Clear, Firm, Intelligent, Decisive, Kind and Gentle
Roger would have been seriously annoyed by all this: a whole lot of people talking and writing about him, rather than concentrating on important work that has been done, or needs to be done. He did everything he could to downplay the seriousness of his condition and keep it from becoming a topic or otherwise interfering with his duties at work or at home. But now that he has headed off to present his code to the Galactic Emperor, I hope that we can all be forgiven for needing to share our feelings of loss: he left us much too soon.
I read the words in the title above in a message written by Rob Hodgkinson on the Jprogramming forum [Jprogramming] Roger Hui – 1953 – 2021 (jsoftware.com), and I cannot think of a more fitting description of Roger, so I hope Rob will approve my reuse here.
I expect that almost everyone who worked with Roger would agree on the first four; I suspect a few people may occasionally have experienced the “Firm and Decisive” aspects of some e-mail communication as less than Kind and Gentle. In his post, Rob does go on to say “funny how that reminds me of both Ken and Roger alike”. As was the case with Ken Iverson, arguing with Roger always carried the risk, if you were not careful enough with the rope that was extended to you, of exposing that you hadn’t quite thought your arguments through.
Roger at Dyalog
Although Rob, Roger and all worked for I.P.Sharp associates back in the 1980’s, we were almost perfectly distributed around the globe at 12°E, 151°E and 122°W. I did not have an opportunity to work with Roger until he sent me a message with the subject line “business proposition” in 2008, proposing that he do some work for Dyalog. Obviously, we leapt at the chance, and enjoyed a very productive period at Dyalog, with the combined intellect of Roger and John Scholes guiding our work on language, before they both departed prematurely.
With Roger’s help, we were able to bring many of the ideas that he and Ken had worked on in J into Dyalog APL, and start to repair some of the wounds that opened up when the community split in two around 1983. Much to my surprise, Roger took an immediate liking to Scholes’ dfns, and the combined efforts of Scholes and Hui have been an important part of creating a foundation for APL to grow again, as a modern programming language. A joint presentation by John and Roger on “Potential Version 14.0 Features” demonstrates the results of some of this work.
Roger and Ken
As Ken started on J, he had Roger at his side to provide the clear, firm, intelligent, decisive, kind and gentle feedback – and the exceptional programming skills – that Dick Lathwell, Roger Moore, Adin Falkoff, Phil Abrams and Larry Breed had provided for the first APL implementation back in 1966. Although I was nowhere close to either of these projects, there is no doubt in my mind that this work made the difference between having a collection of very interesting ideas, and having tools that thousands of people decided to use to solve problems that were important to them.
Roger the Archiver
In addition to being intelligent and spending a great amount of time thinking about things, I believe that a key to his ability to be clear, firm and decisive was that Roger was extremely well read. He was familiar with a huge amount of significant writing about array languages.
Roger clearly felt this was important and tried hard to help the rest of us to benefit. Apart from code that he wrote, and the excellent documents that he authored himself, one of the most significant contributions that Roger has made to the community is his collection of digitized historical documents that have been made available on the jsoftware website, such as Community/Creative Commons – J Wiki (jsoftware.com).
As the doctors started to give him estimates for his remaining time, Roger ramped up his own writing and his work on collecting the writings of others. His biggest project was the paper on “APL Since 1978”, written for the fourth conference on the History of Programming Languages. I’m extremely happy to be listed as a co-author of this paper, but 90% of the work – and nearly all the serious content about the evolution of the language – was provided by Roger. Fortunately, Roger was still well enough in June of this year to be able to do an energetic presentation of this work.
Shortly after the HOPL IV conference, Roger’s health started to deteriorate. I was extremely lucky that the Canadian border was opened to fully vaccinated visitors at the start of September, which made it possible for me to visit Roger and do my best to thank him for his enormous contributions – and also ask whether there were any unfinished tasks we needed to take care of, such as archiving more documents.
According to Roger, it is done. Knowing his level of determination and dedication to any task that he took upon himself, we can probably rely on that. Roger frequently used the acronym “WWKD” (What Would Ken Do?) in discussions. Possibly, Roger ended up being better at answering that question than Ken himself. Now we also need to ask WWRD? Fortunately, Roger has done his best to help us find the answers.
We just need to start reading and get to work!
If I was going to add a couple of words to Rob’s list, it would be these two. Humour is very often the companion of intelligence, and this was definitely true for Roger. In addition to the collection of serious material, Roger was keen to collect humourous content related to his work and his colleagues, such as APL Quotations and Anecdotes (jsoftware.com). The subject line in his first message to me in 2008 (“business proposition”) was clearly a joke, and a dangerous one: if it had not all been in lowercase, I could easily have deleted it unread.
Roger was also a bit of a gourmet, and discovered that he shared this interest with Gitte Christensen, who enjoys cooking delicious meals for the inhabitants of the 5-bedroom house that Dyalog rents in Basingstoke, which until the pandemic was frequently full of visitors from distant places.
When I visited Roger and came home with his final input on my to do list (and perhaps more importantly, my NOT-to-do list). He also provided Gitte with a couple of his favourite recipes. I’m looking forward to remembering Roger by tasting the results!
Roger’s presentations often included jokes, but they were often so dry and sophisticated that they sailed over the heads of most of the audience. His daughter Rachel suggested that he try injecting some more “regular” jokes into his talks. I’m not sure whether that is the reason why Roger also published a collection of the Jokes We Told Each Other (jsoftware.com).
From the collection, we can see that Roger and Rachel had been working on each other for some time:
What does Silly Billy call his pet leopard? Stripes!
Now, what does Silly Jilly call her pet zebra?
When my daughter Rachel was about 4 years old, I asked her this, in that sequence, and to my gratification she answered without hesitation, Spot!