On one hand, sad news. On the other, you can't say he won't be remembered, or question that he did anything truly worthwhile in his life.
Most real computer work done since the mid 80s was directly benefited by C, be it by more robust and stable platforms to work on, quality portability without needing an interpreter, even if porting wasn't always graceful, and/or as a means of writing high performance applications before the likes of C++, Java, .NET (C->C++->Java->C#), etc., when Fortran and the like weren't good options. Today, the only language that might work out better than C for system programming might
be C++, and then only with a sizable coding standards document, practically reducing it to a type safe C.
Beyond that, it has been a solid base from which to develop new application programming languages, since both C, and derivatives of it, have well known properties, including general performance of interpreted and compiled code, ease or difficulty of optimized compilation, ease or difficulty of creating libraries to use various hardware and software systems, gotchas when it comes to language semantics as they occur in real code
, and so on.
It's not just coming up with one idea, either. It's sticking with it to keep its implementation good for many years, and still not losing sight of the fact that his most well-known work is a small part of the big picture, and that it was in the best interests of all that it be the best of that small part it could be, rather than taking over all things a system may need to accomplish. I will go on to say something that many a zealot may be offended by: much of what made Unix and C was not strictly novel. What has made it* special, and able to survive and evolve, was that its originators and early users had somewhat clear vision, and were willing to stick to that vision long enough for their work to get a real foothold before
the standards committees could try to butcher it, keep said butchering in check, yet at the same time, keep platform specific oddities from overwhelming commonalities, before standards can be fully agreed upon. This is not a balancing act for for mere mortals, can't be done with an iron fist, and DMR managed to pull it off long enough for C to become its own living language...the kind that lives in a sturdy brick house, uses coupons, and tells kids with their classes and lambdas to, "get off my lawn!" * so much of the design decisions of C reflect the needs of the original Unix OS, and the few neat features of the PDP-11, that I'm considering the original implementation of both as a tightly coupled unit, rather than an language used to write some arbitrary software, orthogonal to each other, which has become the reality of it, over time.