Sam Knutson Believes Collaboration Helps Mainframes Evolve
By Reg Harbeck / January 16, 2016
Reg Harbeck talks with Sam Knutson, a prominent mainframer who has been involved in both the vendor side and private industry side on the platform.
Reg Harbeck talks with Sam Knutson, a prominent mainframer who has been involved in both the vendor side and private industry side on the platform. Listen to the interview via the orange play button or read the transcript below.
Reg Harbeck: Hi, I'm here—this is Reg Harbeck. I'm here with Sam Knutson today and we are going to get to know him a little bit better. He is somebody who has been active on the mainframe, worked for both private industry and software vendors, is a very well-known and active contributor at SHARE as well. Sam, welcome. Could you maybe give us a little bit of a background of how you ended up both working on the mainframe and attending SHARE and what they meant to you in your persona life and your career?
Sam Knutson: Okay so I started in the 1980s. I came out of a college education at DeVry that was focused explicitly on working with information technology so courses in COBOL, Assembler, using JCL, hands-on in labs and was recruited to come out and work for GEICO as a Programmer/Analyst I programming PL/1 on their auto policy system in 1987. So I started at the bottom and, as many people today, progressed through a number of different roles and a number of different careers. I believe I've had at least 17 or 18 different, categorically different, jobs throughout my career while working on the mainframe from various types of application development, capacity planning, working as a developer for an ISV, working as a team leader and then ultimately moving into product management where I am today. So you know that wouldn't have been possible if I hadn't come into an industry that had so many different roles supporting it.
Reg: Cool. Now, having sort of found your way into a mainframe career at some point, you found your way into SHARE. Maybe you could tell us a little bit about your journey into SHARE and, having gone into SHARE, into the various and numerous roles that you have had at SHARE.
Sam: Well, I used to tell people that I worked with computers when they asked what do you do. I would say, oh, I work with computers. Then I used to say I work with people who work with computers. Now I just say I work with people, so my career today is all about building really cool products to solve difficult problems. As you go through life, you evolve, but it’s always the people that are, I think, the really interesting and memorable parts. It was through SHARE that I found a community of other people.
Early on, I read publications, especially when I started getting into system programming. I read things like NASPA. It’s 30 years, I think, since NASPA was founded and they are still out there. They have a small presence and Scott published a retrospective in one of their magazines. But finding that online community so listserves like IBM-MAIN and SHARE. SHARE was this mythical thing that I heard about as an application programmer and as a junior system programmer, this gathering, where all the mainframers wanted to go. I don't think I actually got to a SHARE until the 1990s but it was a phenomenal experience. It was overwhelming the number of people. There was so much to learn and the connections that you made—then being able to leverage those to help each other solve problems and that evolved to the online experience where you were always interacting with people.
Being able to help other people share ideas, solve problems and, ultimately, I became involved with sharing software. You know, that was one of my passions was open-source software on the mainframe, the CBT tape, connecting with Sam Golob—that was one of the most significant events in my life—is finding a way to help the community. That was a way I could give back by providing a way for people to share ideas. You know Sam and you know Arnie Casinghino before him so I am excited to be part of that effort and now having been part of both private industry and working at a couple of different vendors over the last 30 years, I really can see the ecosystem as a whole. It needs to have that broad support of you know customer contributions, customers sharing ideas, working together to lobby IBM and vendors for their needs, vendors serving a robust market that is competing to provide the absolute best products and bring the best new ideas forth. That's what makes the mainframe valuable today is that it has been constantly reinvented both by IBM, vendors and customers over the years.
Reg: Now, one of the things that you referred to that we don't often think about is the CBT tape. You know that, obviously, the IBM mainframe is very much a business computer and SHARE when it was found in 1955 was founded by business organizations including universities, manufacturers, finance and government to get the most value out of available computing technology but that SHARE wasn't of course and isn't an acronym. It is what they do.
Sam: I love that slogan-
Sam: And they still have it on the sign.
Reg: Exactly and so one of the things that SHARE has done right from the beginning is SHARE not only input but actual code. Maybe if you could give some insight into sort of the history and the continuing relevance of the CBT tape and similar things.
Sam: So people ask about you know how can there be a CBT tape where there is no more Connecticut Bank and Trust. You can go to CBTtape.org and read all about the history but it was fundamentally a way for people to share code. It started with, it started small. It was originally distributed as an actual tape. The origins of sharing code at SHARE and in the community go much further back than that. You know, I learned about the bushel basket mods and you talk to Bill Smith, other people in the industry, and they tell you these stories of how things got started. That was fascinating to me. I love the idea of helping each other solve problems and so I was very eager to be involved in that taking it the web, facilitating this. So over the years we've used tapes, DVD's, CD's, principally now the internet. It is a worldwide, round-the-clock community effort and I love that that is still part of the community today—the act of sharing things that they have created that are of value to your peers.
Reg: Now what else would you say of value to your peers and really in some ways stretching right out to the whole world economy would you say you get out of SHARE and similar user groups because you obviously—you mentioned NASPA as well—similar you know ways of dealing with your peers as fellow mainframers?
Sam: The mainframe is at a very interesting state today and I think it’s—people need to get out and they need to interact not only with other mainframe professionals but to look at the broader changes in IT, e-commerce, the application economy. The mainframe has for many years operated in a very siloed set of organizations. It has been incredibly successful as the system of record for most of the companies that are represented here at SHARE but today what I am seeing is we're in a sea change where the mainframe—it’s incredibly valuable. It is the most efficient transaction-processing platform available but too many people want to just rest on the laurels of the past achievements. The mainframe is the most securable platform, scalable, efficient and unparalleled for transaction processing in business but that is not enough. It's not enough to be secure and stable. It can't be isolated; you know that is a sure way for the mainframe to become irrelevant.
For the mainframe to stay at the forefront of relevance to business, it has to also become agile, to be more responsive to business needs to be fast. Mainframe practitioners all too often have become the experts at answering no to every question. You can't have it. You can't have it for a year. You can't have it for 18 months, so when you compare that to what you see in the world today it’s not acceptable. The old paradigm used to be that big beats small, so companies could accumulate vast numbers of brick and mortar locations, of salespeople; they could dominate an industry and they could dominate their competitors. Many of the hallmark companies that you see today were successful in that model but things have changed. I mean we have seen the rise of the internet, of a global economy and you have seen disruption on a scale that was unimagined years ago so, Uber disrupting the transportation industry, Airbnb disrupting the hotel industry.
The companies that today own mainframes—and these are typically Fortune 1000 customers—¬ need to fend off that disruption and, in order to do that, they want to be both big and fast. They want to leverage the assets they have, often customer data, years of data about buying patterns, business intellectual property that has been built up because they have systems that you can't go buy. If you have spent 20 years, 40 years developing banking systems that embody all of the logic of your business processes, that give the value that you uniquely provide for your customer, you can't go buy that. You can't go get that in the cloud but you have to unlock the potential so you have to be able to deliver to the business, to customers, a great customer experience very rapidly.
So one of the things you are seeing here at SHARE, and it started in San Antonio, was the DevOps track getting a lot of attention. I'm seeing a tremendous number of people who are coming from mainframe AppDev, mainframe systems administration and they want to understand. They are in a discovery phase so they want to understand what is DevOps; why is it relevant to me? Am I already doing it? The reality is there are other ways of developing software that are focused on shorter interactions using, you know, agile methodologies so that you can be more responsive to the business, faster to deliver with results that are going to resonate more closely with customer's needs. As mainframers, we need to embrace this change. We cannot go forward into the future and be successful simply by pointing to what we have done in the past and saying if we continue to do that, we'll be successful because the mainframe must be secure, must be scalable and efficient but it must also be agile.
Reg: Thank you so much for this Sam. Maybe as some closing thoughts, if you could share with us your thoughts on the present and future direction of the mainframe technology, culture, the whole ecosystem and any other additional thoughts you want to just sort of conclude with.
Sam: IBM has done a really great job of evolving the hardware platform today. The z13 is an engineering marvel. I fully expect that the next generation of that technology will surprise many people in its leaps forward that it will take. It needs to be surrounded by a vibrant vendor ecosystem so there are many unmet needs. People need to look at the mainframe and say, ‘How do I bring this into my overall development,’ because that is where the value is released for customers. AppDev is the single most critical area for the mainframe. We have the crown jewels, the best customer data, all of these business processes that have been written into applications, the digital DNA for companies, but it needs to be unlocked now.
You know we've done web enablement; now we need to do microservices. We need be able to work with the systems of engagement in new ways. It is in being successful there that we’ll carry the platform into the future and I think SHARE and the community plays a big part in fostering the development of new ideas, getting people to embrace change.
Much of what we're talking about in the DevOps track here is a change in culture, so tools are absolutely essential for people to unlock some of this value. It is equally important for them to embrace a new mindset, to be willing to look at you know working iteratively to being engaged with customers. Waterfall is dead. It is absolutely dead and as we go forward, you know people need to look at changing their mainframe silo development and administration organizations to be part of a larger hold for the company. I think SHARE plays a big part in this in connecting customers to each other so that we can share proven practices. You know that has always been at the heart of SHARE is sharing ideas so people coming and doing user experiences about how they are doing development, how they are implementing DevOps, how they are getting value from making these changes and incorporating their mainframe into everything else that they are doing.
Essentially customers need to mainstream the mainframe. They need to make it normal in a good way so that it’s approachable by developers who are just entering the workforce because we want to pass the torch. We really want the mainframe to be the platform of the future and in order for that to happen, we have to make mainframe development and administration accessible and equal. The mainframe has to be a first class DevOps citizen. It cannot be siloed. It cannot be put off into some sort of slower mode of operation where you essentially put it in a box and just get data in and out. That is a way that it will become completely irrelevant.
Reg: Well, thank you very much, Sam. This has been really interesting and informative. I really appreciate you taking the time.
Sam: Thanks, Reg.
z/OS / Linux on IBM Z / z/VM / z/VSE / Article / Community / zTalk podcast
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.
See more by Reg Harbeck