Elwyn spent a lot of time at MSRI and this Numberphile video captures him doing something he loved: playing mathematical games.
Elwyn came from Kentucky and KET, the PBS affiliate there, did a really nice retrospective on his life and work.
Thanks for everything Elwyn.
eecs page: https://www2.eecs.berkeley.edu/Faculty/Homepages/berlekamp.h...
home page: http://math.berkeley.edu/~berlek/
math page: https://math.berkeley.edu/~berlek/index.html
It isn't terribly difficult to work through the algorithm mechanically and see that it solves the equations that it's intended to solve, but that is a long way from a good intuitive understanding of it. We spent some time going back through the original literature describing it and related contemporary work, but what we found also focused on a rather procedural description that didn't really strike the insight we were looking for.
At one point, I quipped that I should reach out to Berlekamp and inquire if this work really weren't the product of reverse engineering instruments found in a 1967 UFO crash-- given what a remarkable advance it was. :)
I'm sad to hear that the world will never know for sure now.
Although Berlekamp's work has already been important to the world-- impacting most of the communications and storage devices we interact with--, I expect that its importance will increase in the future.
Many advances in computer science and related fields have their utility gated or multiplied by the power and pervasiveness of computing technology. I am always struck when I read older papers-- even ones as late as the 90s-- and find them discussing problem sizes at the limits of their computer technology which my laptop is solving thousands of times per second as part of my application (or just as part of my research). So I find that my recent work using the berlekamp-massey algorithm casually applies to problem sizes that would have probably been unthinkable in the 80s, even using Cyclotomic's expensive specialized hardware. The increasing power and decreasing cost of general purpose computing increases the importance and broadens the applicability of the algorithms we run on it.
I had the opportunity to meet him while he was at Berkeley Quantitative, and it was an honor.
As an aside, not many in information/coding theory know that he was the advisor of Ken Thompson (of UNIX fame) according to Wikipedia: