analogy I tell them to think about is to think about what is the unique value proposition that your company is looking to deliver in the products that you're putting out there and where do you have within your engineering team, staff, and resources that you have that unique expertise that that is a differentiator? Focus your energy in that area; focus on delivering on that and for the things underneath it, that's commodity. Leverage what's out there, leverage the commodity infrastructure that's out there already that you can tap into. I remember talking to a company years ago and they had their engineering team who was spending all of this time on database connection libraries between I think it was PHP at the time and MySQL and things like that. I'm like, you don't need to rewrite these. These have been written already. That's not an area where you need to focus your energy, you’re a company in this space. Focus your level energy at that level and benefit from what the community is driving it at the bottom level. I mean, still feed back into that open-source community of what you're finding in some of your use cases and contribute to that common good, but that's not a level you need to be solely innovating in. So yeah, it's sort of a mindset to think about that.
Reg: So that the proprietary gets more value, by not reinventing the commodity wheels is what I hear you saying.
John: I think that's a great way to articulate it. It builds from it. It uses that as a base point and it uses that to support functionality, but that's not what the end level deliverable is. The end deliverable is something that is specifically fought with the customer in mind of the finished value that it's trying to drive. A standard relational database, for example. Yes, there can be potential value in there, but a lot of that layer is getting so commoditized, the value might be of that end user-facing application that's using that data and how it's uniquely driving value. That is just a little bit of the mindset that you have to think of and that's often just also a good differentiator for companies that might be building some of this commodity stuff and realizing, boy we're just bending over backwards supporting this and it's not helping us close deals. It's not the thing that my sales people are leading with out there. It's something I need to get into these deals, but it's not something that that's what people are buying. I think that's a little bit of the mindset that needs to happen. That's the mindset that a company that is actively thinking about their open-source strategy that they really start to dig into.
Reg: Cool. Now, we started this conversation sort of talking briefly about your journey into this role that you have. Given the amazing and really indefinite future potential for open source as you look at your career, because my perception is, you're still pretty early on in your career. What are some of the career highlights you can imagine yourself having as you continue on this journey interacting with all of this?
John: My voice must sound a lot younger than I am. That's positive.
Reg: I'm convinced you're younger than I am, but we won't compare notes on this recording.
John: That's a really good question. I have always been one in my career to think about it is, on one half there's always opportunities where you can take leaps within your career of going forward, and I think it's really cognizant to know where those are at and where those are the opportunities to kind of leap. I guess a lot of where I think of it as where the stage of my career is at is, what are the skillsets that I can continue to evolve to help this common greater good, because I think where I came to is I love spending time in open source. I love working with communities. Communities are hard to work with sometimes. It's not all rainbows and unicorns, but there's so much reward of just seeing the benefits and the people that are impacted and it's just a unique way to do it. I'm sure you can do that through driving for-profit products and in other sort of ways, but for me, I guess I'm able to connect to that so much easily there. I think where I see a lot of where I've really turned a lot of my mindset and especially as I've realized that my coding skills are certainly not what they used to be, is what are the ways that I can spend time with communities and help them understand where they're willing to go and be an enabler and facilitator towards that change and towards that direction. What that actually means as a specific job title—I mean, I think that's what the future gets to tell all of us but if I would put into words, sort of a design principle or a design pattern of how I'm thinking about it, it's what can I do to serve? What can I do to help add value? What can I do to advise? What can I do to facilitate because there's great people out there doing great work, and I think the leap I'm seeing that I can provide value is how can I help take that work to the next step and not necessarily on a technical contribution but building out that ecosystem, ensuring that they can operate smoothly and cohesively, sort of all of those functions in there that just really make a community thrive and honestly make a community successful or not. There are so many examples out there in open source where a community might have amazing technology, but if it doesn't have the support around it to take it to that next level, it dies. Then there's also examples out there of open-source projects that have okay technology but the ecosystem around it is really what makes it thrive.
Reg: So, it's all about people.
John: It's all about people. I think somebody asked me in an article what's the best and worst part about open source, and I said people both times, and that's okay. I think that's a great thing of it because that's really what this all comes down too.
Reg: Cool. Well, any final thoughts before we sort of finish up?
John: No, I don't think I have. I love the direction a little bit of these questions here. I think it has sort of taken it in a little bit of a higher level which I appreciate but no. I really enjoyed talking to you and I encourage a lot of your listeners here if you're not as in-depth into open source, if this sort of an area you've always been interested in but not know a lot, I definitely would check out some of the resources provided out by the Linux Foundation broadly. We have a number of groups that focus on a lot of different disciplines from the to do group which focuses on companies building open-source program offices. We have a training arm that really adds training in a lot of open-source technologies and just open source in general. There's a Chaos project which focuses on community health metrics and then just a lot of our various technology space projects. Like I said, we have over 350 of them out there and they really dig into just all sorts of fascinating areas with fascinating people. Particularly the two that I—the couple that I work on here Open Mainframe Project, which we discussed earlier, but I also work with the motion picture industry and the Academy Software Foundation, which is quite fascinating and the big data industry throughout EPI.
Reg: Cool. Well, thank you so much, John. This has been fascinating. I really appreciate you taking the time.
John: Likewise, thank you.
Reg Harbeck is a mainframe enthusiast who has worked IT and mainframes for over three decades. He's the chief strategist at Mainframe Analytics ltd.