Orcmid's Lair: Computer Pioneer Bob Bemer

Orcmid's Lair

Welcome to Orcmid's Lair, the playground for family connections, pastimes, and scholarly vocation -- the collected professional and recreational work of Dennis E. Hamilton

Archives

Atom Feed

Associated Blogs

Edge City

Numbering Peano

Praxis101

Prof. von Clueless in the Blunder Dome

Recent Items

Cattle-Driving the Future

Rich Client, Poor Client, Smart Client, Web Client

Accountability and Transparency for Credibility

In an ALGOL State of Mind

The Ever-Shifting Tide of Communication and Civil Literacy

We Must Like Blogger Because ...

Satisfying Generated Wants

Mining the Torrent

The Inner Dork: Hero in My Own Mind

Opened Academic Middleware Shifts Culture

2004-06-25

Computer Pioneer Bob Bemer

I just learned of the death of Bob Bemer. I think he'd be pleased to know that he was slash-dotted. It is also odd the things people are remembered for, although Bob was tickled to make sure the ESC code was in ASCII, and other things. Only one person who knew Bob seems to have chimed up, and I added my footnote. I need to say it again in this quieter place.

I worked in one of Bob's teams while he was Director of Software at Sperry Univac in the 60's. He was a lot of fun. He kept calling me "Bub." I ran into him at standards meetings a few times after he moved on to GE and Honeywell, but we did not stay in touch. I re-encountered him later on the web, just prior to Y2K ,as the result of an article reporting that he was suggesting a Y2K repair that would not require people to remap existing file records. He wanted to pack the numbers tighter and intercept date accesses in running programs, buying some time.

I exchanged e-mail with him a few times in the last few years, and I had a chance to acknowledge the inspiration he was for me while he was still around. I don't know that he would hang out on slash-dot. When I last exchanged e-mail with him he was frustrated about what it took to maintain his web site.

I guess he was a geek at heart. I had produced a fast decimal-to-binary assembly-language algorithm for a character-oriented machine that didn't have a built-in converter but addressed in binary and calculated in decimal (making subscripting hard). He was the only one of his entire organization that worked it over and took more cycles out of it, and then I took out more using his ideas. He thanked me for giving him a chance to play.

Beside paying attention to those little details that can mean a lot, like character sets and extension techniques, he also worried about improving programming languages, training software developers, establishing software forensics, and making software engineering an activity that exploited reusable piece parts, anticipating components by a good 30 years. He funded Peter Landin and Bill Burge's work on Functional Programming in the US because he saw the possibility of applicative languages as the ultimate in piece-part composers.

Bob's web site is loaded with reminiscences. Some of Bob's recollections are a little off. When I arrived at 315 Park Avenue South, Al Paster was already there, working for C.L. McCarthy. Bemer hadn't shown up yet. I don't remember what the lag was. I know Bob was there in 1962, because I remember him talking to my wife, Bobbi, at an employee party that he initiated to have all of us become better known to each other. But who was already there in 1961 when I moved from Seattle to Manhattan was Bill Lonergan, one of the architects of the Burroughs 5000. Univac was definitely looking to make a move. (It was Lonergan's connections that led to my getting to work with Don Knuth in the summer of 1962. Along with building a Fortran compiler for us, Don said he had some ideas for a definitive book about programming.)

One thing I liked about Bob was he thought he hired me and he liked to tell people that. I was there ahead of him, but as far as he was concerned, I was his find.

Another fascinating thing is that he recreated his years at Univac by going over his expense reports for the time. Dang, I wish I had thought to keep mine over all this time. What a clever way to associate and recall events.

With regard to managing people, Bob didn't believe in burning out developers and he thought there was a lot of life to be had outside of the office. I'm pleased to learn that he was active to the end. I'll never forget him.

The key to immortality is living a life worth remembering.
-- end title to a Bruce Lee film
Who we are is the remembrance for the immortality of those who touch our lives
-- me

This little appendage is provided to force Blogger to see this article as having changed. There's really no new content. I am making this modification so that the Atom feed will supplant the previous modification that was published as a result of this defect.

posted by orcmid at 6/25/2004 02:05:50 AM

Comments:

Bob in Phoenix. I forgot that I also visited with Bob while in Phoenix for some sort of standards-related shindig. He was at GE/HIS at the time, and he invited some of us out to see his home on its little hilltop. I loved the view and the design of his home, but my greatest lust was for his bound volumes of Communications of the ACM. Bob was a solid ACM booster and had been the Standards Editor for a time. Every so often something would show up from him, including a Pracnique back when they featured those.

I couldn't stand the space (nor the expense) of bound volumes, so I did the next best thing and acquired Comm.ACM, J.ACM, and Computing Reviews on microfiche, along with a lot of IEEE and Stanford reports that were available in fiche format. Fortunately, a good chunk of that material is now on-line for ACM Digital Library subscribers, so I'm covered where my fiche subscriptions dropped off around 1986.

I no longer have a viewer and I thought that my hot-stuff HP Scanjet would provide an alternative. No such luck. The scanner's quantization is too course for good viewing of microfiche images. I figure there must be image processing that will do a good job of resampling the scans to good monochrome images, but I haven't found anything obvious. So much for reading an article in one display window while making research notes in another. I bet Bob would have known who to talk to, if I'd thought to mention my problem to him. Sounds like something to cover in nfoWare as an image-processing component demonstration. I'll have to find something to name "bemer," in Bob's honor and in a way that it isn't confused with German automobiles. Maybe I'll call it "Bub."

# posted by orcmid: 6/25/2004 11:42:15 AM