Ray Mullins on Assembler, COBOL and PL/1
Reg Harbeck: Hi, this is Reg Harbeck and today I'm here with Ray Mullins who is a friend of mine and fellow mainframer who I've known from SHARE and various other parts of the mainframe ecosystem for quite some time. It turns out we have a lot of common interests, including languages and of course the mainframe. Rather than telling everybody about you, why don't I get you to tell us about yourself? How did you end up becoming a mainframer?
Ray Mullins: Well I'm older than I look and I started in college which is back around 1980. Back then, mainframe was the only thing so I took plenty of computer courses. I first started working with computers in my last semester of high school and realized that this is the world I wanted to be in. Gradually from there I started taking the courses on the IBM system and you know with the OS VS/2 and then eventually I got a job although my first job was not working with mainframes. It was nascent PCs as the IBM PC had just come out; however, they discovered I had Assembler knowledge and I actually wound up writing some Assembler code for an in-house job scheduler.
Reg: Now how did you pick up Assembler knowledge? In high school?
Ray: In college. I took a course but I finished all the assignments within five weeks.
Reg: Oh wow.
Ray: And this was for a standard semester.
Reg: So this was when you started discovering that you had a talent for a computer languages.
Ray: Especially for Assembler. I had done very well with COBOL and PL/1 and other primary languages at the time. Then you know things happen and I wound up taking an application programmer job. It was batch CICS COBOL but I knew from my previous experiences in college that I was a technical person. I had done system administration work at the college and I wanted to become a systems programmer. Eventually I was able to get that opportunity, enjoyed it very much and then I realized that I was really liking the vendor side of the world because I would call in problems and I had great conversations with the particular company. Eventually I took a job with that company.
Reg: Now I'm trying to think of whether would have been before or after the turn of millennium. I'm going to guess it was before the turn of the millennium.
Ray: Yes, it was. This was 1991.
Reg: Okay.
Ray: So I worked in customer support for almost five years and as things happen in the industry I was laid off but I knew at that point that I wanted to get into development. I held out a little bit and was able to get a development job in Assembler for a company. It was Platinum Technology and then from there I wound up going back to the company that had laid me off earlier but in R&D; ever since then I've been—so we're looking now at 24 years in product development mostly Assembler. I've dabbled a bit with Metal C as IBM has been pushing it and I code a lot of Rexx.
Reg: One of the greatest languages ever invented.
Ray: Yes, it was and I count Mike Cowlishaw as a friend so that's good. I've worked on MVS, OS/390, z/OS the majority of my career, however I have also done development on VSE and a little bit on VM as well including some non-IBM mainframe operating systems.
Reg: So now how many different Assemblers? I gather obviously the vast majority of the Assembler you've done has been System/360-descended mainframe Assembler. Are there other Assemblers you've enjoyed working on as well?
Ray: Not professionally. I dabbled back in the early days of the Intel during the PC era. As it turns out all of the non-IBM mainframe platforms that I've worked on have been IBM compatible.
Reg: Oh, OK. Interesting. Now of course you're a person who likes languages. On the one hand you like a lot of computer languages; on the other hand you like a lot of people languages. Tell me. I don't know if you can tell me off the top of your head how many computer languages but maybe you could just mention some of your favorite computer languages that you know.
Ray: Of course Assembler. I like COBOL in spite of its wordiness as the standard has gone on, the newer standards you can do a lot with COBOL. PL/1 is a nice language because that is how I learned the block type languages – procedures, which was helpful when I started working with C, learning that. One of my favorite niche-level languages which you don't see much of anymore is APL. That was the funny symbolic one where basically the goal was to try and write a program in one line.
Reg: APL makes C seem self-documenting.
Ray: Yes. That is true but you know I've done Perl as well. I started working with Perl about 15 years ago. It has a nice string-like capability, similar to Rexx. That's really where I started working with it a lot and sometimes if I'm not on a mainframe I will use Perl just to write some quick and dirty scripts.
Reg: That makes sense. Now of course as I mentioned you are also into people languages. I don't know very many people who know as many languages as you do. How many are you at now?
Ray: Well let's say I speak English rather well I think. I have a dabbling knowledge of French, Spanish and a little bit of Portuguese which I've lost because I really don't get an opportunity to use that at all. It turns out my backgrounds, my dad is half Argentine. My mom was half German.
Reg: Because you speak German as well, right?
Ray: Ich—Ja. Ein bisschen and my grandmother, my father's mother was upper class Argentine, and was very well educated, and she spoke more languages that I did.
Reg: Cool. Now bringing that all together of course, you found yourself in a pretty important set of roles at SHARE in terms of helping them move forward including being in charge of the languages project. How did you end up with such responsibilities at SHARE?
Ray: Well—
Reg: How did you end up at SHARE?
Ray: I'd always wanted to go to SHARE and the opportunities never presented themselves unfortunately. I did attend a GUIDE back in the late 80s when it was in Anaheim, since at the time I lived in Los Angeles and it kind of worked out that way, but finally a prior employer gave me the opportunity to speak. I started coming and I wanted to volunteer. Actually, do you remember? I did give a presentation a couple of years earlier from that first one that was for the open-source z390 Assembler project which actually for me was a life event because many people who had known me over the years over IBM-MAIN, VSE-L, the mailing lists, the internet mailing lists, actually came to see me present. They wanted to meet me in person and that to me really affirmed my career choice in the mainframe. I wanted to work with SHARE and I first started a little bit later as a project officer in the MVS group.
Reg: OK.
Ray: But I learned at the time that eventually the languages group had become kind of moribund due to some health issues and they needed somebody so I stepped up and took that mantle, originally as a project officer and then I was asked to become the project manager.
Reg: Cool. Now of course that's not the limit of your involvement with SHARE. You've also been on a few committees and a few other things. In fact you even have done some writing for people, particularly about areas of interest to yourself such as when we had SHARE in Sacramento.
Ray: Yes, I was very happy to do that because even though I am a native Angelino, I've moved to the other part of the state. I love the Sacramento area and I'm very enthused about certain things. I believe one of the articles you're referring to is my wine article because our Sierra foothills wines are excellent.
Reg: Cool. Now let's see. I'm just trying to think. There's some other things you've also written articles about. I guess you've certainly written some stuff about languages. You've sort of got a series of articles you've done on languages haven't you?
Ray: Yes. Unfortunately due to life things that's kind of on a hold at the moment but I do want to get back into that where I've started writing on a history of languages for the SHARE newsletter.
Reg: Cool. Now as you look down the road, I mean one of the things—you and I are sort of contemporaries and we're sort of at that point where we've got a lot of years behind us but still a lot of years ahead of us if we want and enough perspective to be looking forward to seeing where we would like things to go. Where would you like to see the mainframe, your career and your role in the mainframe ecosystem to go?
Ray: To be fair probably—I'd like to continue with what I'm doing. I am a coder. I've realized over the years I will not be a manager. I do not like that. During the recession I went back and got my college degree because believe it or not when you're offered a full-time job at age 20 that's paying $20,000 a year in 1983 money—
Reg: Wow. Nice.
Ray: —You kind of go for that but I went back and since I was working full time at the time I did it on the part-time plan so it honestly took four and a half years for me to complete my upper division but I did. I kind of took the easy way being a business major within management information systems (MIS) concentration because it was easier for me with knowledge but one of the things that I learned from my fellow students, most of whom surprisingly were older than the traditional—you know we think of traditional matriculation of 18-21 in four years.
Reg: Yeah.
Ray: Well quite a few of them were also, like me, doing it on the part-time plan or they had come back in and had started later. Quite a few of my classmates were in their late 20s into early 30s surprisingly. Now Sacramento of course is the capital of California. There are many state data processing agencies there. There are two major mainframe shops for the state but one thing was that as some of the students were out looking for jobs, they would find these mainframe IBM z/OS jobs and they were really unsure about it. At the time there was one part-time instructor evenings and I'll say this. Usually the evening instructors, the part-timers are much better because during the day they're still dealing with everything and unfortunately I have seen the walls of academia block things. They start thinking too much on themselves but the faculty there did know of the mainframe and once they knew that I was part of the mainframe, there were times where they would ask me questions if I could fill in certain things from that area plus we had this one instructor who—he was a Greek immigrant. He learned computers in Greece but he knew the opportunities were better in the United States. He was able to get over here and they saddled him in his first job for one of the big eight accounting firms then was Unix System Services and he actually is very good at it.
Reg: Oh. Cool.
Ray: So between the two of us, we were able to fill in these gaps but one of the major problems that was very obvious there was that these students had no way to learn or experience the mainframe in the situation. Unfortunately for many years there has been a high-cost barrier of entry into the ecosystem for students, hobbyists and people who are interested.
Reg: Oh yeah.
Ray: Yeah and people kind of like myself who have some product ideas but no money. We aren't the types that can go pitch to venture capitalists and start ups and that sort of thing. There are gray areas for getting into that but right now if somebody wanted to purchase a system, you are going to be putting out probably over five digits or you could rent time at the IBM Dallas Center but you're still paying probably almost $5,000 a year. I think most of us cannot afford that, certainly not a student.
Reg: Right.
Ray: Unfortunately for many years the case was you were either in a gray area or you were out of luck or maybe you worked for a company and you had good management that said hey, you know I'd like to get into this and they would find a slot for you.
Reg: So what I hear you saying is that one of the important things that has to happen is the mainframe needs to be more accessible not just in terms of being able to find a mainframe you can work on but being able to afford the education and the accessible mainframe to try stuff out as a single contributor.
Ray: Yup. I do have to say there is the Master the Mainframe competition which is very popular. I recently was at a session where two graduates of that eventually got—they got jobs eventually through that but they were introduced to the mainframe ecosystem through the program; however once the program is done, there is really no option other than you can get an audit account as they call it on the system but you're really at the whim of what's going on in the contest and then every year they wipe your stuff away so you either back it up or recreate it. IBM has finally recognized there's a problem and I will say that they're taking steps to get around this so that people will now have a much lower barrier to get into this. I would love to see these opportunities come to fruition. I hope that we can get this done because one of the problems that really was self-inflicted is, by having this high barrier, as people somewhat older than us retired there wasn't new blood coming in because people in the universities or people interested in computers, period, couldn't get in there. It used to be that many major colleges in their computer programs taught mainframe skills and they used the IBM platform. Some things happened in the 80's where basically cheaper options such as Unix came in.
Reg: Right.
Ray: And then eventually you came in with Linux which as an operating system is free. The hardware isn't but the hardware was much less expensive, so during the 80s into the 90s, the mainframe goes away and we bring in Linux. Then of course Windows takes off and because of the high demand there, you have the universities teaching these other operating systems and not the operating system and platform that seriously runs the world as our friend Bob Rogers said. If mainframes disappeared, civilization would collapse. I don't believe that's the exact quote but it's one of the many interpretations.
Reg: Okay well this has been really fun and fascinating. You and I have known each other for long enough we could record this for another three hours easily and have the content but this has been really good. Any last thoughts before we finish up?
Ray: The mainframe is not going away anytime soon. Of course we don't know what will happen down the road, way down the road but I just want to say it's still an extremely viable platform. We have the most resilient and secure platform. We need to evangelize that more. We are getting better at it but again as this keynote another person said that they would go to college campuses and say we work mainframes. He works for a large financial services company and basically the response was either what's a mainframe or I thought those were gone. Until he tells them that, every time they made a charge on their credit card or used their debit card, probably 95% sure there was a mainframe involved somewhere along the way and we need to get this out there.
Reg: Yes, absolutely. Well, thank you so much for your contribution both in this podcast and in all the other ways you get that word out there.
Ray: Thank you very much, Reg. It's been a pleasure.
Reg: Definitely.