[Info-vax] Sharing about an old friend of the group.

[Danielle White]

I’ve never been active here but I have been around the edges for a long time, and a friend pointed me here due to my reminiscing about some AlphaServer and VMS things, such as the SRM process “dead_eater waiting on dead_beef.” However, my attention was then drawn by a familiar name: Bob Curley. I wish I had been here when information was shared by another poster at that university, whom I don’t believe that I ever met, because I had some additional data. I’ll share it now, both for the sake of the historic record and because nobody I worked with remains employed there for a mixture of retirement and moving to other positions.

I started work in the university’s library the mid-1990s. Their DRA library automation system ran on VMS and that server was hosted in the main computer room, thus I soon met Curley (as he was known there, a practice that came from his department of perhaps 20 people having a surprisingly high population of people named “Bob.”) It was during this time that I found the process I mentioned above; the university had a power-down over a holiday in the spring and I was in to do some maintenance work on the DRA library system that Saturday. The server, and AlphaServer 1000, depended on its cluster partners for some resources, and Curley’s work on those had gone slower than anticipated, so I was left with time to kill and an SRM console in front of me, about which I was curious, and explored.

In 2000 I joined his department, where I would remain until a few years after his retirement. I was somewhat surprised I was brought on. They had done so for me to take on the growing Windows install count but I also had earned a reputation for having suggested we consider Linux before it was politic to do so. I had been busted the year before for running a RedHat Linux 5 DHCP server to ease regular re-imaging of the library’s lab systems, preferring to automate that repetitive manual task as much as possible. The systems group had a great dynamic, possibly best demonstrated by the fact that we had decided to have our office doors keyed the same. We would carefully respect the privacy and use it only when necessary, such as when someone out that day had left in their office a manual that another needed.

Curley and I got on well enough. We had a lot of side conversations about tractors and chainsaws, owing to him buying a rural place in northeastern Pennsylvania and I having been a dairy farmer in my teens. We also talked a lot about motorcycles as he rode, as he’d had a Honda (CB360?) in the 1970s and then a BMW (R65) for over 20 years, and he told of riding in the Philadelphia area, when the well-worn cobblestone streets became even more challenging with accumulated wet leaves in the autumn. I was in the process of getting my motorcycle endorsement when I started in his department and still ride to this day. At one point he gave me one of his old DECUS polo shirts and, just before his retirement, a couple tech swag nick nacks he had amassed over the years that he didn’t care to haul home from his office. I remember that he wore the most garish shirt and tie combinations; a quiet protest of the university’s dress code. Invariably the shirts had a breast pocket and clipped to the edge of that would be an IAS pen. When we migrated from CI and the StorageWorks cabinets to one of the earlier SANs (pre-EVA, as I recall) he insisted we run the fiber in flexible corrugated plastic conduit under the floor. The coil of it that arrived was bright orange and he always referred to it as “tiger guts.”

Professionally we butted heads several times, mostly owing to his passion for the tech and me having a lot to learn, though occasionally it was having to convince him that the tech world was changing in some ways, like that Windows had some significant differences from Unix and VMS. The last time was memorable, though.

Curley had written several in house system utilities in Fortran. Our boss wanted them rewritten in C before Curley retired because nobody else knew Fortran. I had never done either at that point but decided to go for it. They say “fortune favors the bold” after all. Should that not hold I’d undoubtedly have a great story to share over drinks while I found my next job. In the course of this I realized he had made an assumption about the DCL command line which probably were true when he wrote the utilities but had since changed. I'm pretty sure he hadn't realized, possibly because with his approaching retirement he was winding down professionally. We had quite the argument about it. As that disagreement dragged on I realized his retirement party was later in the week so it was time to leave the battle and quietly implement my code after he retired.

Please don’t take that the wrong way: we were coworkers, in daily contact, so having those kinds of disagreements was part of the professional relationship. No matter how strong the disagreement there was always professional respect.

I recall a moment of him enjoying dark humor by retelling an exchange from a recent medical examination. He had a cardiac stress test and the doctor at one point said “I’m going to inject you with thallium now.” While necessary for the particular exam there was a local coincidence: around the time Curley was hired by the university a local murder case, mostly stagnate since 1991, moved forward. The victim in that case was also named Robert Curley and he was poisoned with thallium.

At Curley’s retirement party I remember our boss beginning the festivities with a question to me: “have you DISUSERed Curley’s account?” That affirmed, the sharing of stories began.

For various reasons our relationship never reached friendship and so his retirement party would be the last time I saw him. The last I would hear from him was a few weeks later: he and his wife had taken a trip, I believe to the Thousand Islands. He sent an e-mail with a couple photos of him kayaking, saying “the programming of the machine is simple:
while not tired:
  paddle on the right
  paddle on the left"

A few years ago, when I lived in North Carolina, I took a motorcycle road trip to New England, and spent a couple days in Scranton. I had some free time around lunchtime on a Friday and was staying at the hotel across the street from campus so stopped by my old department. I can still picture the hallways conversation with my former boss, who was a few days from retirement, and being informed that Curley had died. I also got a tour of the new computer room which was his design handiwork in 2009. It was nicely done.

If I had to say one thing I truly feel about Curley it would be this: I am a better system administrator for having worked with him.

With his retirement I became the VMS person there, so much as we then truly had one until they brought him back years later. I’ll never claim that I had the depth of knowledge that he did on the system. The phrase “any technology sufficiently advanced is indistinguishable from magic” applied; colleagues were amazed by the performance tuning I did on their batch jobs, like making one that had a twenty minute run time take only two, but nearly all of it was simple logic reorganization informed by the characteristics of VMS, like the impact of image rundown.

Over the next couple years I had a lot of VMS work to do. The biggest upgrade that I remember was moving the campus academic webserver onto Secure Web Services for OpenVMS. I’m afraid that I forget what it had been running before but we had periodic performance issues with the proper one that were solved by the new version.

Unfortunately, here is also the story of how I came to leave VMS.

One day, around the start of 2007, I was assisting a coworker with an issue on a Solaris email server because I wanted to build my skills on other platforms. Our manager stopped in the computer room to check on the status of the problem about that time. He then asked me to step out to discuss something with him, leading me to a closed door meeting in his office in which I was informed that I was not to do that and instead to focus on VMS to the exclusion of all other tech because that was our core operation, the single most important systems, and we were committed to it long-term. However, not long after, I was about to walk past the open door of a conference room and heard the same manager stating to a vendor that his goal was to get VMS out of the computer room.

Perhaps it was merely coincidence but a little later that year a newer coworker was promoted to senior ahead of me. The person promoted had no VMS knowledge which didn’t help the claim of valuing that skill set and long-term importance of the platform as core operation of the university. I also noticed that my coworkers were getting approval to go to a few conferences each year while every request I put in was denied. Overall I was reminded of stories both Curley and our boss shared with me earlier of what preceded the corporate axe falling on them at prior positions, plus what a family friend, an executive of a local bank, had told of his career experiences. 

By then hints of the upcoming financial crisis were showing so I knew I had to quickly make a choice. I decided to carefully ignore my boss’s direction and build Unix/Linux skills so I could move.

Though nobody directly said so, I’m certain that my departure was why Curley was brought back in 2009.

I truly enjoyed running VMS, but I saw that road coming to a bad end for me in that time and place. I also know that I am a better system administrator because of my years with VMS.

Happy Holidays

Danielle White