2005-10-26 #1

Computer Systems Research: Past and Future

Butler Lampson

He mentions that in 10 years we will have chips with 100 processors on them. This has me worried because I have yet to see (or maybe just understand) a viable programming model for how to manage this many processors.

He says that Functional Programming hasn’t worked but I would partially disagree. You are starting to see in increased use of functional programming techniques in new OO languages like Ruby and in some better Java Frameworks. (Not to mention Smalltalk.) Pure functional programming never took off but I think that Object-functional programming will be a big success.

I would argue that we can now move garbage-collection over to the success column and that “reuse” is now a reality, at least in some communities.

His slide on why Systems Research didn’t invent the Web is interesting. (Summary: too simple.)