Bob Rogers on His Longevity With the Mainframe
Reg Harbeck talks with Bob Rogers, who started at IBM in 1969 as a computer operator, about what role the mainframe has played in his life.
Reg: Hello. I'm Reg Harbeck and I'm here with Bob Rogers, who is one of the people who has really made the mainframe what it is today to talk to him about the mainframe, its context and what role it has played in his life. So Bob, maybe if I could start by just getting you to introduce yourself and tell us what the mainframe has meant in your life.
Bob: Thank you. By way of introduction, I started in the mainframe in 1969 as a computer operator and continued for 43 years working for IBM before retiring, was retired for approximately two and a half years before rejoining IBM again because I just missed all the fun. The whole time I have been on the mainframe platform, most of it on the MVS side but more recently on the VM side. It has been very important to me. While I was retired, I found that I really didn't have that many hobbies or anything to occupy my time because I was one of those really unfortunate individuals who really liked their job and treated their job as their hobby. I just had so much fun.
Reg: OK. So basically in those 43 years how would you say the mainframe other than being a source of income impacted your personal life?
Bob: Well the biggest impact is that I have friends all over the world, all over the United States, people that I've spent a lot of time with, people that I've had fun with, people that I've solved problems with and that has been very enriching. That is my community. Most people, I don't know, they join the chamber of commerce or the Elk's club or something like that. I joined SHARE and through the SHARE organization and its members and even other customers and software vendors not through SHARE, that's really where I have spent my social life.
Reg: So then, you would say that probably coming back to SHARE is as much for your own personal satisfaction as for your career and for your contribution to the mainframe context?
Bob: Well, let's make it clear. I have no career now. I had a career. I had a good career. I modestly say that I think I have made substantial contributions to the platform and I had a wonderful time doing it. Right now, I would like to continue contributing based on my background experience and you know the things that I have seen that worked and didn’t work but a lot of it really is social. I was so very happy to rejoin IBM as a retiree supplemental because that means that I can get into the IBM building and get access to all my old MVS friends in Poughkeepsie. I also occasionally drive out to Endicott to spend some time with my VM friends out there, not nearly as frequently, but certainly as enjoyable. Of course, there always is the work. There is finding the problems, defining the problems, solving the problems, testing the solutions and so forth as there always was but the social aspect for me has been very important.
Reg: Very interesting. Now you mentioned that you made some contributions. What would you say are your favorite contributions that you made to the mainframe during your time at IBM?
Bob: I think my greatest contribution to the mainframe platform was to cause there to be a very smooth, safe and intelligent transition into the 64-bit world that by chance and, if you want, I'll briefly tell the story.
Bob: I became the leader on the software side in IBM mainframe for this transition into the 64-bit world. Because we made a transition in which we delivered operating system value, real value to customers as soon as, in fact even before, the 64-bit processors were available, it gave no wiggle room for our competition to say what they did when we came out with ESA which was, “This is really great stuff but by the time you need it, we'll have it,” because there was zero window. As it happened, the PCM competitors for the most part left the mainframe marketplace arena and IBM was left alone standing as the only manufacturer that built—I'll call it ‘real mainframes.’ There have been a number of emulator enterprises. The way I became—I got this position—was sometime early on in other platforms going to 64 bit, a woman came to me who worked with what is called market support and she said ‘Bob, we need like a short whitepaper on why the mainframe doesn't need 64 bit because all of these other guys are touting 64 bit, how great and wonderful it is and everything.’ So I wrote a one-and-a-half page paper explaining that we’re not really a 31-bit system because we have ESA data spaces to give us additional virtual and we have expanded storage to give us more processor memory. I claimed that we were probably about a 35-bit system right then and that the mainframe, being the big guy, had to grow beyond the 31 or 32 bits earlier before it was possible to build a microprocessor chip that had 64-bit capability. If you couldn't build a 64-bit machine, we did these kind of ad hoc solutions, which we were using to buy to the time until we actually could in the year 2000 build a 64-bit machine. So I had written this whitepaper on why we don't need it but there was a young lady who was in that department at the time and then she went off on some assignment. When she came back, she became the manager of MVS design. She knew that I knew something about 64-bit architecture and you know the engineers at the time wanted to build a 64-bit machine so they needed an architecture and an architecture needed to be good for MVS so she asked my manager if she could have me on loan like 10 percent of the time for a short period which ended up being 125 percent of the time for about four years. We got the job done and we did it in a way that was very smooth. I will tell you another anecdote.
Bob: I'm at SHARE, you know we are at SHARE reception sort of thing. Somebody walks up to me, puts his hand out, and says, “You're Bob Rogers? I want to shake your hand. I want to thank you because my manager asked me to do the conversion to 64 bit, then he gave me a big award for doing it successfully, and there really wasn't anything to do.”
Bob: So he thought that was good and that was one of my major objectives was to make the 64-bit system easily consumable. Apparently, we did that and people rolled right into the 64-bit world without much difficulty.
Reg: Now I have to ask you. Back when we went from 16 MB to the full 31 bit of addressability, you know then we talked about 16 MB line. Then of course, when you introduced all that extra addressability with 64 bits, they talked about that big empty and theoretically unaddressible space depending on whom you are running with as the bar. I am wondering if you know why that term bar was chosen because it makes for great puns.
Bob: The term bar was actually coined by Dr. John Ehrman early on in this after I had taken the assignment and got my thoughts together I presented to what was then the mainframe design council where all the designers of the main hardware and software products would come together at one of the laboratories and then we would have a week of meetings. They threw me on the end of the agenda starting at 6:00 in the evening to tell them what I had in mind about this transition to 64-bit. John Ehrman came up with term bar and I had so much respect for John that I said this is good. I'm going to keep this rather than have two different lines. Now we can talk about the line and the bar. There is some argument as to whether the bar is actually the line at two gigabytes or if it's the whole; there is a two gigabyte dead zone that is above two gigabytes. But it doesn't matter.
Reg: It is so much like SHARE because sure enough, you go downstairs this evening and there will be a line at the bar.
Bob: Just don't get caught under the bar.
Reg: Oh yes, exactly. Well a few too many of those beers at SHARE would be a problem. Tell me just a sort of closing thought. What are your expectations and hopes, best hopes, for the future of the mainframe?
Bob: Well you know the mainframe has been quote “dead” for longer than a lot of platforms have been around. It was predicted in 1992 that the mainframe would die, in 1995 that the last plug would be pulled out and you know that is so wrong that I can't help but think that there has got to be at least five, 10 or 15 more years of mainframe vitality, that the mainframe platform based on the work of the IBM engineers and the IBM software developers has really been able to keep up with things, to implement newer technologies onto the mainframe to augment what is already there. Like some people think ‘Oh the mainframe. Isn't that green screens?’ Well, yeah it's green screens but it is also Web serving; it's also just about anything you can imagine in addition to still supporting green screens. I have had the opportunity at previous times to give a little presentation, to be able to say a little bit more on this subject of view of what will happen with the mainframe as it goes out but right at this point it is the only thing that can run Western civilization so I am confident that it will not be very soon that the mainframe is replaced by something else. The other platforms have made every try they could. They've you know spoken ill of the mainframe but yet those who really understand large enterprise computing understand that the mainframe still is the only thing that can really do the job.
Reg: Cool. Well thank you very much, Bob. It's been a real treat to have this chance to listen to your thoughts, your history and your thoughts about the future.
Bob: Thank you.
Like what you just read? To receive technical tips and articles directly in your inbox twice per month, sign up for the EXTRA e-newsletter here.
comments powered by