Lambda the Ultimate

Unexpected and very sad. He drove a lot of interesting discussion on LtU, and was exploring many interesting ideas. My condolences to his family.

Aside from being generally a good guy, driving discussion here and reminding us of some less conventional but valid views, he kept an awesome and interesting blog. It was some of his work (on Kernel) that convinced me some of mine wasn't entirely insane.

.

I have long enjoyed reading John's thoughts, and I enjoyed our conversations. John, you will be missed.

I am so sorry to learn of John's premature passing. He had a sympathetic disposition and I will miss his thinking and intuitions. Condolences to family and friends.

I first encountered his words here on LtU, and then his blog. I was so happy to have done so. (I occasionally read wikinews too but had no idea any of it was edited by him.)

I hope his family will come to hear and ponder how many tens (hundreds?) of thousands of people will have loved -- and will come to love for as long as the internet still carries his presence -- his words and exploratory spirit, even if we didn't know him, nor necessarily know that what we're reading was written/edited by him.

He has become a part of so much interesting thought, may it last forever. Condolences to all who were close to John.

John was the ultimate anti-Northeastern/Scheme view of Lisp.

While I never fully appreciated the benefits of vau calculus, since I prefer programs that have contextual equivalence, I did see some beauty in his line of thought: What if we push back meaning until we have an Environment to run a program? One consequence would be that argument evaluation and argument passing are fully decoupled.

May John's Environment in Heaven have a sound beta rule reduction and he get to meet his vau the ultimate creator.

John was one of the most interesting posters here on LtU for a ling time. His gentle and engaging style made many of the threads here a pleasure to read. RIP

There's not a lot that John and I agreed on technically.

Which, IMO, makes his passing, secondarily to the tragedy for his family and friends, a tragedy for people who share my thinking about the issues. Because whether I agreed with him or not, there was a clarity and a commitment to his view, as many of us saw expressed in Kernel, that sustained and elaborated on an important historical feature of the Lisp family of languages ("FEXPRs") while demonstrating that many of the perceived infelicities of this evaluation approach were not due to the feature itself, but rather to coupling with the historical commitment to dynamic scoping, which, thankfully, both Scheme and Common Lisp later abandoned.

John was the ablest defender of this tradition. Sadly, that's in large part due to the fact that he was essentially the only defender of this tradition. In spite of disagreeing with him technically, I can't help but be saddened to picture his passionate defense of this crucial idea fading into lore, and eventually winking out completely.

Godspeed, John.

So terribly sad. May his memory be a blessing.

I just found out the news from an email I received about Kernel. It's been some years since I worked on anything related but I still remember fondly my involvement in that era. I have email conversations with him going back to 2007 and even while busy with his dissertation he found time to entertain the questions of a complete stranger from half across the world.

I still remember reading his "Decomposing Lambda" paper and being hit with the simplicity and elegance of his presentation. I still hope to see someone pushing that line of thought even though history moved in another direction.