Interview With Donald Knuth
InformIT (04/25/08) Binstock, Andrew
Computer scientist Donald E. Knuth, winner of ACM's A.M. Turing Award in 1974, says in an interview that open-source code has yet to reach its full potential, and he anticipates that open-source programs will start to be totally dominant as the economy makes a migration from products to services, and as increasing numbers of volunteers come forward to tweak the code. Knuth admits that he is unhappy about the current movement toward multicore architecture, complaining that "it looks more or less like the hardware designers have run out of ideas, and that they're trying to pass the blame for the future demise of Moore's Law to the software writers by giving us machines that work faster only on a few key benchmarks!" He acknowledges the existence of important parallelism applications but cautions that they need dedicated code and special-purpose methods that will have to be significantly revised every several years. Knuth maintains that software produced via literate programming was "significantly better" than software whose development followed more traditional methodologies, and he speculates that "if people do discover nice ways to use the newfangled multithreaded machines, I would expect the discovery to come from people who routinely use literate programming." Knuth cautions that software developers should be careful when it comes to adopting trendy methods, and expresses strong reservations about extreme programming and reusable code. He says the only truly valuable thing he gets out of extreme programming is the concept of working in teams and reviewing each other's code. Knuth deems reusable code to be "mostly a menace," and says that "to me, 're-editable code' is much, much better than an untouchable black box or toolkit."
Click Here to View Full Article
No comments:
Post a Comment